Docstoc

LIMS Final Report Draft 28-2-06

Document Sample
LIMS Final Report Draft 28-2-06 Powered By Docstoc
					CONSULTATIVE INVESTIGATION
 INTO LOW INCOME MEDICAL
         SCHEMES


    DRAFT FINAL REPORT


     28 FEBRUARY 2006
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



ACKNOWLEDGEMENTS

The LIMS process was designed as an open and collaborative stakeholder process, and numerous
institutions and individuals have played a significant role in all aspects of the process. I would
like to acknowledge, firstly, all the many individuals who participated in the various Task Groups
and Sub-Groups. These various groups met many times during the course of the process, and
numerous individuals gave generously of their time, both in meetings and in preparation of work
outside of formal meetings.

I would particularly like to acknowledge the extraordinary efforts of the following individuals
who went beyond the normal levels of participation, either in leading a Task Group or Sub-
Groups and/or in working very hard on specific projects within these groups: Wimpie Du Plessis,
Etienne Dreyer, Michelle Scrutton, Karen van Rensburg, Anri vd Walt, Bernie Clark, Derry
Herron, Dan Nefolovhodwe, Michael Settas, Sarah Bennett, Emile Stipp, Pieter Grobler, Brian
Ruff, Alan Pollard, Colin Bullen, Christoff Raath, Roseanne da Silva and Lydia Footman.

The LIMS process would not have been possible without the generous support of the institutional
donors, who are listed in the Report. I would particularly like to acknowledge the very generous
support of the major donors who each committed at least R500,000 to the LIMS process. These
were Netcare, Netpartner, Innovative Medicines South Africa, Old Mutual Healthcare and
Discovery Health (which committed up to R2 million). There were several other generous
donations, including R200,000 from the Pharmaceutical Manufacturers Association, and
R100,000 each from the Hospital Association of South Africa and Medscheme.

Finally, I would like to thank various officials at the Council for Medical Schemes and the
Department of Health, both for their vision in initiating the LIMS process, and for their
enthusiastic support and contributions to the various discussions throughout the course of the
process. In particular, I would like to acknowledge the efforts of Alex van den Heever from the
Office of the Registrar of Medical Schemes, and the Registrar, Patrick Masobe.




Jonathan Broomberg
28 February 2006




                                                                                                1
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



INTRODUCTION

This is the final report of the Consultative Investigation into Low Income Medical Schemes
(LIMS), which was initiated by the Ministerial Task Team on Social Health Insurance (MTT)
during April 2005. The purpose of this report is to provide the MTT, the Council for Medical
Schemes, and all LIMS stakeholders with the final recommendations of the LIMS process, as
well as the supporting arguments and evidence for these recommendations. Fortunately, the LIMS
process has produced a wide consensus on almost all of the relevant issues contained in the
recommendations. This is not true for all issues however, and as required by the Terms of
Reference of LIMS, the report specifically identifies those areas where consensus was not
reached. In these cases, the debates on both sides are summarized, and where relevant, supporting
arguments or documents are included.

Section 1 provides a brief review of the LIMS process. This review covers the history, objectives,
principles, policy framework, and modus operandi of the LIMS process. It also summarises the
agendas and activities of the three major LIMS Task Groups and their respective Sub-Groups,
and provides a report on the timetable and interim outputs of the LIMS process, as well as the
financial aspects of the process including donations and expenditures to-date. Section 2 outlines
the key findings of the various research activities and consultations undertaken, and references
the full reports of these, which are included as Annexures. Section 3 provides the
recommendations of the LIMS process, which are based on both the research findings and the
consultations undertaken during the process.




                                                                                                2
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



SECTION 1: REVIEW OF THE LIMS PROCESS


1.1 HISTORY AND OBJECTIVES OF THE LIMS PROCESS

The LIMS process was formally established through a Terms of Reference issued on 23rd April
2005 by the MTT (Annexure 1), and was launched at an open Public Forum on 4th May 2005. As
summarised in the Terms of Reference, the LIMS process is very clearly located within the broad
context of the Social Health Insurance (SHI) policy framework being pursued by the Department
of Health (DOH) and the MTT. A key objective of the SHI policy is to extend risk pooling
through health insurance as broadly as possible, over time, to all those who can afford to
participate in formal risk pooling arrangements. The full SHI policy framework requires a
combination of risk pooling arrangements (via the Risk Equalisation Fund (REF)) and income
cross subsidies (via the proposed SHI tax and subsidies to low income participants in risk pooling
arrangements). While there has been substantial progress towards the risk pooling component of
the SHI framework, as evidenced by the recent adoption by Cabinet of the REF policy, and the
initiation of the shadow testing period for the REF, it appears that there will slower progress
towards implementation of the required income cross subsidies. As a result, it now appears that
the full implementation of the SHI framework will take significantly longer to implement tha n
was originally envisaged.

As a result of the delay in extending risk pooling arrangements via the full SHI framework, the
MTT identified the need to implement interim measures to extend risk pooling through health
insurance coverage as far as possible while the balance of the SHI framework is negotiated and
implemented. This need to extend risk pooling to those currently unable to afford existing
medical scheme cover provided the specific impetus for the LIMS process.

As identified in the Terms of Reference, the specific objectives of the LIMS process were to:

   Investigate the barriers to low income participation in medical schemes including:

    o   Household affordability and willingness to pay
    o   Employer attitudes and willingness to pay/ provide subsidies
    o   Distribution and other obstacles on the demand side
    o   Supply side behaviour and other systemic constraints
    o   Drivers of high cost within the current private healthcare system
    o   Regulatory obstacles which have a direct impact on affordability
    o   The subsidy framework
    o   Any other obstacles




                                                                                                3
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   Investigate all potential policy options, with supporting technical motivations.

   Review options required to ensure that the resulting market is diverse and does not reflect the
    elements of market concentration prevalent in the current health industry.

   Review potential Governance models, taking due cognizance of proposals put forward by the
    Department of Health and Council for Medical Schemes.

   Identify specific areas of difference between relevant stakeholders.

   Attempt as far as possible to identify where a consensus has occurred on a specific issue.

   Collaborate with independent research projects; and

   Consult with all groups relevant to this issue, with a special emphasis on:

    o   Medical schemes;
    o   Administrators and managed care companies;
    o   Employers;
    o   Trade unions;
    o   Low income employees;
    o   Informal sector workers;
    o   Medical services providers; and
    o   Government stakeholders.

The „target population‟ for the LIMS investigation was identified as households with gross
monthly income of R2,000 - R6,000 per month, since current data indicate low levels of coverage
in these households, despite the fact that in the majority of such households there is at least one
household member with a formal sector job, and in many households, at least one member
already belongs to a medical scheme.

As also noted in the Terms of Reference, the LIMS process is not, in itself, a policy development
process, since this is the prerogative of the CMS and Government. Instead, the purpose of the
LIMS process is to provide the policy makers with a well researched and widely consulted set of
proposals for their consideration in the development of specific policy proposals, and to identify
areas of consensus and disagreement in relation to the various proposals.

In further discussion with the MTT, it has emerged that the intention of the MTT is to develop
proposals and implement any required legislative and/or regulatory changes in time for
implementation in early 2007. As the market typically requires at least 6 months of research and



                                                                                                 4
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



development time prior to launch of new products, the aim would be to have all necessary
changes agreed during the second half of 2006.

1.2 PRINCIPLES AND APPROACH OF THE LIMS PROCESS

Following the requirements of the Terms of Reference for detailed research as well as wide
consultation and development of consensus where possible, the LIMS process was based on the
following key principles:

   Discussions and proposals should be research driven and evidence based wherever possible:
    This approach required reviews of existing evidence, and where necessary, the
    commissioning of extensive new empirical research. These tasks were undertaken within the
    various task groups, and are reported on further below. It is worthwhile noting that several
    ground-breaking pieces of new empirical research were undertaken for the LIMS process.
    These include:

    o   A large scale national household survey of over 3500 low income households, focusing
        on current healthcare expenditure and utilization, as well as on preferences and
        willingness to pay for health insurance cover.
    o   A detailed literature review covering international experience with the extension of health
        insurance coverage to low income households.
    o   An analysis of all existing demographic and statistical data of relevance to LIMS from
        STATS SA and other statistical sources.
    o   A demand model estimating the likely take up of health insurance coverage amongst low
        income households, assuming various levels of premium, employer and National
        Treasury (NT) subsidies, and household affordability constraints.
    o   Interviews with high level executives of 40 large, medium and small scale employers on
        their current approach and attitudes in relation to health insurance coverage for low
        income employees.
    o   A capacity model to test the impact of an expansion of health insurance coverage on the
        supply side of the healthcare market, including in relation to nurses, health professionals
        and hospital beds.
    o   An analysis of the impact of the 2006 changes to the tax treatment of medical scheme
        contributions on health insurance coverage amongst low income households.

   The process should be open and fully transparent: All Task Groups, Sub-Groups or other
    processes were always open to any willing participant, and all documents, research papers,
    minutes of working groups and other documents were at all times in the public domain.
    Where possible, these documents were uploaded to the website of the Council for Medical
    Schemes, and in all cases, these were widely disseminated to all stakeholders via email.



                                                                                                 5
               Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



      The process should involve extensive and broad consultation with all relevant stakeholders:
       This was achieved through the open nature of the Task Groups, the initial public forum, the
       wide dissemination of LIMS updates and the First Interim Report, and a wide range of
       specific stakeholder consultations. It should be noted, however, that not all of the consultative
       processes were as extensive or complete as would have been desirable. In particular, while
       there have been some formal consultation processes with or ganized business and the trades
       unions, these have thus far been limited in extent, and it is hoped that these stakeholders can
       be brought into the debate more actively following the publication of this report. In addition,
       there has unfortunately been no formal consultation with organizations representing civil
       society thus far, and it is also hoped that such engagements will be arranged during the policy
       development process that will follow the publication of this report.

1.3 POLICY FRAMEWORK AND MODUS OPERANDI OF THE LIMS
    PROCESS

In order to address the Terms of Reference and to be consistent with the principles and approach
described above, three broad work streams (and detailed objectives for each of these) were
identified in discussions with the MTT. It was envisaged that a Task Group, open to all interested
stakeholders, would be established to undertake a combination of research, discussion and
consultation for each of the work streams. This overall architecture for the LIMS process, and the
detailed objectives of each Task Group, were presented for discussion at the first LIMS Public
Forum.

Membership of the Task Groups is shown in Annexure 2. 1 The agendas and activities of each
Task Group were as follows:




1
    T he Annexure shows the full list of those volunteering for each Task Group at the outset of the process, as well as those added
during the process. In reality, only a subset of these initial volunteers has been actively involved in each T ask Group and Sub-Group.
It is possible that some participants in one or more of the Task Groups have been left off these lists, due to the frequent changes in
Task Group participation over the 10 months of the LIMS process.



                                                                                                                                     6
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



1.3.1   Demand and Distribution Task Group

This task group was led by Bernie Clark, Derry Herron and Jonathan Broomberg. It was tasked
with examining all relevant demand side obstacles to extending health insurance to low income
households, and proposing solutions to these obstacles. It was also tasked with estimating the
likely demand for low income medical schemes under different scenarios, and to examine issues
pertaining to the distribution of low income medical scheme products to low income households.

The overall agenda of this Task Group was as follows:

For all target households (R2,000-R6,000 per month), gather relevant data on:

   Current medical schemes coverage.
   Current healthcare utilisation patterns and preferences.
   Preferences regarding health insurance packages.
   Assess ability and willingness to pay for health insurance, and likely demand for health
    insurance packages amongst low income households.

For employers, gather relevant data on:

   Current coverage of workforce.
   Current subsidy policies.
   Extent of take-up of subsidies and reasons for this.
   Ability and willingness to extend subsidies to currently uncovered low income employees.

Distribution Issues:

   What are most effective and appropriate mechanisms for distribution and sponsorship of new
    medical scheme products to the targeted low income households?
   What are the roles of brokers, employers, stokvels, retail outlets, service providers etc in
    distribution of these products?

In order to address this agenda, volunteers in the Demand and Distribution Task Group met
during May 2005 and agreed to break into two Sub-Groups, one focusing on employer issues, to
be led by Bernie Clark, and one focusing on Distribution issues, to be led by Derry Herron. In
addition, it was agreed that Jonathan Broomberg would coordinate the development of the
household survey, as well as the commissioning of analysis of existing datasets and any other
baseline data required. The two Sub-Groups met several times during the LIMS process.




                                                                                               7
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



The following activities were undertaken under the auspices of this task group:

   Analysis of existing statistical data sets, including Stats SA Household surveys, Census data,
    Council for Medical Schemes data and other commercial datasets of relevance.

   A national household survey of low income households focused specifically on LIMS related
    issues.

   Inclusion of relevant questions related to employers in the Old Mutual Healthcare survey.

   Interviews with senior executives of 40 employers on issues relating to current and future
    subsidy policy, and take up of medical scheme coverage by low income workers and their
    families.

   Structured consultations with unions and organized business representatives.

   Focus group with brokers working in the existing low income medical scheme market.

   Development of a financial model to estimate likely broker distribution costs for low income
    medical scheme products.

   Development of a demand estimation model to estimate the likely take up of medical scheme
    cover by low income households, using various assumptions as to the cost of a low income
    package, employer and National Treasury subsidies, and household affordability constraints.

1.3.2   Benefit Design, Governance and Regulation Task Group

One of the key assumptions to be tested in the LIMS process was whether the high cost of current
medical scheme cover is one of the fundamental obstacles to extending health insurance coverage
to low income households. In addition to the high costs of medical services and scheme
administration, a key possible driver of the high costs of medical schemes relates to the high cost
of the prescribed minimum benefits (PMBs) that all medical schemes must fund in terms of the
Medical Schemes Act.

One possible policy option, which LIMS therefore investigated, is the establishment of separate
medical schemes, or separate options within medical schemes, open only to low income
households, which could be subject to a different regulatory environment, including a lower cost
prescribed minimum package. This would, naturally, reduce the minimum cost of such scheme s,
even in the absence of reduced costs of medical services, administration etc.




                                                                                                 8
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



The Benefit Design, Governance and Regulation Task Group was tasked with focusing on various
aspects of these potential new schemes (hereafter referred to as LIMS schemes) including issues
such as the nature of a revised approach to a prescribed minimum package for LIMS schemes
(referred to hereafter as the LIMS minimum package, or LMP), the governance of such schemes
in order to ensure adequate „voice‟ for low income members, and critically, how such options or
schemes might be adequately demarcated from the current medical schemes environment.

The overall agenda for Benefit Design, Governance and Regulation Task Group was defined as
follows:

Structure of LIMS schemes and benefit design:

   Should there be a separate class of LIMS schemes and/or options within schemes that would
    be exempt from the current PMB requirements of the Medical Schemes Act?

   What would the pros and cons of such an approach?

   If separate LIMS schemes and/options were to be recommended, what would the nature of a
    revised minimum package for LIMS schemes be?

   What services should be included and excluded from the minimum package?

   What would such a minimum package cost?

   How flexible should LIMS minimum packages be – based on uniform, regulated “modules”
    or entirely flexible?

   How to create consistency and compatibility with the REF?

Demarcation of LIMS schemes

   Demarcation of LIMS Schemes from the current medical schemes market.

   How to avoid adverse risk movements between LIMS schemes and current market.

Governance Issues

   Should LIMS schemes be new schemes only, or should existing medical schemes be able to
    register LIMS options?
   Should there be particular governance requirements for LIMS options/schemes?



                                                                                             9
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report




In order to address this agenda, volunteers for the Benefit Design, Governance and Regulation
Task Group met during May 2005 and agreed to break into two Sub-Groups, one focusing on
Benefit Design and Demarcation issues for the proposed LIMS schemes, led by Brian Ruff and
Silvia Cornejo, and the second focusing on Governance issues in LIMS schemes, led by Michael
Settas. The two Sub-Groups met several times since the beginning of the LIMS process. In order
to expedite progress and as a result of the technical nature of much of its work, the Benefit
Design and Demarcation Sub-Group set up two small technical working groups, one to examine
demarcation issues and the other to examine benefit design issues in LIMS schemes. Both of
these were convened and led by Jonathan Broomberg. These technical groups comprised
volunteer members of the Sub-Groups, as well as additional invited experts from the industry.
Participation in these technical groups, as with all other groups, remained open to all who wish to
participate.

The following activities were undertaken under the auspices of this task group:

   Extensive internal technical analysis and discussions, particularly with reference to the
    findings of the Household Survey.

   Iterative discussions with representatives of the MTT and CMS on key policy issues.

   Detailed costing analysis of the proposed LMP.

1.3.3   Supply-Side Task Group

In addition to the investigation into the possibility of a revised approach to the PMB requirements
for LIMS schemes, the LIMS process assumed from the outset that significant and sustained
reductions in the cost of medical goods and services and scheme administration will be critical if
the overall costs of coverage are to be reduced to a level which is affordable for a meaningful
number of low income households. The Supply Side Task Group was thus established to
examine all aspects of the current high costs of medical services and to examine approaches to
reducing these on a sustainable basis, while still ensuring profitable participation in this
potentially large and growing market by providers of goods and services.

Key issues which were examined by the Task Group included regulatory obstacles to the
emergence of lower cost services, how to achieve sustainable cost reductions within the current
structure of the provider market, the role of alternative delivery models in reducing the cost of
supply, and the ability of the supply side to cope with a significant expansion of demand that may
arise from the establishment of LIMS schemes.




                                                                                                10
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



The overall agenda of the Supply Side Task Group was defined as follows:

Reducing cost of supply in current market structure:

   What can be done to significantly reduce the costs of all medical goods and services on a
    sustainable basis without assuming dramatic short term changes in structure of the provider
    market?

Regulatory obstacles to reduced costs:

   What are the current regulatory obstacles to reducing the cost of health services and what
    changes would be required to address these?

Alternative Delivery Models:

   What alternative delivery models could feasibly be implemented in order to sustainably
    reduce cost?

   What are the obstacles to the development of such models?

Ability of supply side to cope with increased demand:

   How will all elements of current supply cope with a substantial increase in demand?

   What are the limits of current private sector capacity?

   What can be done to extend capacity to deliver services to LIMS members should this be
    required?

In order to address this broad agenda, the Supply Side Task Group, led by Wimpie du Plessis,
assisted by Etienne Dreyer, has been working in four Sub-Groups. One, led by Karen van
Rensburg focused on regulatory issues. The second, led by Michelle Scrutton, addressed the
issues of supply dynamics and the impact of likely growth in demand on the supply of services. A
third group, led by Wimpie du Plessis, interacted directly with various provider organisations to
examine the scope for sustainable cost reductions within the current provider market. A fourth
group, led by Anri van der Walt, examined the potential role of alternative delivery models in
reducing the cost of supply of healthcare goods and services. All of these Sub-Groups met
numerous times during the LIMS process.




                                                                                              11
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



The following activities were undertaken under the auspices of this Task Group:

   Extensive review of existing legislation.

   Discussions between working group and representatives of key provider organizations.

   Internal technical discussions.

   Development of a detailed supply side model to estimate likely impact of expanded medical
    scheme coverage on all parts of the supply chain, including nurses, doctors, other health
    professionals and hospitals.

1.4 TIMETABLE OF THE LIMS PROCESS

The key milestones of the LIMS process were as follows:

   23 April 2005: Terms of Reference issued by MTT and CMS and Coordinator appointed

   4 May 2005: First Public Forum to launch the LIMS process

   19 August 2005: Policy workshop with MTT/CMS. This workshop was held as a small,
    closed workshop attended by senior officials from the MTT, CMS, and DoH, and the LIMS
    Task Group and Sub-Group leaders. The aim of this session was for the LIMS task groups to
    present headline findings and conclusions that were available at that stage, and to discuss and
    debate these with the policy makers. This resulted in very valuable interaction and detailed
    discussions were held on some key conceptual issues, including issues such as:

    o   The acceptability of a revised approach to a minimum benefit package for LIMS
        schemes, and the nature of such a revised approach.

    o   Demarcation between LIMS schemes and current medical schemes.

    o   Governance issues for LIMS schemes, including the question of whether LIMS schemes
        must be new standalone schemes, or could also be options within existing schemes.

    o   The acceptability of formalized price differentiation from suppliers to the LIMS market.

   7 September 2005: Interim Update to LIMS Stakeholders published

   31 October 2005: First Interim Report circulated to Stakeholders for comment



                                                                                                12
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report




   1 December 2005: Interim Update to LIMS Stakeholders published

   12 December 2005: Final LIMS Research Report

Discussions with MTT and CMS officials suggest that the following process will take place
subsequent to the publication of this report:

   28 February 2006: Draft Final report circulated for comment

   Early March 2006: Second Public Forum to discuss Draft Final Report

   Late March 2006: Final LIMS Report published

   Late March/early April 2006: Final Policy Workshop to be convened to discuss Final LIMS
    Report and make policy recommendations to the MTT and CMS. This workshop will be
    under the auspices of MTT and CMS, and will thus no longer be under the jurisdiction of the
    LIMS process.

1.5 FINANCIAL ASPECTS OF THE LIMS PROCESS

1.5.1   Funding commitments, budget and expenditure to date

As noted above, it was always the intention of the LIMS process to base recommendations on
well researched evidence and data. To this end, substantial efforts went into raising funds from
private sector stakeholders, and several organizations responded very generously. This section
provides an update on funds raised, the proposed budget for LIMS research, and actual
expenditures against the budget to date.

As the table below indicates, commitments of R4.425 million were pledged to the LIMS process.
The table also shows that the total expenditure to date (including some expenditures incurred but
not yet paid) has been R2,879,504 against the budgeted total amount of R3, 345,000.




                                                                                              13
               Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



              Fundi ng Commitments by LIMS Donors
                                                          Rands
    Discovery Hold ings                                2,000,000
    Innovative Medicines SA                              500,000
    Mediscor                                              25,000
    Medscheme                                            100,000
    Netcare                                              500,000
    Netpartner                                           500,000
    Old Mutual Healthcare                                500,000
    PMA                                                  200,000
    HASA                                                 100,000
    Total                                              4,425,000


    Budget and Expendi ture to date: 28 February 2006
                                                           Initial         Revised
                                                          Budget            Budget     Expendi ture          Vari ance
    Literature Review and related research                 75,000           75,000            74,362                 638
    Household Survey                                    2,400,000        2,524,260         2,444,458            79,802
    Emp loyer Focus Groups                                320,000          150,000           120,000            30,000
    Additional Research and contingency                   550,000          550,000           240,684          309,316
    Total                                               3,345,000        3,299,260         2,879,504          419,756


Notes:
      The budget for the household survey was revised upwards due to a much greater expenditure of time
       on development of the survey instrument than originally budgeted. This has led to an increase in costs
       fro m the consultants assisting on the survey. The budget for employer survey was revised downwards
       due to scaling down of the originally envisaged survey, which was limited to 40 emp loyers only.
       Expenditure in the Additional Research and Contingency category was on the costs of two
       international consultants who participated in the development of the household survey instrument, and
       the consultancy work on the database analysis, the preparation of the LIMS submission to NT re the
       tax proposals, and the demand estimat ion model. The total expenditure in this category was below
       budget due to less work being co mmissioned than was originally envisaged.


      The expenditures shown above include some which have b een incurred but not yet paid out. Some
       small additional expenditure is still anticipated between the date of publication of this report and the
       finalisation of the LIM S process. A final reconciliation will be provided to all donors, and detailed
       accounts for all expenditure are available for inspection. To the extent that there are any funds left over
       once all expenditures are finalized, these will be refunded on a pro rata basis to the donors who have
       thus far provided R500,000 each to the LIMS process.




                                                                                                               14
             Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



1.5.2     Operating principles for expenditure against research budget, and for draw downs
          from donors

The following principles for utilizing the funding commitments from donors were agreed with the
donors:

   As and when funds were needed, these were drawn down, in equal tranches, from all of those
    who had committed funds.

   To the extent that new donors made commitments after draw downs had taken place, the next
    draw down was made against new donors first, until they had reached the same level of
    participation as existing donors, subject to their own maximum commitment.

   Once any organization‟s commitment ceiling was reached, they no longer participated in any
    new draw downs.

   Once all requirements were met, no further draw downs were made.

The draw down schedule below indicates the total amounts drawn down from each of the donors,
and the remaining balances against the total commitments from each donor.

Draw down schedule
                                                      Current          Ceiling           Balance
                                                    drawdown       Commi tment         remaining

Discovery Hold ings Limited                         500,000.00      2,000,000.00     1,500,000.00

Old Mutual Healthcare                               500,000.00       500,000.00                 0.00

Innovative Medicines SA                             500,000.00       500,000.00                 0.00

Mediscor                                              25,000.00        25,000.00                0.00

Medscheme                                           100,000.00       100,000.00                 0.00

Netcare                                             500,000.00       500,000.00                 0.00

Netpartner                                          500,000.00       500,000.00                 0.00

PMA                                                 200,000.00       200,000.00                 0.00

Hospital Association of South Africa                100,000.00       100,000.00                 0.00

Total                                              2,925,000.00     4,425,000.00




                                                                                                       15
                Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



SECTION 2: FINDINGS OF                                                         LIMS              RESEARCH                       AND
CONSULTATION ACTIVITIES

2.1 LITERATURE REVIEW ON LOW INCOME                                                                PARTICIPATION                     IN
    VOLUNTARY HEALTH INSURANCE SCHEMES

LIMS commissioned a detailed literature review in order to survey the current state of knowledge
and experience with regard to expanding access to health insurance to low income communities,
in both developed and developing countries. It was commissioned in order to highlight any
experiences on which South Africa could draw as it develops its own policy approaches. 2 South
Africa is one of the few low and middle income (LMIC) countries to have a well-developed and
regulated private health insurance (PHI) market, and so whilst micro-insurance schemes may hold
some lessons for the process, the main focus of the review is on the expansion of large-scale and
formalised private health insurance. In particular, the review considers the demand side, supply
side and regulatory challenges faced by countries that have attempted to expand health insurance
coverage to low income populations.

This section briefly summarises the key findings of the Literature Review. The full literature
review is provided in Annexure 3.

       Healthcare can be financed by out of pocket (OOP) payments or though pre-payment
        systems. Pre-payment systems include general taxation, SHI and voluntary or private health
        insurance.

       In developed countries, governments generally have the ability to collect sufficient revenues,
        through some combination of general taxation, SHI or other social security systems, to
        finance the healthcare demands and needs of most or all of the population.

       In the majority of middle and low income developing countries, governments are not able to
        mobilise sufficient revenues to provide adequately for the healthcare needs and demands of
        their populations. However, the underlying demand for adequate quality healthcare remains.
        Thus the ability and willingness to pay for healthcare in most developing countries is far
        greater than the capacity of governments of those countries to mobilise resources through
        various forms of taxation.

       This unmet demand tends to be satisfied through OOP expenditures by households. This
        pattern is confirmed by the relatively high levels of OOP on healthcare in developing


2
    T he literature review was carried out by Health and Development Africa (Pty) Ltd, and has been edited for purposes of this report.




                                                                                                                                      16
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    countries, where such expenditures may account for up to two thirds of total health
    expenditure. By contrast, the average for developed countries is approximately one third.

   OOP expenditures on healthcare are amongst the most regressive forms of financing, since
    they do not allow households to benefit from any form of income or other risk related cross
    subsidies. As a result, where OOP expenditure is a dominant form of financing, low income
    households tend to spend a disproportionately high percentage of their income on healthcare.

   These observations have stimulated a growing body of research and policy analysis that is
    promoting the role of various forms of private health insurance as a means of extending risk
    pooling and pre-payment in developing countries that are characterised by some of the
    features outlined above. Some specific arguments in this regard are:

    o   PHI can provide access to financial protection and additional resources for healthcare,
        where collection of general taxation and/or additional social security contributions is
        difficult.

    o   With appropriate regulation and cross subsidies, PHI can be used to overcome income
        related inequalities in healthcare.

    o   PHI can provide consumers with greater choice of providers and services, and has the
        potential to free up public sector capacity, as well as to promote innovation and
        efficiency, and to encourage the retention of a professional workforce within developing
        countries.

    o   Countries with high rates of PHI tend to have lower levels of OOP expenditure on
        healthcare, suggesting that PHI may substitute for OOP and assist in moving the
        healthcare system towards more equitable financing.

    o   PHI is recommended by some analysts as a method for moving towards universal
        coverage for countries with large informal sectors and relatively constrained tax
        resources.

   Different forms of PHI appear to be more or less suitable in different economic and
    developmental circumstances. One form of PHI is community health insurance, in which
    small groups form to share health related risks, with some form of re-insurance arrangement
    to pool these risks across the small groups. There is a growing body of literature and support
    for this approach in some low income settings. This approach has relevance in South Africa,
    particularly at the lower ends of the income spectrum, and is the subject of a parallel
    investigation to LIMS, led by the Social Re project.



                                                                                               17
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   However, South Africa is relatively unique amongst developing countries in having a well
    organised and well regulated formal PHI system, based on not-for profit medical schemes.
    The formal PHI system is already well entrenched, and could provide an efficient and
    effective platform for the expansion of health insurance coverage to low income households
    without current cover.

   South Africa therefore meets several of the criteria identified by policy makers as creating the
    appropriate conditions for the expansion of healthcare coverage through the PHI system.
    These include:

    o   A well organised and regulated existing PHI system.

    o   Significant demand for healthcare cover that is not met through current public financing,
        as demonstrated by high OOP expenditures.

    o   A well developed government plan to develop an SHI system over time, but obstacles to
        extending this in the short to medium term.

   Studies on demand for health care amongst low income groups suggest that where there is
    high OOP expenditure, this will facilitate willingness to pay health insurance premiums.
    However, the key to stimulating the demand for health insurance is to set premiums at rates
    close to consumers‟ expected OOP expenditures. This suggests that unless premiums are
    suitably low and closely related to expected OOP expenditure, there will be limited demand
    for PHI, since premiums will be unaffordable, and households will tend to continue with their
    current OOP expenditure patterns.

   Very few countries have historically used PHI to expand coverage to low income groups, and
    there is thus limited experience for South Africa to draw on in the development of this
    approach. This is because in most developing countries, even though demand exists and
    public systems are under funded, PHI systems are insufficiently developed and inadequately
    regulated. In the majority of developed countries, on the other hand, PHI systems are well
    developed and regulated, but publicly financed healthcare systems are usually adequate to
    meet healthcare demands of the population. The USA is the exception to these observations.
    It has a well developed and regulated PHI system, but no adequate public financing
    mechanism. As is well known, tens of millions of US citizens lack adequate healthcare cover.
    For this reason, US states have experimented extensively over the years with various
    mechanisms for using PHI to extend cover to low income groups. Some of the approaches
    and instruments summarised in this review are based on that experience.




                                                                                                 18
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   Efforts to expand cover to low income groups need to recognise that low income populations
    are not homogenous, and that there may be diverse groups within the target population that
    require targeted interventions if they are to be successfully brought into cover.

   There are many entry points and channels to reach the uninsured within low income
    populations. Employers play a key role for several reasons: they can facilitate access to cover
    and provide information to employees; group cover should also be less costly to distribute
    and support, due to economies of scale. Collection of premiums is also much easier in
    employer based groups. However, employers cannot be the only entry point to cover in the
    SA context, due to the high levels of informal employment, as well as the important role
    played by other social structures.

    There are a number of strategies for reducing the cost of PHI for low income households.
    These include direct and indirect government subsidies, product design strategies, and
    strategies to reduce the cost of healthcare goods and services.

   Direct subsidies can take the form of premium subsidies to employers who cover their
    employees, or premium subsidies directly to individuals. The US experience with direct
    subsidies to employers has not been favourable. Employers tend to fear that these subsidies
    will be temporary. Also, it appears that a high proportion of the target market do not
    necessarily work for larger, better organised employers who obtain access to these subsidies.
    It is also complex to design and target the subsidies, since proxies such as employer size,
    used to set the subsidies, appear to not to correlate that well with percentage of low income
    employees in the workforce.

   Direct subsidies can also be paid directly to low income members of medical schemes. This
    could be in the form of a fixed amount per member, or of a fixed percentage of the premium
    paid by the member. This latter approach, which has been implemented in Australia, tends to
    promote inequity, and would also run counter to the current trend of tax reforms regarding
    medical scheme contributions in South Africa. The former approach may however be very
    well suited to the South African environment, since it will result in direct reductions in
    premiums of low income households, and can thus be expected to increase demand for
    medical scheme cover, provided that the net premium is affordable.

   Indirect subsidies, using the taxation system, are an alternative to the direct subsidy approach.
    Tax incentives of one form or another have been used in almost every country with a PHI
    system, and have the advantage of using the existing tax system, thereby minimizing
    additional administrative requirements. Tax incentive instruments include tax deductibles and
    tax credits. The tax deductible approach is used in SA. In the case of tax credits, the credit
    can apply even where an individual has no or low tax liability. The tax deductible approach



                                                                                                  19
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    often has limited impact on low income households, since the household usually has a very
    low tax liability, and tax deduction thus has no impact on the high cost of the premiums. On
    the other hand, tax credits can directly reduce the cost of premiums for low income
    households. The use of tax credits would be a new approach in South Africa, and might
    involve substantial administrative complexity. The extensive literature on tax credits also
    raises questions as to the cost effectiveness of this approach.

   Another form of indirect subsidy which has been attempted in the US is the development of a
    reinsurance system aimed at providing stop loss insurance for schemes catering to low
    income members. Such re-insurance could be provided directly by government, or by the
    private market within a regulated framework.

   Overall, the evidence suggests that in order to be cost effective, subsidies must be carefully
    designed and targeted. Where this is the case, such approaches can have substantial impact,
    but where subsidies are poorly targeted, this can be an expensive approach with limited
    impact on overall coverage targets.

   In addition to subsidies, making private health insurance affordable for low income
    populations requires intervention at the level of product design. A key issue here is the trade -
    off faced by government, between the affordability and the comprehensiveness of cover. As
    in SA, the US has experience of regulated minimum benefit packages which have created
    barriers to affordable healthcare for low income groups. In response to this, some states,
    including New York, have implemented arrangements in which schemes aimed at low
    income groups are exempted from the minimum benefit requirements. This is precisely the
    dilemma faced by SA policy makers as they confront the question of the trade-of between the
    comprehensiveness of cover offered by the PMBs in the current medical scheme market, and
    the affordability barrier that these PMBs create for low income households. This is reflected
    in the current debate within LIMS, as to the pros and cons of a different minimum package
    for LIMS schemes.

   Other aspects of product design that can promote lower costs and more affordable packages
    include limited provider networks, other managed care interventions, and cost sharing
    approaches. In the SA context, where there is evidence that demand for health cover may be
    specifically related to OOP expenditure for day to day care, the use of significant cost sharing
    approaches does not seem appropriate.

   A key element of affordability is also the cost of healthcare goods and services. There are
    many approaches to reducing the cost of supply, including government intervention in pricing
    of goods and services, the use of tiered pricing approaches for low income environments, and
    negotiation of volume discounts by low income schemes. One approach, which has been used



                                                                                                  20
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    in some countries and has great relevance in South Africa, is the use of public facilities as a
    key provider of services to members of low income schemes. This can be facilitated through
    direct contracting, and would be enhanced through the use of significant discounts to such
    schemes. In addition to reducing costs significantly, this has the obvious advantage of
    strengthening public sector capacity by increasing revenue flow to public hospitals. In the SA
    context, there are several constraints against using this approach on a large scale in the short
    term. These include the poor quality, both perceived and real, at many public hospitals, as
    well as the fact that, in practice, many potential members of low income schemes receive
    public hospital services at no charge, since user fees are not effectively implemented.
    However, these constraints are not necessarily permanent, and it is quite possible that public
    hospital quality could be improved. In these circumstances, contracts between low income
    schemes and public hospitals could be attractive for both parties, and the resulting low costs
    would allow affordable premiums.

   The review revealed a number of supply side, demand side and regulatory barriers, which
    have prevented PHI from serving the needs of those beyond the highest income groups.
    These barriers will need to be overcome through careful intervention and regulation if South
    Africa is to be successful in expanding a PHI market that provides equitable access to health
    care for low income households.

2.2 RESULTS OF LIMS HOUSEHOLD                           SURVEY AND             ANALYSIS         OF
    EXISTING STATISTICAL DATASETS

The LIMS process commissioned two key pieces of household level research: the first was an
analysis of all existing relevant information from official Stats SA and other major databases and
surveys (the base line analysis); and the second was a LIMS specific national household survey
(the HH survey). Both of these exercises were aimed at providing LIMS stakeholders with a
sound evidence base on the following fundamental issues:

   A detailed understanding of the demographics of the targeted low income population,
    including details on household structure, income distribution and employment patterns.

   Patterns of medical scheme coverage and non coverage amongst low income households.

   Patterns of healthcare utilization and expenditure amongst low income households.

   Preferences, willingness to pay and affordability of medical scheme coverage amongst those
    with and without current medical scheme cover.




                                                                                                 21
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



The results of the base line analysis were summarized in the First Interim Report. This section
provides a summary of the key findings of the HH survey. For the most part, the findings of the
base line analysis and the HH survey are well correlated. Where discrepancies between the HH
survey and the baseline analysis were identified, there are highlighted and explained where
possible. The full report of the base line analysis is provided in Annexure 4, while the report of
the HH survey is provided in Annexure 5.

2.2.1         Size and demographic characteristics the target market for LIMS schemes

   The HH survey estimates some 5.1 million non-rural households with gross household
    income of R6,000 or less per month. This total comprises 2.5 million households with
    monthly income of R2,500 or less, and 2.6 million households with monthly income of
    between R2,500 and R6,000 (vs. GHS estimate of 2.3 million and AMPS estimate of 2.9
    million).
    Distribution of the Target Population by Monthly Household Income
        Monthly Gross Househol d Income                N             %              SE          Es t. size
        < R2,501                                   1,806           47%           0.011         2,393,848
        R2,501 – R3,500                              780           22%           0.009         1,105,823
        R3,501 – R4,500                              487           16%           0.008           794,592
        R4,501 – R6,000                              457           14%           0.008           700,907
        Missing                                       62            1%           0.003            71,599


   The size and characteristics of the target market correlate well with previous estimates from
    household surveys with larger sample sizes. Most differences can be explained by different
    years of the surveys, and slightly different means of defining income-related cut-offs for
    inclusion. In this study, gross household income included income from wages, remittances,
    grants and all other income sources.

   There are an estimated 21.9 million individuals residing in all non-rural households with
    incomes below R6,000 per month. Of this total, there are 11.5 million individuals living in
    households earning between R2,500 and R6,000 per month (vs GHS estimate of 8.8 million
    and AMPS estimate of 13.4 million).

   For the sample overall, around 50% of households earning below R6,000 per month had one
    or more household members employed in the formal sector. For households in the target
    population (R2,501-R6,000), 86% had one or more members working in the formal sector.

   Inter-generational households are common in this target group. Of all individuals living in the
    sampled households, some 13% of all individua ls were grandchildren of the household head.
    Only 0.3% was not related to the household head.



                                                                                                       22
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   Remittances to and from individuals living outside the household are a fairly substantial
    feature of the urban households surveyed here. Around 14% of all households reported
    sending money to someone living elsewhere (e.g. parent or child living elsewhere); sending
    money elsewhere was more common amongst the lowest income households (17% compared
    to 11% of households with income of R2,501-R6,000); in most cases the person remitting
    was the son or daughter of the head of household. Some 16% of households reported
    regularly sending money outside the households e.g. to children or the elderly elsewhere; this
    practice was more common amongst the higher income group (22% of households earning
    R2,501-R6,000 compared to 9% of those with income R2500 or less).

2.2.2       Levels of health insurance cover and barriers to cover

   For the sample overall, some 7.3% of individuals (an estimated total of 1,525,985) were
    members of medical aid schemes. In the target population of households earning R2,500 to
    R6,000 per month, 13% of individuals (1,487,258 individuals) were members of medical
    schemes. This compares to the GHS data cited in the baseline survey which shows that in the
    target market, 16% of individuals (1.38 million) living in 25% of households (55,700) have
    medical scheme cover.

   Conversely, 92.7% of the sample overall had no medical scheme cover (20,055,804
    individuals). In the target market, the equivalent figures are 87% (9,953,188 individuals).
    This compares to the estimate from the GHS of 7.35 million individuals in the target market
    with no cover.

   Of those without cover, the household survey shows that 39.9% (587,778 individuals) live in
    households where one or more members do have medical scheme cover. This compares with
    the GHS data which show that, in the target market, 550,000 individuals without cover in live
    in households where someone is covered. This indicates that extending medical scheme cover
    to the uncovered household members in household where there is already partial cover would
    in its own right make a significant impact on coverage levels.

   Further examination of those households where there is partial or no coverage indicates that
    51.5% of these households have at least one person in formal employment. This suggests that
    extending coverage to uncovered formal sector employees and their dependents would make
    a dramatic impact on coverage patterns.

   Within the highest household income band (R4,500 – R6,000), 23% of individuals were
    members of medical schemes. In the lowest income band (0 - R2,500 per month), only 1%
    were members of schemes.




                                                                                               23
                  Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Proportion of Individuals Covered by Medical Aid by Gross Monthly Household Income
                                                                       No Medical Ai d              Medical Ai d                 Total
    < R2,501                                                                   98.6%                        1%                 100.0%
    R2,501 – R3,500                                                            93.8%                        6%                 100.0%
    R3,501 – R4,500                                                                 88.2%                      12%             100.0%
    R4,501 – R6,000                                                                 76.7%                      23%             100.0%
    Total                                                                           92.7%                     7.3%             100.0%
N=14,280; Population size=21,799,787

       Medical scheme coverage of individuals varies significantly by age group. Coverage was
        found to drop steeply for ages 20-29 years and after age 50 years. This pattern is well
        recognized from the current data on medical scheme coverage, and this confirms that these
        patterns are present in low income households as well.

       Whilst, overall, some 7.3% of all individuals were members of medical schemes, only 3% of
        those aged 20-24 years were scheme members, and 5% of those aged 25-29 years. The
        elderly also had low coverage. In households earning R2,500 to R6,000 per month, the
        highest coverage is seen in the 30-49 year age group. Low levels of coverage are evident in
        the older age groups, with a downward decline evident from age 50 years. Low coverage is
        also evident in the 20-24 year old group, with coverage less than 6% of individuals, compared
        to around 18% for the 35-44 year age bands.

Proportion of Individuals Covered by Medical Aid by Age Band (full sample)


                    10-     15-     20-     25-     30-     35-     40-     45-      50-    55-       60-      65-     70-     75-
0-4       5-9                                                                                                                           >79
                     14      19      24      29      34      39      44      49      54      59        64       69      74      79
years     years                                                                                                                         years
                    years   years   years   years   years   years   years   years   years   years     years    years   years   years
7%        9%        7%      6%      3%      5%      10%     11%     12%     11%      7%     7%        5%       4%      3%      0%       1%




                                                                                                                                       24
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Proportion of Individuals Covered by Medical Schemes By Age Band (Individuals Living in
Households with Gross Monthly Income R2,500-R6,000 only)




     It is notable that 6% of individuals overall, and 13% of individuals in the highest household
      income band had previously been a member of a medical aid scheme, but are no longer. The
      fact that the proportion of those previously covered but now without cover is similar to the
      total proportion with cover underlines the well recognized pattern of high turnover in medical
      scheme membership amongst low income members.

Proportion of Individuals who were previously Members of a Medical Aid/Sick Fund
                                            Yes           No         Don’t          N/A         Total
                                                                     know
    < R2,501                               3.3%        96.3%         0.2%          0.0%        100.0%
    R2,501 – R3,500                        8.0%        90.7%         0.1%          1.2%        100.0%
    R3,501 – R4,500                        7.9%        91.7%         0.2%          0.0%        100.0%
    R4,501 – R6,000                       13.1%        86.2%         0.7%          0.0%        100.0%
    Total                                  6.2%        93.2%         0.3%          0.3%        100.0%
N=13,110; Sub-Population Size=19,805,103

     Some 7% of individuals overall had been given the opportunity to join medical aid schemes,
      but had chosen not to do so. In the highest income group (R4,500-R6,000), this proportion
      was 12.3%, whereas it was only 3.1% in those households earning below R2,500 per month.




                                                                                                   25
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Proportion of Individuals Not Covered by Medical Aid that have been Offered Opportunity
to Join Medical Aid Scheme
                                            Yes           No         Don’t          N/A         Total
                                                                     know
    < R2,501                               3.1%        96.0%         0.6%          0.2%        100.0%
    R2,501 – R3,500                        9.9%        87.6%         2.5%          0.0%        100.0%
    R3,501 – R4,500                       11.7%        85.8%         2.4%          0.1%        100.0%
    R4,501 – R6,000                       12.3%        85.3%         2.5%          0.0%        100.0%
    Total                                  7.0%        91.3%         1.6%          0.1%        100.0%
N=13,110; Sub-Population Size=19,805,103

     Within partially covered households, there appear to be distinctive patterns determining
      choice of which members of the household should obtain cover. A key feature here appears to
      be selection of those members in poor health. Although, in general, the proportion of
      individuals in self reported poor or very poor health was not markedly different for lower
      income households compared to higher income households, there does appear to be a pattern
      of adverse selection in partially covered households. Medical scheme members who live in
      partially covered households appear to be higher service users than expected. This is
      illustrated by per capita utilisation rates for this sub-population, which are higher than for
      those medical scheme members in fully covered households. To the extent that service
      utilization is a proxy for health status, this would confirm this pattern of adverse selection.

     Multivariate analysis of those individuals who are not covered by schemes and are living in
      households where other members are covered, indicates the following in relation to
      households that have partial cover :

      o   Individuals in poor or very poor health are 1.5 times more likely to have medical cover
          than those who are in good health.

      o   Older (above 64 years) and young adults (20-29 years) are significantly less likely to be
          covered than those aged 30-64 years.

     When asked why some members of the household were not members of medical aid schemes,
      lack of employment was the most commonly cited barrier to scheme membership. Lack of
      employment was understandably mentioned more commonly by households with no
      members formally employed (accounting for 75% of reasons given by respondents in these
      households).

     In households with no household member in formal employment, lack of employment
      accounted for 72% of reasons given, followed by “never heard about medical aid” (10.7%)
      and high cost (7.2%).




                                                                                                   26
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   Where the household had one or more members formally employed, the most common
    reasons for not being covered related to the employer not offering medical scheme cover
    (27% of reasons), lack of employment (28.5%) and the high costs associated with scheme
    membership (15%).

   Barriers to scheme membership reported by respondents thus included both lack of
    affordability associated with lack of employment or low income earning, and also a
    perception that scheme membership was only open to people in the formal sector.

   Overall, the key obstacles of lack of employment, high cost, and lack of employer assistance
    accounted for 75% of all the reasons for not belonging to medical schemes. By contrast,
    conceptual barriers (such as indications of lack of understanding of risk-pooling, or
    complexity of medical schemes, or lack of knowledge of schemes at all) were mentioned
    much less frequently by respondents.

Distribution of Reasons Why Some Household Members are Not Members of Medical Aid
Schemes




                                                                                             27
             Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Distribution of Reasons Why Some Members of the Household Are Not Medical Scheme
Members, by Household Employment Status

                                                                     One/ more
                                                     One/ more
                                       No Household                  members
                                                      members
                                         member                       working      Total
                                                    working formal
                                          working                     informal
                                                        sector
                                                                       sector
Not working                                   72.0%         28.5%          46.8%       42.1%
Never heard about it                          10.7%          3.6%          6.3%            5.8%
It is to expensive                             7.2%         17.7%         15.2%        14.9%
Too complicated                                2.4%          6.5%          4.1%            5.1%
Employer doesn't offer it                      0.5%         27.1%         13.2%        18.2%
Other                                          7.2%         16.7%         14.4%        14.1%
Don't need                                     1.7%          5.6%          4.4%            4.5%
Deduct money even if you aren't sick           1.6%          1.2%          1.3%            1.3%
Don't know                                     1.5%          3.1%          2.2%            2.6%
Don't qualify/self employed/casul              1.5%          1.9%          5.2%            2.6%
Main member died/lost job/retired              0.7%          1.8%          0.4%            1.3%
Don't have info on med aid                     0.1%          0.4%          0.2%            0.3%
Never thought about it                         0.1%          0.6%          0.3%            0.4%
Refused to answer                              0.0%          0.4%          0.1%            0.3%
Don't trust med aid                            0.0%          0.0%          0.1%            0.0%
Required id/don't have bank account            0.0%          0.0%          0.1%            0.0%
Divorced/too old                               0.0%          1.0%          0.0%            0.5%
Not family member                              0.0%          0.2%          0.0%            0.1%
In process of getting med aid                  0.0%          0.4%          0.0%            0.2%
Total                                        100.0%        100.0%        100.0%      100.0%



   When asked who in their households they would like to see covered by medical aid,
    respondents chose hypothetically to cover approximately 80% of the total population across
    their households. The highest household income quintile was most selective about whom
    they would choose to cover. More specifically, these higher income households proposed
    lower rates of coverage for the elderly and young adults (20-29 years), confirming the
    tendency to select those with higher risk for coverage.




                                                                                                  28
               Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Percentage of Household Members that Respondent Would Choose to Have on Medical Aid
by Age Group of Household Members

      90%
      80%
      70%
      60%
      50%
      40%
      30%
      20%
      10%
        0%
                0-19 years   20-29 years   30-64 years   65 years +   Overall

                       R0-2500   R2501-3500    R3501-4500    R4501-6000




2.2.3          Attitudes to Health Insurance

     Across all income groups, there were reasonably high levels of interest in obtaining medical
      scheme coverage. The data suggest that over 60% of respondents across all household
      income groups would be willing to „risk pool‟ for healthcare risk. It is important to note
      however, that this result is based on the proportion who “agreed” with the relevant statement
      in the survey, whereas only a low proportion of respondents „strongly agreed‟ with these and
      other statements that pertained to their likely participation in health insurance schemes.

Proportion of Respondents willing to Risk Pool by Household Income
                                            Strongly        Agree     Disagree       Don’t         Total
                                               agree                                 know
    < R2,501                                   8.8%         47.1%         30.0%      14.1%        100.0%
    R2,501 – R3,500                           12.6%         55.8%         26.8%       4.7%        100.0%
    R3,501 – R4,500                           13.7%         60.6%         22.2%       3.5%        100.0%
    R4,501 – R6,000                           13.5%         64.0%         18.4%       4.1%        100.0%
    Total                                     11.1%         53.5%         26.4%       8.9%        100.0%
N=3,560; Population Size=50,101,885

     Some 72% and 82% of respondents in the low income group and higher income group
      (R2,500-R6,000) respectively, strongly agreed or agreed with the statement that „I would be
      willing to pay a small amount each month for free medical care even if I am not sick now.‟



                                                                                                      29
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Proportion of Households Prepared to Pay a ‘Small Amount’ Each Month for Medical Aid
by Household Income
                                       Strongly        Agree      Disagree        Don’t         Total
                                          agree                                   know
    < R2,501                             11.5%         55.0%         22.8%        10.8%        100.0%
    R2,501 – R3,500                      15.4%         59.9%         20.8%         3.9%        100.0%
    R3,501 – R4,500                      17.3%         63.5%        16.9%          2.3%        100.0%
    R4,501 – R6,000                      18.6%         65.5%        13.9%          2.0%        100.0%
    Total                                14.3%         58.8%        20.2%          6.6%        100.0%
N=3,557; Population Size=50,009,829

Proportion of Respondents who Agree That They Do not Need to be Members of Medical
Aid by Household Income
                                       Strongly        Agree      Disagree        Don’t         Total
                                          agree                                   know
    < R2,501                              7.4%         15.2%         48.8%        28.7%        100.0%
    R2,501 – R3,500                       4.3%         12.1%         52.6%        30.9%        100.0%
    R3,501 – R4,500                       3.1%         13.0%        51.7%         32.2%        100.0%
    R4,501 – R6,000                       3.9%          9.6%        52.3%         34.2%        100.0%
    Total                                 5.5%         13.3%        50.5%         30.7%        100.0%
N=3,563; Population Size=50,140,627

     In relation to private hospital cover, approximately 72% of the sample agreed or strongly
      agreed with the concept of risk pooling to fund private hospital care so as to avoid out of
      pocket payments for these services.

2.2.4     Preferences and Willingness to Pay for various coverage options

     The HH survey attempted to identify preferences of respondents to different components of a
      medical scheme package, as well as their willingness to pay for these components in several
      different ways. Firstly, respondents were asked about what components of a package would
      be essential if they were to purchase medical scheme cover. Secondly, respondents were
      asked to indicate the Rand amount they would be willing to pay per member per month for
      selected components of a medical scheme package. Further, more comprehensive data on
      these issues was obtained through a Discrete Choice Experiment (see below).

     In relation to questions as to what components of a package would be essential for the
      respondent to consider purchasing medical scheme cover, the following results were
      identified:

      o   20%-25% indicated that they would not purchase cover without General Practitioner
          (GP) visits, with a higher proportion in the higher income group.



                                                                                                   30
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    o   Approximately 15% of the sample indicated that they would not purchase medical
        scheme cover unless it included private hospital cover.

    o   Around 14% indicated that comprehensive cover; including both GP and hospital cover
        was an essential requirement of medical scheme cover.

   Respondents were also asked what benefits they would like to have, depending on cost. Here,
    responses were as follows:

    o   55% of the population indicated that they would like to purchase GP cover, depending on
        cost. This ranged from 57% in the lowest income group to 41% in the highest income
        group.

    o   61% of the population indicated that they would like to purchase private hospital cover,
        depending on cost. This ranged from 62% in the lowest income group to 51% in the
        highest income group.

    o   64% of the population indicated that they would like to purchase comprehensive cover,
        including GP and private hospital cover, depending on cost. This ranged from 63% in the
        lowest income group to 58% in the highest income group.

   In response to questions regarding various set amounts that respondents might be willing to
    pay for various packages, the key conclusions were as follows:

    o   80% of the total sample indicated that they would be willing to pay R25 per month for
        GP visits. As the price increased, the proportion willing to pay the amount decreased,
        with 41% of the sample indicating that they would be willing to pay R45 per month for
        GP cover. These responses vary by income group. In the highest income group, for
        example, 89% indicated that they would be willing to pay R25 per month, while 59%
        indicated that they would be willing to pay R45 per month.

    o   In the case of a comprehensive package, including both GP and hospital cover, 10% of
        the sample indicated that they would be willing to pay R100 per month. As the price
        increased to R200 per month, the proportion willing to pay this decreased to 4%. As
        would be expected, in the highest income group, the willingness to pay is higher, with
        23% willing to pay R100 per month, and 11% willing to pay R200 per month.

    o   The survey also tested willingness to pay for improved care at public hospitals. Here the
        data indicate that 18% of the overall sample would be willing to pay R40 per month for
        this benefit, while 10% would be willing to pay R80 per month. For the highest income
        group, 30% would be willing to pay R40 per month, and 20% would be willing to pay
        R80 per month.



                                                                                              31
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Willingness to Pay for Medical Insurance for General Practitioner Visits by Gross Monthly
Household Income

  100%
                       88%
  90%                    89%
                86%
  80%                    80%
                         73%                           69%
  70%
                                                       66%
  60%                                                  58%                       59%
  50%                                                  51%                       52%
                                                                                 46%
  40%                                                                            41%
                                                       38%
  30%                                                                            29%
  20%
   10%
   0%
                  R 25                              R 35                     R 45


                      Total Sample                 <2501                   2501-3500
                      3501-4500                    4501-6000



                      Frequency            Overall            <R2501       R2501-3500 R3501-4500 R4501-6000
           R 25                1036                80%               73%            86%      89%       88%
           R 35                    365             51%               38%            58%      66%       69%
           R 45                1455                41%               29%            46%      52%       59%




Willingness to Pay for Medical Insurance for Private Hospital Cover by Gross Monthly
Household Income
  100%
  90%                  86%
                         89%
                88%
  80%                        79%
  70%                    71%                               68%
                                                           69%
  60%                                                      60%                      57%
                                                            51%                     53%
  50%                                                                               48%
  40%                                                                                 40%
                                                           37%
  30%                                                                               28%
  20%

   10%
   0%
                      R 20                           R 30                      R 40


                       Total Sample                <2501                   2501-3500
                       3501-4500                   4501-6000


                Frequency                Overall            <R2501      R2501-3500 R3501-4500 R4501-6000
         R 20                1023             79%                 71%          88%          89%      86%
         R 30                 371             51%                 37%          60%          69%      68%
         R 40                1427             40%                 28%          48%          53%      57%




                                                                                                              32
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Willingness to Pay for Medical Insurance for Comprehensive Cover (Private Hospital and
General Practitioner) by Gross Monthly Household Income

  25%                   23%
                                                 23%


  20%


  15%                                           13%
                13%       12%                    12%
  10%                     10%                    10%                  11%



  5%                                                                  5%
                          4%                     4%
                                                                      4%
                                                                      2%
  0%
                      R 100                 R 150                 R 200


                        Total Sample       <2501                2501-3500
                        3501-4500          4501-6000


                                      Total
                 Frequency                          <R2501    R2501-3500 R3501-4500 R4501-6000
                                     Sample
        R 100                   0         10%            4%         13%        12%        23%
        R 150                  179       10%             4%         13%        12%        23%
        R 200                  159        4%             2%          5%         5%        11%




                                                                                                 33
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



2.2.5    Results from Discrete Choice Experiment

   As discussed in detail in the Report of the HH survey, a discrete choice experiment (DCE)
    was carried out, in which respondents were asked to choose between various options of
    different medical scheme packages or having no medical scheme at all. Table 2.2.5.1 below
    summarises the main results from the discrete choice experiment (DCE).

   The objective of this part of the study was to determine which components of different
    medical aid packages were most important to respondents from low-income households. The
    results are expressed in economic terms (Willingness to Pay for different package
    components) rather than the more familiar rankings produced by other methods of
    comparative valuation. It is argued that economic approaches to valuation are more realistic
    and consistent than ranking methods but Willingness to Pay (WTP) should still be interpreted
    as a measure of the relative importance of different choices rather than a direct assessment of
    what households are actually prepared to pay or, indeed, can afford to pay. It should be
    emphasised that these results reflect „stated preferences‟ rather than „revealed preferences‟. It
    is not possible to analyse „revealed preferences‟ for LIMS because no real market for LIMS
    exists at present so we have to rely on respondent‟s hypothetical choices of such packages if
    they were offered by the industry.

   In the DCE, respondents were asked to choose between a particular medical aid package at a
    specific price or having no medical aid. The analysis utilises regression techniques to model
    these choices as a function of the package components. The results of these regression
    analyses are summarised in Table 2.2.5.1. The regression coefficients represent the relative
    importance of each component in the respondents‟ choices. There is obviously a confidence
    interval around all estimates. Non-significant (p>0.05) coefficients are shown in grey
    indicating which components were not statistically important in the respondent‟s choice of
    medical aid package (respondents would not pay for that component and would use the public
    sector). The relative rankings can also be expressed as the Willingness to Pay (per member
    per month) for each component if the component coefficient is divided by the cost
    coefficient.

   Table 2.2.5.1 compares the valuations of the total study sample (top two lines) with different
    sub-segments of the sample. Differences in WTP for a specific component between different
    sub-populations could be due to differences in the valuation of that component, differences in
    the importance of cost in respondent‟s choices (indicated by the coefficient on the cost
    variable), or both.

   The basic results for the individual package components are shown on the left side of the
    table. One of the advantages of the DCE methods is that part worths can be added or
    subtracted to compare the utility of different package combinations. Some likely medical aid



                                                                                                  34
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    package combinations are show on the right side of the table. It should also be noted that due
    to the study design and choice of coding schemes for the regression, some of the package
    components on the left side of the table are already composite measures. For example, „GP +
    Specialist Care‟ incorporates „GP care‟ and the valuation of the addition of specialist care to
    basic GP care is measured by the difference between the two coefficients (or WTP estimates).

Overall Results

   The results for the total study sample are shown in the top lines of Table 2.2.5.1. All
    coefficients were statistically significant which means that all the package components
    included in the study were important to respondents‟ choices. As expected the coefficient on
    cost is negative indicating that respondents would prefer cheaper packages. The coefficients
    for all other variables were positive reflecting that if cost were not a factor respondents would
    prefer packages with more cover.

   The relative importance of different components is evaluated by comparing coefficients or
    WTP. For the total sample, emergency hospital cover and basic GP services were valued
    most highly - the WTP was estimated at R36.10 for emergency hospital cover and R31.56 for
    GP cover. Interestingly, optometry cover was also quite highly valued with a WTP of
    R26.65. Although still statistically significant in influencing respondents‟ choices, the
    inclusion of hospital maternal and neonatal care was least important to respondents (WTP of
    R10.90). The value of adding specialist care to GP cover can be calculated as R15.36
    (R46.92-R31.56). Similarly adding in-patient hospital cover to emergency and maternity
    hospital cover is valued at R18.56 (R61.64-R43.08). These last two components were in the
    same order as dental cover which had a WTP of R16.39.

   The definition of an appropriate LIMS benefit package will require the combination of
    individual package components. The analysis indicates that primary care packages would be
    valued more highly than hospital packages. The WTP for basic primary care cover (including
    GPs, optometry and dental care) is R74.61 which is higher than the WTP for full hospital
    cover (R61.64). Adding specialist cover to the PHC package would be valued at R89.97.
    Overall, the combined WTP for a full private care package was R151.61 per member per
    month.

Segmentation Analysis

   The rest of Table 2.2.5.1 shows the results of the regression analysis repeated for different
    sub-groups.

   The overall patterns discussed above were consistent across different sub-groups. However,
    in some cases, the coefficient for maternal and neonatal care was no longer statistically
    significant. Package cost was more important to poorer households resulting in much lower


                                                                                                  35
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    utilities and WTP (respondents were more likely to choose the no medical aid option). Not
    surprisingly, households with higher income, some formal employment or some existing
    cover expressed higher WTP for package components. For all groups, comprehensive out-of-
    hospital benefits were valued more highly than private hospital cover.

   Households in the R2500-R6000 income range, with some formal employment, and no or
    incomplete household cover probably represent the primary market for LIMS. This sub-group
    expresses more utility from medical aid cover than the total sample. It values comprehensive
    PHC at R105.49, private hospital care at R72.74, and full private cover at R178.23 per
    member per month.




                                                                                             36
                                 Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report




       Table 2.2.5.1: Results from Di screte Choice Experiment Analysing Relative Importance of Different Package Components
                                                                          PACKAGE COMPONENTS                                              COMBINED PACKAGES
                                                                                             Maternity
                                                                                                                 Emergency
                                                             GP +                               +      Emergency                Full
                                                GP                      Dental     Optometry                     + Maternity               PHC       PHC      Full
                                      Price                Specialist                        Neonatal Hospital                 Hospital
                                                Care                     Care       Services                      Hospital                (-Spec)   (+Spec) Package
                                                             Care                             Hospital Care Only                Care
                                                                                                                    Care
                                                                                             Care Only
Total                                 -0.0096    0.3040       0.4519     0.1579       0.2567    0.1050     0.3477     0.4150     0.5937
                             Coeff
(n=3522)
                             WTP                  31.56        46.92      16.39        26.65     10.90      36.10      43.08      61.64     74.61     89.97   151.61

Male Respondent              Coeff    -0.0095    0.2600       0.4139     0.2061       0.2202    0.0556     0.3340     0.3924     0.5665
(n=1264)                     WTP                  27.29        43.45      21.63        23.11       5.84     35.05      41.19      59.47     72.03     88.19   147.66
Female Respondent            Coeff    -0.0097    0.3305       0.4763     0.1295       0.2775    0.1342     0.3590     0.4285     0.6110
(n=2257)                     WTP                  34.07        49.10      13.35        28.61     13.83      37.01      44.17      62.99     76.04     91.07   154.06

HH Income < R2500            Coeff    -0.0121    0.2063       0.3524     0.1647       0.2652    0.1242     0.2374     0.3895     0.5175
(n=1770)                     WTP                  17.06        29.14      13.62        21.93     10.27      19.63      32.20      42.79     52.60     64.68   107.47
HH Income R2501-R6000        Coeff    -0.0084    0.3676       0.5230     0.1737       0.2618    0.1131     0.4153     0.4463     0.6566
(n=1694)                     WTP                  43.99        62.60      20.79        31.33     13.53      49.70      53.42      78.59     96.11    114.72   193.30

HH None employed             Coeff    -0.0110    0.0506       0.2402     0.1961       0.3185    0.0795    -0.0290     0.2880     0.5028
(n=683)                      WTP                    4.61       21.89      17.87        29.02       7.24      -2.64     26.24      45.81     51.50     68.78   114.59
HH Informal employed only    Coeff    -0.0123    0.2702       0.4429     0.1177       0.3285    0.3063     0.2355     0.4102     0.4501
(n=767)                      WTP                  22.03        36.10        9.59       26.77     24.96      19.20      33.43      36.68     58.39     72.46   109.15
HH S ome formal employment   Coeff    -0.0089    0.3568       0.5045     0.1777       0.2297    0.1367     0.4523     0.4481     0.6640
(n=2069)                     WTP                  40.06        56.65      19.96        25.79     15.35      50.79      50.32      74.57     85.81    102.40   176.97

HH No MA                     Coeff    -0.0101    0.2677       0.4233     0.1534       0.2626    0.1076     0.2863     0.3821     0.5364
(n=3144)                     WTP                  26.55        41.98      15.21        26.04     10.67      28.40      37.90      53.20     67.81     83.24   136.44
HH Partial MA cover          Coeff    -0.0085    0.5467       0.7488     0.1240       0.2567    0.2188     0.6723     0.8688     0.9960
(n=179)                      WTP                  64.39        88.19      14.61        30.23     25.77      79.19     102.32     117.31    109.23    133.03   250.34
HH Full MA cover             Coeff    -0.0066    0.4913       0.5561     0.3867       0.2255    0.0618     0.6045     0.6278     1.0376
 (n=188)                     WTP                  74.57        84.40      58.70        34.23       9.38     91.76      95.29     157.49    167.50    177.33   334.83

HH R2501-R6000, S ome formal          -0.0087    0.3673       0.5157     0.1562       0.2409    0.1520     0.4406     0.4557     0.6295
employed, No or partial MA   Coeff
(n=1300)                     WTP                  42.45        59.60      18.05        27.84     17.57      50.92      52.67      72.74     88.34    105.49   178.23




                                                                                                                                                        37
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



2.2.6   Current patterns of utilisation of health care services and attitudes to these services

Levels of satisfaction with healthcare services

   In order to assess levels of satisfaction with current healthcare services, respondents were
    asked to indicate where they had obtained healthcare as out-patients in the last month (and as
    in-patients during the past year), and to indicate their degree of satisfaction with the services
    used.

   Responses to these questions indicated that 16% of medical scheme members and 3% of non
    medical scheme members had visited a private GP in the past month, and that 95% of these
    were either satisfied or very satisfied with the services obtained.

   In relation to public sector clinics, 3% of those with medical scheme cover and 11% of those
    without cover had visited a public clinic in the past month. Here satisfaction levels were
    slightly lower, with 72% indicating that they were either satisfied or very satisfied with the
    services obtained.

   Visits to private specialists were very rare amongst both medical scheme members and non
    members, although satisfaction was rated very highly by the small number of individuals who
    had visited a private specialist in the past month.

   4% of medical scheme members had used a private hospital during the past year, while only
    17 people (less than 1%) without cover had used a private hospital. Satisfaction levels were
    high here as well, with 95% indicating that they were either satisfied or very satisfied.

   1% of those with medical scheme cover and 4% of those without had used a public hospital
    during the past year. Satisfaction levels were lower than for private hospitals, but still
    reasonable, with 68% indicating that they were either satisfied or very satisfied.

   Respondents were asked why they had used that service, and not the alternative. Users of
    public clinics said that they did not use private GPs primarily because of cost. However,
    there was a commonplace perception expressed that the public sector was good enough for
    routine care, and minor illnesses, whilst the private sector was necessary for more serious
    health problems.

Use and perceptions of Public Sector Health Care

   Choice of service provider, as well as attitudes to medical scheme coverage seemed to be
    highly influenced, not only by cost, but also by the respondents‟ perceptions and experiences
    of public sector health care available to them. Some of the reasons cited by respondents for
    their choices are shown in the following tables.


                                                                                                  38
           Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Distribution of Reasons Given as to Why Private Facilities were used and not Public,
amongst users of Private Sector Out Patient Facilities

                                                No Medical Ai d        Medical Ai d           Total
Medical Aid pays                                            10%                53%             22%
Told to go there by doctor/nurse                             4%                 1%              3%
Told to by med ical aid                                      0%                 8%              2%
Emergency                                                    7%                 1%              6%
No public facility close by                                  3%                 0%              2%
Good quality                                                68%                36%             59%
Sent by Emp loyer                                            1%                 0%              1%
For Special t reat ment                                      1%                 0%              0%
Working – no time to go to public facilities                 0%                 0%              0%
Public facilities always full                                2%                 0%              1%
Public facilities don‟t offer quality care                   1%                 0%              1%
Don‟t know                                                   1%                 0%              1%


Distribution of Reasons given as to why Public facilities were used, and not Private,
amongst Users of Public Sector Out Patient Facilities
                                                No Medical Ai d        Medical Ai d           Total
Don‟t have to pay                                           93%                56%             92%
Referred by doc                                              2%                 9%              2%
Referred by medical aid                                      0%                11%              0%
Emergency                                                    2%                10%              2%
No private facility close by                                 1%                 9%              1%
Good quality                                                 1%                 2%              1%
Don‟t know                                                   0%                 4%              0%
Cheaper                                                      1%                 0%              1%
Minor problem                                                0%                 0%              0%


Distribution of Reasons why Public Hospitals were used for most recent Hospitalisation by
Medical Aid Status of the Patient
                                                    Medical Ai d             Total
Don‟t have to pay                                         35.2%               76%
Referred by doc                                           20.2%             6.20%
Referred by medical aid                                   17.0%                0%
Emergency                                                  0.0%               14%
No private hospital close by                              14.7%                1%
Good quality                                               0.0%                1%
Don‟t know                                                 0.0%                0%
Cheaper                                                   12.9%                2%
No difference                                              0.0%                1%
Total                                                   100.0%              100%



                                                                                                39
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Distribution of Reasons why Private Hospitals were used for most recent Hospitalisation
(medical aid members only)

                                                         Total
Medical aid pays                                         75.7%
Referred by doctor                                        9.4%
Referred by medical aid                                   2.7%
Emergency                                                 3.9%
Good quality                                              8.5%
Other                                                     0.0%
Total                                                  100.0%



Frequency of Out-Patient Health Care Visits in Preceding 3 Months

In order to assess current utilization patterns in more depth, respondents were asked to indicate
which services they had used as an out-patient during the past 3 months, and as an in-patient
during the last year. The responses to these questions can be summarized as follows:

Non medical scheme members

   Some 7% of individuals who were not on medical aid reported using a private GP one or
    more times in the preceding 3 months; there were on average 0.12 visits per individual to
    private GPs in the preceding 3 months. This equates to an annualized visit rate of below 0.5
    per annum, which is much lower than that observed amongst current medical scheme
    members, which ranges from 2-5 visits per annum.

   This pattern of usage of private GPs by non medical scheme members was also confirmed by
    analysis at the household level. For households with no medical scheme cover, some 12%
    reported that their last visit to a healthcare provider in the last month was to a private GP,
    whereas 28% of these households reported that their last visit was to a public sector clinic.
    For household with partial cover, 31% reported that the last visit was to a public sector clinic,
    whereas 39% reported that the last visit was to a private GP. Where the household had full
    cover, only 8% of last visits by a household member were to a public sector clinic, with 31%
    to a private GP. In addition, approximately 5% of households with no medical scheme cover
    are purchasing chronic medication through the private sector.

   Some 15% of those not on medical aids had used a public sector government clinic within
    the preceding 3 months; the average number of visits was 0.32 visits over three months per
    individual. This equates to an annualized visit rate of 1.2 per annum.




                                                                                                  40
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   Less than 0.5% of individuals who were not on medical aid reported using a
    workplace/factory clinic in the preceding 3 months, with average number of visits at 0.007.

   Utilisation of specialists was similarly low for those not on medical aid, with less than 0.5%
    using specialists and average number of specialist visits per individual over three months
    estimated to be 0.05 visits (95% CI 0.003;0.007)

   Around 1% of individuals who were not on medical aid reported that they had used
    traditional healers in the preceding 3 months; however reports from the interviewers
    suggested that these data may be unreliable since respondents might not have wished to
    reveal their usage of traditional healers to the interviewers. As a result, actual utilization of
    traditional healers may well be higher than reported here.

   The table below indicates the distribution of out patient visits between different service
    providers for non medical scheme members only, by income group. As the table shows, non
    medical scheme members accessing out patient care overwhelmingly utilize public sector
    clinics (76% of all visits), and this pattern is correlated with income (lowest income category
    had 82% of visits in public sector clinics vs. 63% in highest income category).

   The table also shows that 22.4% of all out patient visits by non medical scheme members
    were to private GPs, and that this pattern is also correlated well with income. In this case,
    16.4% of all out patient visits by the lowest income category were to private GPs, whereas
    the equivalent figure for the highest income category was 34.3%.

   The table also indicates that non medical scheme members make very limited use of other
    outpatient services, including private specialists, pharmacists, dentists etc.

Distribution of Most Recent Out Patient Visits, by Service Provider - Individuals without
Medical Aid Cover Only
                        Private
             Public               Clinic at     Private     Traditional
                        Doctor                                            Dentist   Pharmacist    Total
              Clinic                 work     S pecialist       Healer
                          (GP)
<R2,501       82.2%      16.4%       0.1%          0.1%           0.1%      0.2%         0.4%    100.0%
R2,501-
              72.8%      26.0%       0.0%          0.0%           0.0%      0.0%         1.2%    100.0%
R3,500
R3,501-
              68.3%      28.6%       0.0%          0.0%           0.8%      0.6%         0.8%    100.0%
R4,500
R4,501-
              62.7%      34.3%       1.0%          0.0%           0.0%      1.0%         1.0%    100.0%
R6,000
Total        75.8%      22.4%        0.2%         0.1%           0.2%      0.3%          0.7%    100.0%
N=1,827; Population Size= 2,846,802



                                                                                                     41
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Medical Scheme Members

   Utilisation amongst medical scheme members was much higher than for non scheme
    members. Overall, 34% of individuals who were medical aid members had used a private GP
    one or more times during the preceding 3 months. Individual medical aid members had an
    average of 0.56 visits to GP services over three months, equivalent to an annualized visit rate
    of 2.24.

   Some 6% of medical aid members had used a public sector government clinic within the
    preceding 3 months; the average number of visits to public sector clinics over 3 months for
    all medical aid members was 0.12, equivalent to an annualized visit rate of 0.48.

   Around 2% of medical aid members had used workplace/factory clinics in the preceding 3
    months.

   Some 5% of medical aid members reported consulting a private specialist during the
    preceding 3 months, with 3 month utilization rates at 0.07 per individual (95% CI 0.04;0.10),
    equivalent to an annualized visit rate of 0.28.

   The table below shows data on the distribution of the most recent out-patient visit between
    different providers for medical scheme members only. Here, as would be expected, the vast
    majority of most recent out patient visits are to private GPs (79.5%), with 12.8% to public
    sector clinics and 3% to pharmacists.

   The relatively high usage of public sector clinics by medical scheme members is also
    noteworthy. This does not show any particular correlation with income level.

   This pattern of usage of public sector clinics by those with medical scheme cover is
    confirmed by analysis at the household level. Here, approximately 8% of households with full
    medical scheme coverage reported that their most recent visit to a healthcare provider during
    the past month was to a public sector clinic. It is very likely that this pattern is related to these
    members obtaining chronic medications not covered by their medical scheme from public
    sector clinics. In fact, the data show that 9% of households in the income range of R2,500 to
    R6,000 per month with full medical scheme coverage have one or more members obtaining
    chronic medication from the public sector. This compares with 24% of households where
    there is partial scheme coverage, and 25% of households with no scheme coverage.




                                                                                                      42
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Distribution of Most Recent Out Patient Visits, by Service Provider – Medical Aid Members
Only
                       Private
             Public              Clinic at     Private     Traditional
                        Doctor                                           Dentist   Pharmacist    Total
             Clinic                 work     S pecialist       Healer
                          (GP)
<R2,501       6.0%       92.6%      0.0%          0.0%           0.0%      0.0%         1.4%    100.0%
R2,501-
             10.6%       65.6%     10.9%          0.0%           6.2%      0.0%         6.8%    100.0%
R3,500
R3,501-
             19.3%       79.1%      0.0%          0.8%           0.0%      0.8%         0.0%    100.0%
R4,500
R4,501-
             11.5%       82.8%      0.0%          1.6%           0.0%      0.0%         4.2%    100.0%
R6,000
Total        12.8%      79.5%       1.9%         1.0%           1.1%      0.2%          3.5%    100.0%

N=194; Population Size= 311466

Frequency of hospital admissions during the preceding year

Respondents were asked to report on admissions to hospitals amongst household members during
the past year. A hospital admission was defined as including at least one overnight stay. Fairly
low numbers of overnight admissions were recorded for the sample overall.

Non medical scheme members

   For those individuals without medical scheme cover, 4.5% of the sample reported a hospital
    admission in the preceding year;

   Analysis of admissions to private and public hospitals indicate that for those not on medical
    schemes, 98.5% of all hospital admissions were to the public sector, as would be expected.
    This pattern is positively correlated with income, with 99.3% of admissions in the lowest
    income category being to public hospitals, whilst in the highest income category, this
    proportion decreased to 92.3%.

   As would be expected, cost is the major reason for choice of public sector hospitals by those
    without medical scheme cover. When asked reasons for their choice of a public hospital, the
    fact that public hospital does not require payment accounted for 70% of reasons given. The
    next major reason for this choice was emergency, which accounted for 14% of reasons.

Medical scheme members

   For those individuals with medical scheme cover, 6.1% of the sample reported a hospital
    admission in the preceding year;



                                                                                                    43
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   In the case of medical scheme members, admission rate to hospital is positively correlated
    with income. Here, those in the lowest income category had an admission rate of 0%,
    whereas those in the highest income category had an admission rate of 7%.

   Analysis of admissions to private and public hospitals indicates that for those with medical
    scheme cover, 82.5% of admissions were to private hospitals.

   It is also interesting to note the relatively high proportion, 17.5%, of medical scheme
    members where the most recent admission was to a public hospital. Several reasons were
    given by respondents with medical scheme cover for their choice of a public hospital. The
    main reason was still the fact that public hospital cover did not require payment. This
    accounted for 35% of reasons. This is presumably due to those whose medical schemes only
    cover public hospital cover or perhaps require co-payments for private hospital care. The next
    major reason was “referred by doctor” (20% of reasons), followed by “referred by medical
    aid” (17% of reasons). It can probably be assumed that all of these categories apply to
    medical scheme members with limited or no benefits for private hospital cover.

Type of Hospital/Clinic Used for Most Recent Hospital Stay (No Medical Aid)
                                                 Workplace
                  Government        Private                     Don't Know       Total
                                                  Hospital
<R2501                    99.3%           0.4%             0%              0%      100.0%
R2501-R3500               98.9%           1.1%            0%               0%      100.0%
R3501-R4500               98.3%           1.7%            0%               0%      100.0%
R4501-R6000               92.3%           7.2%            0%               0%      100.0%
Total                      98.5%          1.3%          0.0%              0.2%     100.0%

    N=580; Population Size=90,471,432

Type of Hospital/Clinic Used for Most Recent Hospital Stay (With Medical Aid)
                                                 Workplace
                  Government        Private                       Total
                                                  Hospital
<R2501
R2501-R3500                 9.2%         90.8%          0.0%         100.0%
R3501-R4500               17.4%          82.6%          0.0%         100.0%
R4501-R6000               20.3%          79.7%          0.0%         100.0%
Total                      17.5%         82.5%          0.0%         100.0%


    N=580; Population Size=90,471,432




                                                                                               44
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Distribution of Reasons Given for Public Hospital Use at Last Hospitalisation (<1 year) for
Medical Aid and Non-Medical Aid Members




    N=501; Population Size=79,922,714

   The survey asked respondents to indicate their level of satisfaction with the hospitals used by
    household members during the past year. On the whole, satisfaction levels with services were
    relatively high. In the case of public hospitals, 69% of respondents were moderately or very
    satisfied, with approximately 29% indicating that they were slightly or very dissatisfied with
    their experience in a public sector hospital. In the case of private hospitals, the experience
    appears to have been much better, with 97% of respondents indicating that they were
    moderately or very satisfied and only 0.3% indicating some level of dissatisfaction.

Use of health facilities for antenatal care and child birth

   31.5% of all households with female members contained at least one woman that had given
    birth in the preceding 5 years. Most deliveries (91%) had occurred in government hospitals,
    with the bulk of the remainder in private facilities.

   Amongst households with all members of the household currently covered by medical aid,
    25% of births took place in public or government hospitals or clinics, and the most commonly
    cited reason for this choice of facility was emergency. Data disaggregated by medical aid
    status must be interpreted with caution as we have no way of knowing if the mother was a
    medical aid member at the time of the delivery.

   As would be expected, usage of public or private facilities for childbirth correlates strongly
    with both income and with medical aid coverage status. For those households with no
    medical scheme coverage, 97% of the most recent childbirths in the household occurred in
    public hospitals or clinics, with 0.9% in private hospitals/clinics and 0.7% at home.

   For households with partial medical scheme coverage, 63.6% of births were in public
    hospitals/clinics, and 35.8% in private hospitals/clinics.




                                                                                                45
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



      Where the household was fully covered, 66% of the most recent childbirths occurred in a
       private hospital. Interestingly, 25% of the most recent births in fully covered households
       occurred in public hospitals and clinics. The most important reason for childbirth in a public
       hospital, cited by household with medical scheme coverage, was “emergency” (37% of
       reasons) followed by not having to pay (20%). Another key reason may well be that the
       medical scheme packages of those using public facilities may not provide cover for private
       hospitals.

      Similar patterns were observed for ante-natal care. Here, for households with no medical
       scheme coverage, antenatal care was obtained in the public sector in 91.4% of cases, vs. 2.4%
       in the private sector.

      For households with partial coverage, antenatal care was obtained in the public sector in 67%
       of cases, and in the private sector in 32% of cases.

      For households with full medical scheme coverage, antenatal care was obtained in the private
       sector in 53% of cases, with 38.2% still occurring in the public sector. Data disaggregated by
       medical aid status must be interpreted with caution as we have no way of knowing if the
       mother was a medical aid member at the time of the delivery. This is because we have
       information on births in the preceding 5 years, but we only have information on medical aid
       status at the time of the survey – it could be that these households were not covered by
       medical aid at the time of the health care use, but have become members subsequently.

Proportion of Most Recent Births by Health Facility Used and Household Medical Aid

                     Public          Private Hospital/
                  Hospital/Clinic         Clinic         Home          Don't Know     Elsewhere     Total
None                         97.3%                0.9%          0.7%           0.9%          0.2%      100.0%
Partial                      63.6%              35.8%           0.6%           0.0%          0.0%      100.0%
Full                         24.8%              66.0%           0.0%           5.0%          4.3%      100.0%
Total                        90.9%                7.0%          0.6%           1.0%          0.4%      100.0%

N=1,119; Population Size=15,502,117




                                                                                                            46
             Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Reasons for Giving Birth in a Public Hospital by Household Medical Aid




Proportion of Households Obtaining Antenatal Care at Public and Private Facilities by
Household Medical Aid
                      Public             Private         Private
                   hospital/clinic    doctor/midwife    specialist     No antenatal care   Don't know     Total
None                          91.4%              2.4%           0.0%               4.5%            1.7%      100.0%
Partial                       66.9%             32.1%           0.0%               0.0%            1.1%      100.0%
Full                          38.2%             53.0%           0.8%               8.0%            0.0%      100.0%
Total                         86.4%              7.6%           0.1%               4.4%            1.5%      100.0%


N=1105; Population Size=1526955.2

2.2.7      Health Care Expenditure

Out of Pocket Payment for Out Patient Health Care – Individual level data

      Information from recent health care visits (within the preceding month) was used to estimate
       how much was spent on individuals‟ health care, for those covered and not covered by
       medical aid. These costs are presented separately for different items (e.g. fees, medicines,
       etc).

      There were 2026 individuals (14% of the sample) who had attended some out-patient facility
       in the month preceding the survey.

      Looking at all out-patient visits combined, the greatest contributor to the out of pocket (OOP)
       cost of out patient visits for current service users is the cost of transport, which accounted for
       43% of total OOP costs in the case of medical scheme members, and 46% of OOP costs for
       non medical scheme members respectively.

      After transport, professional fees were the next major OOP cost item, accounting for 33% of
       OOP cost for those with medical scheme cover, and 41% for those without cover. These data
       suggest that medical scheme members still face some significant OOP costs for out-patient


                                                                                                                  47
           Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    health care, which may be related to limited scheme coverage for out patient care, and/or to
    co-payments charged by healthcare practitioners.

   The survey also collected expenditure data on professional fees for 1765 visits (87% of all
    recent visits). Where these cost data were supplied, they indicate that approximately 84% of
    those who were on a medical scheme did not incur any additional costs for fees from that
    visit.

   On average, where the patient did not have coverage, households spent approximately R112
    on fees per visit to GP‟s. The corresponding figure where patients were on medical schemes
    was R27. The tables below show estimated OOP expenditure by scheme and non scheme
    members at different service providers.

Average Out of Pocket Payment for Fees Associated with Various Out-Patient Provider
Visits: Individuals Without Medical Aid (Rands)

                                                                            Population
                           Mean     S.E.       95% C.I.            Deff
                                                                              Size
Public/Government Clinic     5.76    0.85      4.09     7.42         1.00     1861953
Private Doctor             111.66    3.68    104.45   118.87         1.65      533679
Workplace Clinic            38.78   37.47    -34.71   112.26         1.29         4904
Private Specialist          48.91    1.43     46.11    51.71         0.13         1520
Traditional Healer          39.33   18.21      3.61    75.06         0.41         5154
Dentist                    115.40    8.91     97.93   132.87         1.83         9326
Pharmacist                  25.16    8.19      9.10    41.21         1.71       20378
Other                       60.00    0.00     60.00    60.00 .                    3384



Average Out of Pocket Payment for Fees Associated with Various Out-Patient Provider
Visits: Individuals With Medical Aid (Rands)
                                                                          Population
                           Mean     S.E.       95% C.I.            Deff
                                                                            Size
Public/Government Clinic    11.89     5.93     0.26    23.52         0.59     31829
Private Doctor              27.44     7.77    12.19    42.68         1.97    223111
Workplace Clinic             0.00     0.00     0.00     0.00   .                2558
Private Specialist          16.09     4.45     7.36    24.82         0.65       3180
Traditional Healer          50.00     0.00    50.00    50.00   .                3384
Dentist                      0.00     0.00     0.00     0.00   .                 622
Pharmacist                   0.00     0.00     0.00     0.00   .              10779



   Medicines are the third major source of OOP cost, accounting for 20% of OOP costs for
    scheme members and 10% for non-scheme members.

   Expenditure data on medicines were available for 1708 visits (85% of all recent visits).
    Where cost data were supplied, some 80% of all visits did not incur separate costs for
    medicines. Approximately 27% of those who were not on a medical scheme and visited a
    private doctor, did not incur additional OOP costs for medicines – this is most likely to reflect
    visits to those general practices where the costs of medicines is included in the fee, whereas




                                                                                                  48
           Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    the balance of 68% did incur some additional OOP expenditure on medicines when visiting a
    private GP

   In the case of medical scheme members, 92% of visits by medical aid members had no out of
    pocket payment for medicines by the patient.

   On average, where the patient was not on a medical scheme, households spent around R34 on
    medicines associated with visits to Private GPs. The corresponding figure where patients
    were on a medical scheme was R13. The tables below show the corresponding OOP
    expenditures on medicines arising from visits by scheme and non scheme members to various
    other service providers.

Average Out of Pocket Payment for Medicines Associated with Various Out-Patient
Provider Visits: Individuals Without Medical Aid (Rands)
                                                                         Population
                           Mean     S.E.      95% C.I.          Deff
                                                                           Size
Public/Government Clinic     1.83    0.41     1.02     2.64       1.63     1787361
Private Doctor              33.76    5.33    23.31    44.22       1.96      505235
Workplace Clinic            29.11   29.06   -27.89    86.10       1.30         4794
Private Specialist           8.35    0.86     6.67    10.03       0.13         1520
Traditional Healer           0.63    0.74    -0.82     2.08       0.35         5154
Dentist                     16.33   12.79    -8.76    41.42       2.13         9326
Pharmacist                  33.14   13.71     6.24    60.04       2.08       20378



Average Out of Pocket Payment for Medicines Associated with Various Out-Patient
Provider Visits: Individuals With Medical Aid (Rands)
                                                                         Population
                           Mean     S.E.      95% C.I.          Deff
                                                                           Size
Public/Government Clinic     1.34    1.21    -1.04     3.71       0.26       31207
Private Doctor              13.35    4.76     4.01    22.69       1.55      226233
Workplace Clinic             0.00    0.00     0.00     0.00 .                  2558
Private Specialist           7.24    2.00     3.31    11.17       0.65         3180
Traditional Healer           9.00    0.00     9.00     9.00 .                  3384
Dentist                      0.00    0.00     0.00     0.00 .                   622
Pharmacist                  88.64   52.79   -14.89   192.18       2.16       10779



   Some 11% (N=1574) of all individuals living in the sampled households were on some form
    of chronic medication; information on monthly cost of medicines was available for 1319 of
    these individuals.

   The average monthly cost of chronic medication per patient is shown in the tables below for
    patients not on medical aid, and on medical aid respectively.

Average Out of Pocket Payment per Patient for Chronic Medication (Individuals Without
Medical Aid only) (Rands)
                                                                       Population
                           Mean     S.E.      95% C.I.          Deff
                                                                         Size
Public Sector Clinics        9.35    1.44    6.52     12.18       1.84   1503153
Private Pharmacy           114.22    9.49   95.60    132.83       1.53    220356
Other                       28.23    7.75   13.03     43.43       1.85       9326
                                                                                              49
           Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report




Average Out of Pocket Payment per Patient for Chronic Medication (Individuals With
Medical Aid only) (Rands)
                                                                           Population
                           Mean     S.E.      95% C.I.              Deff
                                                                             Size
Public Sector Clinics       38.34    24.45   -9.63    86.30           1.91     39457
Private Pharmacy           100.61    24.87   51.83   149.40           1.81    113401
Other                       95.00    31.87   32.47   157.53           2.21       6768




   Expenditure data on tests and investigations were available for 1754 visits (86% of all recent
    visits). Where cost data were supplied, some 68% of those who were not on a medical
    scheme and visited a private doctor, did not incur costs for tests, compared to 94% of those
    who were members of medical aid schemes.

   On average, where the patient was not on a medical scheme, households spent around
    R7 on tests associated with visits to Private GP‟s. The corresponding figure where patients
    were on a medical scheme was just under R2.

Average Out of Pocket Payment for Tests Associated with Various Out-Patient Provider
Visits: Individuals Without Medical Aid (Rands)
                                                                           Population
                           Mean     S.E.      95% C.I.              Deff
                                                                             Size
Public/Government Clinic    0.91      0.25    0.43    1.40            1.01   1773720
Private Doctor              6.85      1.80    3.32   10.37            1.82    450121
Workplace Clinic            0.00      0.00    0.00    0.00      .                4794
Private Specialist          0.00      0.00    0.00    0.00      .                1520
Traditional Healer          0.00      0.00    0.00    0.00      .                5154
Dentist                     0.00      0.00    0.00    0.00      .                5942
Pharmacist                  2.48      1.75   -0.96    5.91            1.26     15933
Other                       0.00      0.00    0.00    0.00      .                3384


Average Out of Pocket Payment for Tests Associated with Various Out-Patient Provider
Visits: Individuals With Medical Aid (Rands)
                                                                           Population
                           Mean     S.E.      95% C.I.              Deff
                                                                             Size
Public/Government Clinic     0.11     0.12   -0.12       0.34         0.25     28649
Private Doctor               1.79     1.17   -0.51       4.09         1.17    226066
Workplace Clinic             0.00     0.00    0.00       0.00   .                2558
Private Specialist           0.00     0.00    0.00       0.00   .                3180
Traditional Healer           0.00     0.00    0.00       0.00   .                3384
Dentist                      0.00     0.00    0.00       0.00   .                 622
Pharmacist                   0.00     0.00    0.00       0.00   .              10779


   Expenditure data on transport were available for 1740 visits (86% of all recent visits). Those
    who were members of medical schemes incurred on average, lower costs for transport per
    visit than those who were not medical aid members; this may be due to a greater provider
    choice amongst this subpopulation, or may be due to other factors, most likely socio-




                                                                                               50
                Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



        economic factors, where medical aid members, who tend to be wealthier, may also therefore
        live in closer proximity to service providers.

Average Out of Pocket Payment for Transport Associated with Various Out-Patient
Provider Visits: Individuals Without Medical Aid (Rands)
                                                                                     Population
                                    Mean       S.E.       95% C.I.          Deff
                                                                                       Size
Public/Government Clinic             28.55        2.74   23.17    33.92       1.60     1829558
Private Doctor                       59.93        6.99   46.21    73.65       1.52      484141
Workplace Clinic                      9.88        4.11    1.82    17.94       1.30         4794
Private Specialist                   11.13        1.14    8.89    13.37       0.13         1520
Traditional Healer                    2.66        3.11   -3.45     8.76       0.35         5154
Dentist                               7.26        5.69   -3.89    18.41       2.13         9326
Pharmacist                           81.11       40.44    1.79   160.44       2.09       20378
Other                                20.00        0.00   20.00    20.00 .                  3384

Average Out of Pocket Payment for Transport Associated with Various Out-Patient
Provider Visits: Individuals With Medical Aid (Rands)
                                                                                     Population
                                    Mean       S.E.       95% C.I.          Deff
                                                                                       Size
Public/Government Clinic              7.04        3.79   -0.40    14.47       0.56       28649
Private Doctor                       29.97        6.76   16.72    43.22       1.88      238404
Workplace Clinic                      7.97        4.85   -1.55    17.49       1.91         5942
Private Specialist                    7.24        2.00    3.31    11.17       0.65         3180
Traditional Healer                    0.00        0.00    0.00     0.00 .                     24
Dentist                               0.00        0.00    0.00     0.00 .                      4
Pharmacist                            2.94        3.26   -3.46     9.34       0.30       10779




Household-Level Data on Expenditure for Out Patient Health Care

       This section presents data at the household level; here cost data from recent health care visits
        for all members of the household are combined with information on frequency of visits per
        household over the previous 3 months, to obtain estimates of household OOP expenditure
        over time.

       The table below shows information on health care expenditure from those households which
        reported recent use of outpatient services. The data indicate that:

For households with no medical aid cover:

       Approximately 7% of lowest income households (R2,500 per month and less) reported recent
        use of private GPs; on average, these households paid R103 per GP visit, including fees,
        medicines and tests associated with the visit. For all households with medical expenditure on
        private GP visits, the average amount paid out of pocket amongst these low earning
        households was R93 per month , equivalent to an average of 9.8% of household income.




    Total visits over the past 3 months divided by 3.



                                                                                                     51
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   Around 12% of the higher earning households in the target market (R2,500-R6,000), reported
    recent use of private GPs, with an average OOP expenditure of R150 per visit and an average
    of R105 out of pocket payment per month, equivalent to an average of 2.9% of household
    income. This is a slightly higher proportion of household income than shown in the IES data
    (1.8%).

   OOP Expenditure on visits to public sector clinics represented around 0.6% and 0.4% of
    household income for each of the income groups respectively. Around 1/3 of the lower
    earning households reported recent use of public sector clinics and 28% of the higher earning
    households reported such use.

For households partially covered by medical aid:

   Medical expenditure data for the lowest income group with partial medical aid coverage
    should be interpreted with caution owing to small numbers of households in the lowest
    income group with medical cover. Data for the lowest income group are presented in the
    table for completeness, but are not repeated here.

   For households earning R2,500-R6,000 and partially covered by medical aid, some 35%
    reported recent visits to private GPs; OOP expenditure per visit was lower than amongst
    those with no cover, in keeping with partial medical aid coverage of these households, the
    average household out of pocket payment was around R40 per month, or 1% of household
    income.     Interestingly, OOP in this group was very similar to that for fully covered
    households (shown below); this is probably the result of some adverse selection (discussed
    earlier). This suggests that there may not be financial incentives for partially covered
    households to bring the rest of their members into cover.

For households with full medical aid cover:

   In the R2,500-R6,000 household income band with full medical scheme cover, just under 1/3
    reported recent visits to private GPs, with an average monthly payment of R38 or 1.3% of
    household income. However, owing to small numbers, these estimates are imprecise (see
    confidence intervals in the table).




                                                                                              52
             Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Household Out of Pocket Health Care Expenditure by Household Medical Aid Coverage –
Users of Services Only
Household Out Patient Health Care Expenditure by Income Band: Households with no medical aid only
              Household
                                                                                                  Proportion of
               Average                                                                                           Av. out of
                                                                                         est.      households
                Out of                        % HH                                                                pocket
                             SE      95% CI              SE        95% CI        deff population    using the                   95% CI
                Pocket                       income                                                              payment
                                                                                         size    service in past
               Payment                                                                                            per visit
                                                                                                     month
              per month
Expenditure on GP Services (fees, medicines, tests related to GP visits)
<R2501                 93    9.187 75 111       9.8% 0.015 6.7% 12.8% 1.95               357898        7%               103      83 123
R2501-R6000           105    9.879 85 124       2.9% 0.003 2.3% 3.4% 1.42                531401       12%               150     126 174
Expenditure on public sector clinics (fees, medicines, tests related to clinic visits)
<R2501                  6    1.951     2 10     0.6% 0.002 0.2% 1.0% 1.94                310641       35%                   7     4   11
R2501-R6000            16    4.524     7 25     0.4% 0.0011 0.2% 0.6% 0.98               141938       28%                26       6   46

Household Out Patient Health Care Expenditure by Income Band: Households partially covered with medical aid only
              Household
                                                                                                  Proportion of
               Average                                                                                           Av. out of
                                                                                         est.      households
                Out of                        % HH                                                                pocket
                             SE      95% CI              SE        95% CI        deff population    using the                   95% CI
                Pocket                       income                                                              payment
                                                                                         size    service in past
               Payment                                                                                            per visit
                                                                                                     month
              per month
Expenditure on GP Services (fees, medicines, tests related to GP visits)
<R2501                202 93.806 18 387 10.1% 0.0469 0.9% 19.3% 0.995                       3980      31%               202     18 387
R2501-R6000            39 18.144        3 75    1.0% 0.005 0.0% 2.0% 2.248                47104       35%                40      5 75
Expenditure on public sector clinics (fees, medicines, tests related to clinic visits)
<R2501            _           _       _   _      _        _       _       _       _        _           6%                10     10 10
R2501-R6000            11    4.613      2 20    0.3% 0.001 0.1% 0.5% 1.085                39804       19%                13      1 24

Household Out Patient Health Care Expenditure by Income Band: Households fully covered with medical aid only
              Household
                                                                                                  Proportion of
               Average                                                                                           Av. out of
                                                                                         est.      households
                Out of                        % HH                                                                pocket
                             SE      95% CI              SE        95% CI        deff population    using the                   95% CI
                Pocket                       income                                                              payment
                                                                                         size    service in past
               Payment                                                                                            per visit
                                                                                                     month
              per month
Expenditure on GP Services (fees, medicines, tests related to GP visits)
<R2501                249 82.721 86 411 12.4% 0.0413 4.3% 20.5% 0.775                       3242      41%               249      86 411
R2501-R6000            38 30.738 -22 98         1.3% 0.0102 -0.7% 3.3% 2.213              15919       31%                31     -15 76
Expenditure on public sector clinics (fees, medicines, tests related to clinic visits)
<R2501            _           _       _   _      _        _       _       _       _        _            _
R2501-R6000            19       21 -24 61       0.5% 0.005 -0.6% 1.5% 0.503                 4601       8%                   9   -12   31




Note: The cost data included in the table include the total costs of fees, medicines and special
tests related to the out patient visits. Costs were asked for all visits to all providers within the
preceding month. Frequency of visits over a three month period were used to obtain an average
number of visits per service per month, and previous month cost data for the household was used
to extrapolate average monthly cost, together with frequency of visits. Where cost data were
unavailable, average cost data for that household type (by medical aid coverage) was assigned.




                                                                                                                                      53
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Household Expenditure on Hospitalisation

Private hospital admissions

   1.5% of households in the sample reported one or more admissions to private hospitals (59
    households); of these, 85% were in households with medical scheme coverage, while the
    remaining 15% (16 households) had no medical scheme cover. For these households, the
    reasons given for using private rather than public hospital facilities predominantly related to
    trust and perceived better quality care in the private sector (8 households); 3 of these
    admissions were emergency admissions.

   Costs related to private hospital admissions for households not on medical schemes, ranged
    from R250 to R3,000, with one outlier of R35,000. The average cost was R1,175 per
    admission, excluding the outlier.

   Households partially covered by medical schemes that reported expenditure on private
    admissions (7 households only) paid on average R177, with a range of R98 to R648 per
    admission.

   Households fully covered by medical schemes, who reported expenditure on private
    admissions (6 households only) paid an average of R1673, ranging from R98 to R7,098.
    However, if we include all admissions in this group where no OOP cost was incurred,
    households on average paid R75 per admission.

Public sector hospital admissions

   486 households (14% of the sample) reported one or more admissions to a public sector
    hospital. Of these, 58% reported some OOP expenditure. Where expenditure occurred, this
    averaged R162 per household, ranging from R2 to R998.

Affordability of current healthcare utilization patterns

   The survey provides evidence that the burden of OOP expenditure for healthcare is
    substantial, particularly for non scheme members, but in some cases for medical scheme
    members as well.

   When households with no medical scheme cover were asked how the costs of the most recent
    hospitalization were paid for, approximately 29% indicated that these were paid from
    salaries, wages, pensions or grants. A further 14% indicated that they still owed money for
    the most recent hospitalization, and 0.5% indicated that the household had borrowed to pay
    for these costs.




                                                                                                54
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



      For medical scheme members, 90% of all costs had been paid by the medical scheme.
       However, these households still incurred OOP expenditures, as outlined above. These were
       paid using salaries etc in 7.4% of households, with money still owing by 2.7%.

      The table below provides data on household borrowing patterns for funding of medical
       expenses. The table shows that 9.3% of households in the lowest income bracket borrowed
       money in the past 3 months to fund medical expenses, and that this figure drops to 1.1% in
       the highest income bracket, with an overall average of 5.6% of the sample.



Proportion of Households that took Loans in the Past Three Months to Pay for Medical
Expenses by Household Income
                         Yes              No            Don't Know        NA             Total
<R2501                         9.3%            90.6%           0.0%            0.1%            100.0%
R2501-R3500                    3.8%            96.3%           0.0%            0.0%            100.0%
R3501-R4500                    1.3%            98.5%           0.3%            0.0%            100.0%
R4501-R6000                    1.1%            98.9%           0.0%            0.0%            100.0%
Total                          5.6%             94.3%          0.1%            0.0%            100.0%
N=3,542; Population Size=49,839,772

      When these data are analysed by medical scheme coverage, they indicate that 6.2% of
       households with no coverage borrowed in the past 3 months to fund medical expenses, while
       those with full coverage did not borrow at all for this purpose.

Proportion of Households that Took a Loan in the Past Three Months to Pay for Medical
Expenses by Household Medical Aid
                         Yes              No            Don't Know        NA             Total
None                           6.2%            93.8%           0.1%            0.0%            100.0%
Partial                        1.4%            98.6%           0.0%            0.0%            100.0%
Full                           0.0%            100.0%          0.0%            0.0%            100.0%
Total                          5.5%             94.4%          0.1%            0.0%            100.0%
N=3,529; Population Size=49,617,872


      The high burden of OOP expenditure for healthcare is also demonstrated by the unmet need
       for healthcare experienced by many households. The table below indicates that nearly 11% of
       household members who it was felt needed to see a doctor had not done so due to
       affordability constraints. This figure is significantly higher for the lowest income category,
       where 15.6% of such needed visits were not made, and lower in the highest income group,
       where the corresponding figure was 5.9%.




                                                                                                        55
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Proportion of Individuals in a Household that have Needed to see a Doctor but have not
seen a Doctor by Household Income
                        Yes               No           Don't Know         NA             Total
<R2501                        15.6%            84.2%           0.2%            0.0%            100.0%
R2501-R3500                   6.4%             93.6%           0.0%            0.0%            100.0%
R3501-R4500                   7.5%             92.2%           0.3%            0.0%            100.0%
R4501-R6000                   5.9%             94.1%           0.0%            0.0%            100.0%
Total                         10.8%            89.0%           0.1%            0.0%            100.0%

N=3,540; Population Size=49,838,248


      As would be expected, unmet need also correlates closely with medical scheme coverage.
       The table below shows that where there is no coverage in the household, 11.9% of individuals
       who it was felt needed to see a doctor did not do so due to affordability constraints. This
       figure drops to 5.4% for partially covered households and 0.8% for fully covered households.


Proportion of Individuals in a Household that have Needed to see a Doctor but have not
seen a Doctor by Household Medical Aid
                        Yes               No           Don't Know         NA             Total
None                          11.9%            88.0%           0.1%            0.0%            100.0%
Partial                       5.4%             94.6%           0.0%            0.0%            100.0%
Full                          0.8%             99.2%           0.0%            0.0%            100.0%
Total                         10.8%            89.0%           0.1%            0.0%            100.0%


N=3,527; Population Size=49,616,348

      This pattern of unmet need is also reflected in the proportion of individuals who have seen a
       doctor, but who were unable to follow that doctor‟s advice regarding treatment or medicine,
       presumably as a result of affordability constraints. The table below shows that 3.4% of the
       overall sample of individuals was in this position, with a higher percentage (4.3%) in the
       lowest income category.


Proportion of Individuals in a Household that have not Followed a Doctor's Advice
Regarding Treatment or Medicine by Household Income
                        Yes               No           Don't Know         NA             Total
<R2501                        4.3%             95.7%           0.1%            0.0%            100.0%
R2501-R3500                   3.2%             96.8%           0.0%            0.0%            100.0%
R3501-R4500                   1.7%             98.1%           0.3%            0.0%            100.0%
R4501-R6000                   2.8%             97.2%           0.0%            0.0%            100.0%
Total                          3.4%            96.5%           0.1%            0.0%            100.0%


N=3,538; Population Size=49,745,693


                                                                                                        56
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report




      When medical scheme coverage is taken into account, the data show that 3.6% of those with
       no cover were not able to either follow advice or purchase medicine, whereas this prob lem
       occurred in only 0.9% of individuals in households with full scheme coverage.


Proportion of Individuals in a Household that have not Followed a Doctor's Advice
Regarding Treatment or Medicine by Household Medical Aid
                         Yes              No             Don't Know       NA             Total
None                           3.6%              96.3%          0.1%           0.0%            100.0%
Partial                        3.4%              96.6%          0.0%           0.0%            100.0%
Full                           0.9%              99.1%          0.0%           0.0%            100.0%
Total                           3.4%             96.5%          0.1%           0.0%            100.0%


N=3,525; Population Size=49,523,792

      The pressure on low income households arising from the high cost of healthcare is also
       illustrated by the perceptions of households as to whether their household could afford
       sufficiently good treatment in hospital should a household member require this. Here the data
       show that 77.6% of the whole sample either strongly agree or agree that their household
       would not be able to afford sufficiently good hospital care should a household member
       require such care. This proportion is highest in the lowest income category (87.1% either
       strongly agree or agree with the statement) and lowest in the highest income category (56.6%
       either strongly agree or agree).

      As would be expected, responses to this question are also correlated with medical scheme
       coverage status. Where the household has no cover, 81.8% of households either strongly
       agree or agree that they would not be able to afford sufficiently good hospital care should a
       household member require this. Where the household has full cover, 6.4% strongly agree
       with the statement, while 27.3% agree with it. Presumably, these households have cover that
       provides limited private hospital benefits, and these concerns might therefore imply a lack of
       confidence in the public hospital system by scheme members whose benefits only fund
       treatment in public hospitals.

Proportion of Respondents Reporting That Their Household Would not be Able to Afford
Sufficiently Good Hospital Care if one Member got ill by Household Income
                    Strongly agree       Agree            Disagree     Don't know        Total
<R2501                         37.7%             49.4%         11.0%           1.9%            100.0%
R2501-R3500                    28.4%             47.9%         20.2%           3.5%            100.0%
R3501-R4500                    19.5%             49.1%         25.3%           6.1%            100.0%
R4501-R6000                    19.3%             37.3%         33.4%           10.1%           100.0%
Total                          30.0%             47.6%         18.4%           4.0%            100.0%

N=3,556; Population Size=5,005,480

                                                                                                        57
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report




Proportion of Respondents Reporting That Their Household Would Not be Able to Afford
Sufficiently Good Hospital Care if one Member got ill by Household Medical Aid
                   Strongly agree        Agree           Disagree     Don't know         Total
None                         32.2%               49.6%        15.6%            2.5%            100.0%
Partial                      21.6%               35.3%        32.0%           11.1%            100.0%
Full                          6.4%               27.3%        46.1%           20.3%            100.0%
Total                        30.1%               47.5%        18.4%            4.0%            100.0%
N=3,543; Population Size=49,832,899

      A similar pattern can be seen in relation to the affordability of treatment by a private
       doctor should this be required by a household member. Here, 77.4% of the whole
       sample either strongly agreed or agreed that their household would not be able to
       afford to pay the bill of a private doctor should someone in the household require this.
       As expected, this proportion was highest in the lowest income group (83.8% either
       strongly agree or agree), and lowest in the highest income group (65.1% either
       strongly agree or agree).

      Here too there is a strong correlation with medical scheme coverage. Where there is no
       scheme coverage, 79.7% of households strongly agreed or agreed that their household would
       not be able to afford to pay the bill of a private doctor should someone in the household
       require this. Where the household is fully covered, 40.4% either strongly agreed or agreed
       with this statement. This may suggest that the existing coverage does not provide adequate
       out of hospital cover, and/or the extent of co-payments still faced by medical scheme
       members when obtaining private medical care.




                                                                                                        58
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Proportion of Households who would be Unable to Afford to pay the Bill of a Private
Doctor by Household Income

                    Strongly agree       Agree           Disagree     Don't know         Total
<R2501                        33.9%              49.9%        11.8%            4.5%            100.0%
R2501-R3500                   25.8%              48.7%        22.1%            3.4%            100.0%
R3501-R4500                   19.5%              52.8%        21.2%            6.5%            100.0%
R4501-R6000                   19.4%              45.7%        28.9%            6.0%            100.0%
Total                         27.8%              49.6%        17.9%            4.7%            100.0%


N=3,559; Population Size=5,006,857


Proportion of Households who would be Unable to Afford to pay the Bill of a Private
Doctor by Household Medical Aid
                    Strongly agree       Agree           Disagree     Don't know         Total
None                          29.0%              51.7%        15.1%            4.1%            100.0%
Partial                       26.3%              36.1%        29.9%            7.7%            100.0%
Full                          11.0%              29.4%        48.4%           11.2%            100.0%
Total                         27.8%              49.5%        18.0%            4.8%            100.0%


N=3,546; Population Size=4,984,667


2.2.8      Key Conclusions of the LIMS Household Survey

This survey highlights some critical findings which are relevant to the conclusions and
recommendations of the LIMS process:

      There are approximately 5.1 million non-rural households with gross monthly income of
       R6,000 or less. These households comprise 21.9 million individuals. In the „target
       population‟ for LIMS, namely a monthly household income of R2,500 – R6,000, the survey
       estimates that there are 2.5 million households, comprising 11.5 million individuals.

      There is a relatively strong link between these households and the formal employment sector.
       In the target population, 86% of households had one or more members working in the formal
       sector. For the sample population as a whole, the equivalent figure is 50%. This suggests that
       formal employment will be a critical channel for the extension of medical scheme cover to
       the currently uncovered, and that effective employment based arrangements could reach up to
       86% of all households in the target population.

      The demographic structure of these households is an important consideration in developing
       policy regarding extension of medical scheme coverage. The survey has indicated that
       intergenerational household patterns are common, with as many as 13% of all individuals


                                                                                                        59
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    being grandchildren of the household head. This pattern must be taken into account in
    development of definitions of dependents, as well as in determination of subsidy policies.

   Within the target population, the survey indicates that 13% of individuals (1.49 million) are
    currently members of medical schemes. For the sample as a whole, 7.3% (1.53 million) are
    members of medical schemes. This indicates the extent of the opportunity presented by LIMS
    to extend cover to those currently without cover. For the target population, the survey
    suggests that the uncovered population amounts to 87%, equivalent to 9.95 million
    individuals. For the sample population as a whole, the uncovered population amounts to 93%,
    or 20 million individuals.

   These data also indicate that coverage is actually quite high at the upper end of the income
    ranges within the target market. For the highest income group (R4,500 to R6,000 per month),
    23% are medical scheme members. The equivalent figure for the next income group (R3,500-
    R4,500) is 12%. These data must be taken into account when considering the appropriate
    income threshold for proposed LIMS schemes, since there is a risk that if this is set too high,
    those in the current scheme environment may qualify for LIMS schemes, and may thus buy
    down to LIMS schemes. While the data did not provide details on the precise nature of
    schemes to which respondents belonged, it is possible that a proportion of those identified in
    the survey as having cover belong to Bargaining Council type schemes offering relatively
    limited coverage, or to medical schemes that offer only public sector hospital benefits.

   Of those without cover, 39% (589,000) live in households where there is already at least one
    member of a medical scheme. This group is a very high priority for extending cover via the
    LIMS process, since these households already have exposure to the medical scheme
    environment and at least one member already enjoying cover. It should thus be relatively
    easier to reach those without cover in these households, than to reach those in households
    where no-one currently has medical scheme cover.

   Scheme membership varies significantly by age group, and the pattern in the survey
    population closely resembles that of the current medical schemes market, with relatively low
    coverage amongst young adults and the elderly. Households with partial cover appear to
    carefully select those for whom to purchase cover, and there is some evidence that this
    selection is partially based on perceived or actual healthcare risk. Thus young adults are kept
    out of cover, with the main focus being those in employment (for whom productivity is
    essential) and those with higher healthcare need (as reflected in the data on adverse selection
    patterns). It will be important to ensure that pricing of LIMS products, and proposed subsidy
    policies, address these household coverage patterns, with the aim of ensuring that the young
    and the elderly are brought into coverage to the fullest extent possible.




                                                                                                60
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   The main barriers to entry into insurance coverage appear to be lack of employment by
    household members, and where one or more household members are employed, the main
    barriers appear to be absence of medical scheme subsidies/benefits from employers and
    affordability.

   Critically, there do not appear to be significant conceptual barriers or other objections to
    medical scheme coverage. Over 60% of respondents indicated a willingness to „risk pool‟ for
    health care costs and appeared to understand and support the concept of paying a relatively
    small amount each month to insure against larger expenses, even if this meant that the
    contributions would not necessarily be claimed back in any one time period.

   Together, these various results suggest that there is a strong interest amongst the target
    population in gaining access to medical scheme cover, and that lack of employment and/or
    facilitation by employers, and affordability, remain the major obstacles to extension of
    coverage in the current environment.

   As regards preferences for different medical scheme packages, the survey provides
    compelling evidence that low income households place greater value on coverage for out of
    hospital benefits than for hospital benefits. This was demonstrated in several sections of the
    survey. In response to a question about the essential components of a medical scheme
    package, cover for GP visits was regarded as essential to the purchase decision by twice as
    many respondents as compared to the proportion who regarded either cover for private
    hospitals, or comprehensive cover, as essential. This pattern of preference was confirmed by
    the willingness to pay questions, and by the discrete choice experiment, in which the relative
    value placed on comprehensive out of hospital care was approximately 50% higher than that
    placed on private hospital cover.

   It is important to note that these results should not be interpreted as indicating that low
    income households have no or low interest in cover for private hospital care. In fact, several
    areas of the survey indicate that low income households have significant interest in private
    hospital cover, driven at least in part by their perceived superior quality. However,
    respondents appear to appreciate that private hospital cover would likely be unaffordable for
    most.

   As expected, the responses to the willingness to pay and discrete choice questions
    demonstrate a positive relationship between household income and willingness to pay for
    medical scheme cover. There is a similarly positive relationship between formal employment
    and willingness to pay. The key target market for LIMS schemes, namely households earning
    between R2,500 and R6,000 per month, with at least one member in formal employment,
    valued a comprehensive PHC package at R105.49. This compares to a valuation of R58.75


                                                                                               61
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    for those households with no member in formal employment and with household income of
    below R2,500 per month.

   This pattern of preferences for out of hospital care makes sense in the context of the evidence
    on current healthcare utilization and expenditure patterns. The survey indicates that low
    income households with no medical scheme cover are still using private GPs to a significant
    extent, both for acute consultations and for obtaining chronic medication. As noted above,
    22% of all outpatient visits by non scheme members were to private GP‟s.

   Expenditure patterns on GP care further explain the strong preference for out of hospital
    cover amongst non scheme members. Here the data show that, for uncovered households, the
    average GP visit cost R112, compared to R27 for scheme members. Similar patterns were
    identified at the household level, where households in the target population who had used a
    GP reported spending an average of R105 per month on GP visits, equivalent to 2.9% of
    gross household income. These data suggest that at least some uncovered households are
    already spending a meaningful proportion of their gross income on private GP and related
    care, and this provides support for the strong interest expressed in purchasing cover for these
    services.

   The survey also provides evidence that current expenditure on private healthcare poses a
    significant affordability burden to low income households. This is confirmed by the
    significant proportions of households that had either not paid for out of pocket expenditure
    incurred, or had borrowed funds to pay for these expenses. Further evidence of this is shown
    in the relatively high levels of unmet medical needs experienced by many households without
    current medical scheme coverage, and by the large numbers of households who believe that
    they would not be able to afford adequate hospital or out patient care should one of their
    household fall ill and require such treatment.




                                                                                                62
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



2.3           RESULTS OF MODEL TO ESTIMATE POTENTIAL DEMAND FOR
              MEDICAL SCHEME COVERAGE AMONG LOW INCOME
              HOUSEHOLDS

The First Interim Report provided results of an initial demand forecasting model, which utilized
the existing official data (from the GHS 2004 and LFS 2004) to predict the likely take up of
LIMS membership, using different assumptions as to the premium per life per month, employer
subsidy (which together give the net premium payable per household per month), and assuming
different levels of household affordability and willingness to pay for medical scheme cover
(measured as the maximum percentage of total household income that a household will be willing
to spend on medical scheme premiums).

This demand simulation model has now been updated to incorporate the demographic data from
the LIMS HH survey. This updating was carried out by Deloitte, and the report on this work is
included as Annexure 6. The results of this model, using the LIMS HH survey data, are shown in
the tables below. It should be noted that the projections using the LIMS data are lower than those
using the GHS data. These differences are highlighted in table 2.3.1 below, using a baseline
example of a premium of R200 per month, with a 50% employer subsidy. The lower LIMS
estimates are mainly attributable to the fact that the LIMS survey was conducted on non-rural
households and has been weighted up to the total non-rural population of SA, whereas the GHS
survey represents the entire population. As the LIMS HH survey has been conducted more
recently, and focuses on those households most likely to be in the target market for LIMS, the rest
of this analysis and discussion focuses on these data.

Table 2.3.1: Comparison of GHS and LIMS data for R200 per month premium, 50%
employer subsidy
Total percentage HH income on             Total new li ves under cover    Total new li ves under cover
      medical scheme premiums                         LIMS HH Data                             GHS Data
                5%                                          1,549,630                          1,834,444
                8%                                          1,847,561                          2,705,765
                10%                                         3,166,450                          3,429,345
                12%                                         3,265,079                          3,731,190
                15%                                         3,589,066                          4,252,983


Table 2.3.1 shows the results for scenarios in which the premiums for LIMS schemes are
assumed to be R200 per life per month, with a 50% employer subsidy available to the employee
and for as many dependents as the employee is willing to cover. The table then shows the results
at different assumed household affordability levels. This implies that the household would be
willing to spend up to the defined percentage of gross household income on medical scheme
premiums.


                                                                                                     63
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



As the table demonstrates, even assuming a very low affordability rate of 5%, with premiums at
R200 per life per month and if employers provide the 50% subsidy assumed, the LIMS HH data
suggest that an additional 1.55 million lives should be able to afford medical scheme cover. If we
increase the affordability level to 8% of household income, (just above the 7.7% shown in GHS
2004 to be the current percentage of household income spent by those households in the target
market who already have full cover), then this number more than doubles, to 1.85 million
additional lives (GHS: 2.7 million). At higher affordability levels, these numbers increase quite
dramatically, with 3.2 million potential additional lives if household affordability is assumed to
be 10% of gross household income.

Table 2.3.2 shows results assuming that the LIMS scheme premium is reduced to R150 per life
per month, and again assuming a 50% employer subsidy. As discussed in more detail below, the
work of the Benefit Design Task Group has led to recommendations on a LIMS Minimum
Package, and detailed costing of this package suggests that LIMS schemes might well be able to
be offered to the market at premiums in the order of R150 per life per month in 2007. As a 50%
employer subsidy is also regarded as realistic, this table perhaps provides the most likely
estimates of the potential take up of new LIMS schemes by low income households, assuming
varying affordability levels.

As the table shows, at the lowest assumed affordability level of 5%, LIMS schemes could expect
to attract at least a further 1.85 million new lives, of which 183,000 would be from households
with current partial cover and the balance of 1.66 million from households without any cover at
all at present. If household affordability is assumed to be 8%, these numbers rise dra matically.
Here, we might expect a total of 3.2 million additional lives, with 295,000 from households with
partial current cover and the balance of just fewer than 3 million for currently uncovered
households. As the table also demonstrates, increasing assumed affordability levels from 8% to
10% or 12% does not make a substantial difference to the likely take up of LIMS schemes.
However, this pattern changes once the affordability level is increased to 15% of HH income, at
which point there is a substantial increase in potential take up, to a total of 5.13 million new lives.




                                                                                                    64
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report




Table 2.3.2: New lives under cover: R150 per month premium, 50% employer subsidy:
LIMS HH Data
                               New li ves from HH
 Total percentage HH
                              wi th current parti al          New li ves from        Total New Li ves
   income on medical
                                              cover           Uncovered HH               under Cover
   scheme premiums
                                   LIMS HH Data
          5%                               183,230                  1,664,331                1,847,561
          8%                               295,597                  2,969,482                3,265,890
          10%                              348,955                  3,240,111                3,589,066
          12%                              348,955                  3,240,111                3,589,066
          15%                              481,788                  4,654,335                5,136,123


Table 2.3.3, by contrast, shows the much lower take up of LIMS schemes that might be expected
in the absence of any employer subsidy, at the same level of medical scheme premiums as
modelled above, and with the same affordability assumptions. As this table shows, us ing the 8%
affordability level as an example, and assuming a premium level of R150 per life per month, the
absence of an employer subsidy leads to a 51% reduction in the expected take up of new medical
scheme cover, with an estimate of only 1.6 million lives coming into cover, as compared to the
3.26 million predicted with a 50% employer subsidy. A similar pattern applies at all affordability
levels and to the higher premium level as well.

Table 2.3.3: New Lives Under Cover: No Employer Subsidy
                                                       Total new li ves under   Total new li ves under
Total percentage HH income on medical scheme
                                                                       cover                    cover
                  premiums
                                                             R200 premium             R150 premium
                       5%                                            563,538                  678,433
                       8%                                            678,433                 1,600,867
                       10%                                         1,549,630                 1,847,561
                       12%                                         1,600,867                 1,847,561
                       15%                                         1,847,561                 3,166,450




                                                                                                   65
              Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



2.4              IMPACT OF TAX SUBSIDIES ON DEMAND FOR MEDICAL
                 SCHEME COVERAGE AMONG LOW INCOME HOUSEHOLDS

In September 2005, the National Treasury (NT) issued a paper3 containing proposals for reform
of the current system of tax deductions for contributions to medical schemes. These proposals
have subsequently been legislated and will become effective on 1 March 2006. A key motivation
of the reforms, according to the NT paper, is to encourage risk pooling through increased health
insurance coverage among low income employees and their families. The NT paper argues that
the proposed move to a capped, flat rate tax deduction on contributions would increase the
effective subsidy available to low income earners, and that this would significantly enhance the
take up of medical scheme coverage by low income earners and their families. Estimates in the
NT paper ranged from a 5% to a 20% increase in the number of principal members joining
schemes as a result of the proposed reforms, equivalent to an increase in total covered lives of
between 444,000 (at the 5% increase assumptions), and 1.43 million (at the 20% increase
assumption).

In response to the NT paper, LIMS commissioned Deloittes to undertake a brief analysis of the
likely impact of the proposed reforms on take up of medical scheme cover by low income
households in the LIMS target market. On the basis of this analysis, LIMS made a submission to
NT, which is attached as Annexure 7.

The key arguments made to NT were as follows:

      The NT proposals will make medical scheme coverage marginally more affordable to the
       relatively small group of low income employees who already obtain subsidies from their
       employers.

      As a result of low marginal tax rates, the incremental tax benefit of the NT proposals will not
       be sufficient to encourage presently uncovered employees to take out cover, nor to extend
       cover to uncovered dependents. The fundamental obstacle to extending coverage is the net
       cost to the employee of the medical scheme premiums. These remain too high for most low
       income employees, given low disposable income and other competing claims on such
       income.

      For these reasons, the current NT proposals are unlikely to make any material progress
       towards the stated objective of expanding risk pooling amongst low income families. The NT
       projections of incremental cost to the fiscus due to increased medical scheme take-up are
       therefore unlikely to materialise.
3
    National Treasury: Discussion Document on the proposed tax reforms to medical scheme contributions and medical expenses,
September 2005



                                                                                                                         66
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report




The LIMS submission therefore took the opportunity to make a new proposal to NT for a direct
premium subsidy to members of the proposed LIMS schemes, which would have the effect of
directly reducing the net premium payable by employees (and their employers where such
employers do provide a subsidy). The Deloittes paper commissioned by LIMS undertook an
analysis of the impact of such a direct premium subsidy. The key points made to NT regarding a
direct premium subsidy were as follows:

   To the extent that NT is willing to make an incremental contribution to the total quantum of
    subsidies available for medical scheme coverage, a much more direct and effective
    mechanism would be to provide a direct premium subsidy to members of qualified low
    income medical schemes.

   The LIMS process is developing proposals for a new class of medical schemes (LIMS
    schemes), open only to those earning below a defined threshold income. Such schemes may
    be exempted from the requirements of the PMBs, and would hence be able to offer lower cost
    and more affordable packages. This will hopefully be assisted by cost reductions throughout
    the healthcare chain, including administration costs, costs of medical services etc. It is hoped
    that such schemes could be introduced during 2007.

   NT should consider providing a direct premium subsidy to the members of such LIMS
    schemes. This could take the form of a defined Rand amount per beneficiary per month,
    provided directly to the medical scheme from NT, on the basis of returns from the scheme.
    The impact of these subsidies would be to directly reduce the premium payable by the
    member. The significant increase in affordability would also likely encourage employers to
    make further contributions to employee subsidies, and the net effect could therefore be
    substantial reductions in the net premiums actually payable by low income employees.

   The modelling carried out by Deloitte initially used LIMS data obtained from the Stats SA
    General Household Survey 2004, and showed the likely expansion of coverage amongst low
    income families in the household income range of R2, 000 – R6, 000 per month, based on the
    following assumptions:

    o   The total cost per beneficiary for LIMS package is R150, or R200 per month.

    o   Households can afford to spend 5%, 8% or 12% of gross household income on medical
        scheme cover.

    o   Employers provide a 50% contribution subsidy for employees and dependents.




                                                                                                 67
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    o   NT contributes a direct premium subsidy of either R100 or R50 per beneficiary per
        month.

   Using a baseline set of assumptions (households can afford 8% of gross household income
    and NT contributes R50 per beneficiary per month), the model provided the following results:

    o   A total of 3.27 million additional lives would be brought into cover, representing a 50%
        increase in current medical scheme membership

    o   The total cost of the premium subsidies to the fiscus would be approximately R2.07
        billion. This is in the same order of incremental contribution suggested in the NT
        discussion paper, which we argued will not be utilised as a result of the current NT
        proposals.

   Based on these observations, NT should seriously consider the alternative of a direct
    premium subsidy to LIMS schemes, and the LIMS stakeholders would be keen and willing to
    work directly with NT in developing this idea further, should this be required.

The LIMS submission acknowledged that this proposal for a direct premium subsidy could not be
incorporated into the current tax reforms, given the tight timetable for those. From subsequent
verbal discussions with NT officials on this issue, it appears that NT has an interest in pursuing
discussions on this concept, perhaps with a view to incorporating proposals along these lines into
the 2007/8 budget.

Subsequent to the submission of this paper to NT, Deloitte have updated the demand simulation
model to incorporate and adjust for the LIMS household data. The results of this work are
summarised in Tables 2.4.1 and 2.4.2 below.

Table 2.4.1 shows the projections assuming a 50% employer subsidy, and at premium levels of
R200 and R150 per month. These projections indicate the following key results:

   At a R200 per month premium, and at the 5% affordability level, some 1.55 million lives
    would be expected to be brought into cover in the absence of an NT subsidy. The addition of
    a premium subsidy of R25 per month makes relatively little impact here, increasing total lives
    under cover by just over 50,000. A subsidy of R50 per month would make a greater impact,
    increasing total lives under cover by approximately 300,000, but this impact remains
    relatively muted.

   At a R200 per month premium, and at the 8% affordability level, the impact of a premium
    subsidy is shown to be much greater. Here, lives under cover would increase from 1.85


                                                                                               68
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    million without a premium subsidy, to approximately 3.166 million with a R25 per month
    premium subsidy. The addition of a further subsidy at the R50 per month level would make
    only a margina l improvement on this initial impact, increasing total lives under cover to
    approximately 3.3 million.

   Where the premium is lower, at say R150 per month, the addition of a NT premium subsidy
    has a different impact. Here, the major impact is seen at the 5% affordability level, where a
    subsidy of R25 per month makes no impact at all on total lives under cover, but where an
    increase in this to R50 per month has a dramatic impact, increasing total lives under cover to
    approximately 3.166 million.

   With a premium of R150 per month, and at the 8% affordability level, the addition of a NT
    premium subsidy makes a relatively small impact, increasing lives under cover from 3.3
    million to approximately 3.6 million.

Table 2.4.1: New lives in medical scheme cover assuming 50% employer subsidy
                     Premium = R200 per month                    Premium = R150 per month
               NT subsidy     NT subsidy    NT subsidy     NT subsidy     NT subsidy     NT subsidy
                       R0            R25            R50             R0           R25              R50
    5%           1,549,630      1,600,867     1,847,561      1,847,561      1,847,561        3,166,450
    8%           1,847,561      3,166,450     3,265,890      3,265,890      3,589,066        3,589,066
    12%          3,265,079      3,589,066     3,589,066      3,589,066      5,136,123        5,251,823


A similar pattern is seen if we assume no employer subsidy and then model the impact of varying
levels of NT premium subsidy at the R200 and R150 per month premium levels. Here again, the
major impact of the NT premium subsidy is seen where premiums are R200 per month and at the
8% affordability level, where a R25 per month premium subsidy more than doubles the lives
under cover from 678,000 to 1.55 million. Where the premium is R150 per month, the major
impact is again seen at the lower affordability level of 5%, and with the higher NT premium
subsidy, where lives under cover now increase from 678,000 to 1.55 million.




                                                                                                   69
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Table 2.4.2: New lives in medical scheme cover assuming zero employer subsidy
                       Premium = R200 per month                    Premium = R150 per month
                 NT subsidy     NT subsidy    NT subsidy     NT subsidy     NT subsidy     NT subsidy
                         R0            R25            R50             R0           R25              R50
      5%             563,538       563,538        678,433        678,433        678,433        1,549,630
      8%             678,433      1,549,630     1,600,867      1,600,867      1,847,561        1,847,561
      12%          1,600,867      1,847,561     1,847,561      1,847,561      3,166,450        3,265,890


These data make a strong case for a direct premium subsidy from NT to LIMS schemes.
However, they also clearly indicate that the level of such subsidies should be carefully calibrated
in order to ensure maximum return to this fiscal investment, and that the key parameters to take
into account would include the net premium faced by households (which would result from the
combination of the monthly scheme premium and the level of employer subsidy), as well as the
proportion of household income that households are likely to set aside for medical scheme
premiums.

Further detailed results of this modeling exercise are shown in the various scenarios in Annexure
6. In particular, these show further breakdowns of the sources of new lives in medical scheme
cover (new lives from partially covered households, and new lives from uncovered households),
as well as the cost to the fiscus of varying levels of premium subsidy under different scenarios.

2.5           INTERVIEWS ON EMPLOYER ATTITUDES TO LOW INCOME
              MEDICAL SCHEMES

In addition to a detailed understanding of the needs of low income households, the needs and
attitudes of South Africa‟s employers are obviously also critical to the outcomes of the LIMS
process. This was clearly demonstrated by the data provided above, which show the relatively
large proportion of currently uncovered households in the target market where at least one
household member is in formal employment. In addition, the fundamental impact of employer
subsidies on potential extension of coverage was demonstrated in the results of the demand model
presented above.

In order to deepen our understanding of the needs and attitudes of employers to these issues,
LIMS commissioned Markinor (Pty) Ltd to conduct detailed, in-depth interviews with senior
executives at 40 employers using a questionnaire developed by the Demand Side Sub-Group. The
sample of employers was selected to as to ensure appropriate distribution between large, medium
and small employers, as well as geographical representivity. The sample was selected with the
generous assistance of the three major medical scheme broker and consulting houses – Alexander
Forbes, NMG and GlenrandMIB, who provided lists of their clients and approached the senior



                                                                                                     70
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



executives amongst those clients to arrange for their participation. The full report provided by
Markinor is attached as Annexure 8.

The key conclusions of this survey can be summarised as follows:

Current medical scheme coverage patterns, subsidies and attitudes to medical scheme coverage:

   8 companies (20% of those surveyed) indicated that 100% of their low income employees
    were covered. Coverage patterns varied widely in the balance of companies, with half of the
    overall sample indicating that less than 50% of their low income workforce is currently
    covered.

   The majority of companies (83%) indicated that they do offer medical scheme benefits to all
    employees, while 12% indicated that they do not. The balance have some of their workforce
    on medical schemes, and some on bargaining council arrangements and/or in-house provision
    of medical services to low income employees.

   All companies that do not provide a medical scheme benefit do offer low income employees
    the full range of standard employee benefits, including pension/provident fund, funeral and
    disability benefits, group life cover, housing subsidies etc. These companies indicated that
    medical scheme cover does not appear to be highest priority for either employees or trade
    unions at this stage. Lack of affordability of premiums was cited as a key reason for this low
    priority accorded to medical scheme cover.

   With one exception, all companies that do offer medical scheme benefits to low income
    employees do so on a voluntary basis. Most of these companies provide varying forms of
    assistance to their employees in their choice of medical scheme option. Such assistance
    includes advice or counseling by company staff, and/or the use of healthcare brokers and
    consultants to advice staff on medical scheme choice.

   36 of the 40 companies in the survey indicate that they provide a subsidy to employees who
    wish to take up a medical scheme benefit. The nature of the subsidy varied between
    employers, but 58% indicated that they offer a 50% subsidy on the cost of the medical
    scheme contribution. 22% offered a „two thirds‟ subsidy in line with then offered tax
    subsidies for medical scheme premiums, while the balance had a range of different subsidy
    arrangements. The four companies which did not provide subsidies indicated that they
    provide the medical scheme benefit on a „cost to company‟ basis.

   The majority of companies who do provide a subsidy indicated that they would not consider
    increasing their present subsidies in order to encourage additional employees to take up


                                                                                                71
             Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    cover. The major reason given was the current subsidies are already generous enough, and
    that the impact of additional subsidies would be negative for company profitability.

   The majority of respondents indicated that they provided subsidies for employees as well as
    dependents, with most allowing up to three dependents, and some allowing more.
    Respondents indicated that employee take up of the subsidy for dependents varied. Some
    companies reported that employees tend to cover dependents, and attribute this to the
    generous employer subsidy. Others indicate that employees find it increasingly difficult to
    maintain cover for all of their dependents due to high cost of premiums, and that this results
    in removal of dependents from time to time, and/or to patterns in which children are removed
    from cover once they are older.

   When asked for views on the maximum premiums affordable by low income employees,
    most respondents suggested that premiums should not exceed R150-R200 per employee per
    month, and should be below 10%-15% of the employees‟ income.

   In response to questions as to why employees do not take up medical scheme benefits when
    these are offered, respondents indicated that the major reasons were:

    o   Affordability, which appeared to be the key reason cited by most respondents

    o   Some employees being relatively satisfied with currently available free public sector
        medical services, particularly where employers allow employees time off to attend
        clinics.

    o   Lack of education and understanding regarding the benefits of medical scheme
        membership, and the concept of health insurance etc.

   Respondents were asked for their views on how important healthcare cover ranks amongst the
    priorities of their low income employees. Over two thirds indicated that healthcare cover is
    relatively low down in the order of priorities for low income employees.

   The respondents gave the following ranking to various priorities of their low income
    employees:

        1.     Education (School fees)
        2.     Electricity rates and water
        3.     Groceries
        4.     Taxi fare / Transport Costs
        5.     Health Care


                                                                                                72
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



        6. Repayment of Debt
        7. Housing costs
        8. Funeral Cost / Insurance

   Some respondents noted that their companies are frequently requested to provide loans to
    employees to fund medical expenses, and that this is one reason why they would prefer their
    employees to belong to a medical scheme.

   The majority of respondents believed strongly that increased take up of medical scheme cover
    by low income employees would be of benefit to the company, due to some combination of:

    o   Higher employee satisfaction and peace of mind.
    o   Better employee health, leading to increased productivity and reduced absenteeism.
    o   Reduction in requests for loans to fund medical expenses.

Medical Scheme Benefits

   In response to questions on which elements of a medical scheme package low income
    employees would prioritise, the majority of respondents indicated that day to day cover for
    GPs, medicines, dentistry and optical benefits etc was the most important element of private
    medical scheme cover. This was followed by cover for chronic medication and then private
    hospitalization. This was not the unanimous consensus of all respondents, with some
    indicating that private hospital cover would be a higher priority for low income employees.
    However, those with this view were in the minority.

Role of Trade Unions in Medical scheme decision making and distribution

   There were differing views on the role of Trades Unions in medical scheme negotiation,
    decision making etc. Approximately half of the respondents indicated that unions did not play
    a significant role in these decisions. The balance highlighted different roles for unions, with
    the majority of these indicating that unions appear to play a significant role in t he negotiation
    of scheme benefits, and decisions to change these benefits. However, unions then appear to
    play no or very limited roles in the implementation and administration of scheme
    arrangements, most of which is undertaken by the employers themselves.




                                                                                                   73
           Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Distribution Channels and Costs

     75% of respondents indicated that the costs of distribution and commission should be paid by
      medical schemes, rather than employers or employees. A majority also felt that such costs
      should be „built in‟ to their current premiums, rather than paid over separately by employers
      and/or employees.

2.6       OLD MUTUAL HEALTH CARE ANNUAL SURVEY

Old Mutual Healthcare (OMHC) agreed to incorporate, within its annual survey, some specific
questions relating to LIMS issues. Key conclusions of relevance for LIMS from the Old Mutual
2005 Healthcare Survey can be summarized as follows:

     Out of the total sample of 100 employers interviewed, some 84% of the workforce enjoys
      cover on company associated medical schemes. These may be either in the open or restricted
      membership schemes environment. Some 13% of employees have no medical cover that
      employers are aware of, and the balance of 3% of employees are covered by schemes not
      associated with the company for which they work. These would mainly be in respect of
      employees covered on their spouses‟ schemes.

     As shown in the figure below, when asked reasons why employees do not take up medical
      scheme benefits, the main reasons in order of importance were:

      o   The company provides on site primary care.

      o   Employees cannot afford contributions.

      o   Company does not provide medical scheme benefits to contract workers and/or to part
          time workers.

      o   Medical benefit offered but low take up by employees.




                                                                                                74
                                           Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Reasons why employees not covered


                                                       Reasons why employees are not covered




                                                                                                                                                                 Number of respondents
                         Number of employees

                                               50000                                                                                    15                  16
                                               40000                                                            13
                                                                                                                               41331                        12
                                               30000
                                                                                   18412
                                                                                       8                                                                    8
                                               20000
                                                            9333                                             8691
                                               10000           4                                                                                            4
                                                                                                                                                 4473
                                                   0                                                                                                        0
                                                         Contract      Part time Employeres                                Company Medical cover
                                                         workers -     workers - cannot afford                            provides on offered but
                                                         Company       company         the                                site primary low take up
                                                       policy not to policy not to contribution                            care clinic      by
                                                       provide cover provide cover                                                     employees
                                                                                                                                         targeted


   In response to a question as to why those employees who are offered scheme benefits do not
    take them up, the major reason cited by employers was the high cost of medical scheme
    premiums, leading to lack of affordability.

   Responses on the most important type of cover for their employees are shown in Figure 2.6.2
    below. This indicates that the majority of employers believe that day to day primary care is
    most important, and most would prioritise cover for these services with freedom of choice of
    provider for employees. This is followed by primary care cover in a designated provider
    network, and then by cover for private hospital care.

Importance of type of cover offered
                                                  Importance of type of cover offered to low income employees


                                 Hospitalisation in a private
                                                                              11            5            8           11
                                          hospital


                                      Hospitalisation in a public
                                                                          8        2    4       3
                                               hospital


                        Freedom of choice for day to
    % of respondents




                                                                          8         3               12                         26
                         day primary care provider


                       Day to day primary healthcare
                                                                         7          5               11                    19
                                 via DSP


                                                   Dental benefits 02 2



                                                   Optical benefits 02   3




                                                                                                                    10              9        8          7




                                                                                                                                                                                         75
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



2.7       DISTRIBUTION ISSUES

The Distribution Sub Group worked on various issues relating to the role of brokers, and
alternative distribution channels for the low income market. Activities carried out by this group
included:

     A focus group on the experience of brokers within the existing low income market.

     A review of the underlying cost drivers amongst brokers, in order to assess the feasibility of
      brokers adequately serving the low income market

     An analysis of alternative options for distribution of medical scheme products to the low
      income market, including some financial modeling of the costs of such distribution.

2.7.1     Broker Focus Group

TWIG SA were commissioned to facilitate a focus group comprising several brokers actively
engaged in the current market for low income medical scheme products, as well as some of the
providers operating in that market. The full report of the focus group, which generated very
interesting discussion and insights on a range of relevant issues, is attached as Annexure 9.

Some of the key issues identified in the report, of relevance to the question of distribution of
medical scheme products to low income households, can be summarised as follows:

Role of brokers and costs of distribution

     Brokers already have existing distribution channels and have perfected the ability to operate
      cost effectively in the lower income market.

     Brokers already have existing relationships with the unions and members and have their trust.

     Union endorsement of the product would be essential for buy in. Brokers have the knowledge
      to enable them to service the market and answer industry related questions.

     Brokers are already highly efficient in the take up of new business within this market sector.

     To create a sales and servicing force for a new product will be a duplication of manpower and
      costs.




                                                                                                   76
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   A separate sales force for a new product would be competing for the same market share as
    existing brokers. Take up of a new product could perhaps be hindered due to competition
    where income was being taken away from the current sales force servicing this industry.

   There would be unemployment if existing distribution networks were not used.

   Brokers act as a watchdog to ensure consistency of value for the member and quality of
    product and service delivery. It is in the broker‟s best interest for members to be happy.

   Brokers are able to give members independent best advice in line with current legislation as
    they market a suite of products and can ensure members are offered a choice.

   General consensus from respondents is that to remain cost effective in this market a broker
    has to deal with large volumes of business.

   Clients in this market actually require a high level of servicing, both at the time of sale and
    once they are members of the scheme. If service levels are high, retention levels tend to be
    high.

   The need for this level of service increases costs for brokers. If a low distribution cost is the
    intention, then brokers need to be relieved of the obligation to provide ongoing support to
    scheme members. Schemes should take on this responsibility themselves.

   The selling of the product (once union approval is received) is perceived to be cheaper and
    easier than the on-going service needed in this market, which is time consuming.

   Interaction with a large percentage of members in lower income options takes place at pay
    points or other work related areas.

   While simple brochures are useful in this market they cannot be used in isolation. There is a
    high level of personal interaction needed with employees in this market sector. This role is
    mostly filled by the broker.

   Limited access to telephones, illiteracy and language barriers are deterrents that limit the
    usage of call centres as a cost effective servicing tool in this market.

Attitudes of employers

   Employer willingness to pay subsidies for medical scheme membership is dependent on the
    size of the company as well as industry sector. Where there is strong union activity and


                                                                                                  77
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    unions operate on a national basis, most employers take it as a given that they will have to
    offer medical scheme subsidies. Where union influence is weak, employers have a stronger
    say in company policy.

   If a company moves to a cost to company package, brokers find it more difficult to persuade
    potential members of the necessity of joining a medical scheme.

   Respondents believe government needs to offer incentives that will encourage employers to
    offer healthcare subsidies for lower income employees.

   In larger companies where there is a company policy to offer healthcare, there is normally an
    allocated budget. Employers in this category feel fairly safe in offering a subsidy for medical
    scheme membership across all employees on a voluntary basis, as they know that only 50%,
    at most, of lower income employees will join the medical scheme due to affordability.
    Unused budgeted subsidy is channelled back into company profits.

   Medium and smaller companies with no existing policy in place regarding medical scheme
    membership for lower income employees, find it difficult to raise budget to accommodate a
    subsidy as it impacts on their bottom line.

Role of the trade unions

   Selling the benefits of medical scheme cover to union representatives does not involve as
    much documentation and in depth industry knowledge as in the more traditional medical
    scheme sector.

   Without union sanction, a broker will find it almost impossible to sell medical scheme
    products in the lower income market.

   If union influence is strong in a company then the broker can almost be guaranteed of closing
    a deal. Where union influence is weak there is a greater likelihood that the broker may lose
    the deal to the companies‟ official brokers.

   Consensus from respondents is that only in 1 out of 20 companies will a broker be able to
    approach the company directly and work from management downwards.

   Union endorsement of a product does not guarantee a book of business for brokers. However
    it does provide the approval and access to the workforce and the company which enable the
    broker to sell the product.




                                                                                                78
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   Regardless of union endorsement of a broker‟s product, company employment conditions
    take precedence.

   Usage of unions to endorse medical scheme products has happened because brokers have had
    very few other alternative distribution channels due to the perceived stranglehold on the
    market by larger corporate brokerage companies.

   Union endorsement of medical scheme products enhances the image of the union as they are
    seen to be embracing the national agenda of empowering black employees.

Overall conclusions

   Any new product offered into the lower income market would have to offer superior benefits
    cost effectively, with services to be provided mainly in the private healthcare sector. State
    facilities should only be incorporated as service providers where they are on a par with the
    private sector.

   There should be no duplication of existing distribution structures that would threaten the
    livelihood of brokers serving the low income market.

   Brokers would have to accept a lower commission.

   Without legislation to control costs, enforce membership and incentivise employers, success
    of a lower income product is questionable.

   Brokers should be enabled to negotiate with management and not have to fight their way up
    from the bottom of the organisation.

   There must be fair distribution opportunities and no exclusivity.

   A product for this market must have no co payments, levies or savings plans. It must be
    simple and facilities must be easily accessible.

   Rural dependents must be adequately provided for.

   There must be a national footprint for all providers of services.




                                                                                              79
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



2.7.2   Investigation of distribution channels

The Distribution Sub-Group conducted discussions and investigations into the pros and cons of
different distribution channels for medical scheme products in the low income market. A report of
the conclusions of these activities is attached as Annexure 10. The following conclusions can be
drawn from the analysis:

   Brokers and employers appear to be the most suitable channels for ongoing and expanded
    distribution of medical scheme products to low income households.

   Several other possible channels have been superficially investigated, and none of these appear
    suitable relative to brokers and employers.

   Further work is required in relation to some channels not yet investigated (e.g. stokvels,
    burial societies, NGOs or other civic groupings, trade associations etc).

   Financial modeling of the costs of servicing low income members by brokers suggests that
    this could be achieved at approximately R3.00 per member per month. For a single member,
    assuming a premium of R200 per month, this would amount to commission of 1.5%, well
    below of the regulated maximum of 3%. As the cost would not increase for a principal
    member with dependents, the overall percentage of premiums could be even lower, on
    average. Assuming a family premium of R500, for example, the corresponding percentages
    payable on commission would be 0.06%

   These analyses do suggest that brokers are an important channel for ongoing distribution of
    medical scheme products to low income households, and that they should be able to provide
    this service within the ceiling of the current 3% of premiums as required by the Medical
    Schemes Act.

   Employers are clearly a fundamentally important component in the effort to extend health
    insurance coverage to low income households. This is obviously the case in relation to
    subsidies, but also in relation to the distribution of medical scheme products. In the latter
    case, employers‟ active support for the broker and/or trade union can make a significant
    impact on the success of otherwise of efforts to extend coverage.

   Successful distribution of low income medical scheme products will therefore require a
    careful re-look at the roles, division of labour, and relative costs borne by brokers, scheme
    administrators and employers.




                                                                                               80
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



2.8       FINDINGS OF THE SUPPLY SIDE TASK GROUP

The Supply Side Task Group examined several aspects of the current high costs of medical
services, and approaches to resolving these. Key issues which were examined by the Task Group
included regulatory obstacles to the emergence of lower cost services, how to achieve sustainable
cost reductions within the current structure of the provider market, the role of alternative delivery
models in reducing the cost of supply, and the ability of the supply side to cope with a significant
expansion of demand that may arise from the establishment of LIMS schemes.

The Supply Side Task Group has produced a comprehensive report covering the methodology,
findings and recommendations of the four major work streams within the Task Group. The full
report is included in Annexure 11. In this section, the major findings are summarized and
interpreted.

2.8.1     Supply side constraints and capacity planning

This work stream carried out a detailed analysis of the current patterns of supply of all key
healthcare resources, including hospital beds, and all healthcare professionals, as well as a
detailed modeling exercise to ascertain the impact of an expansion in demand through the LIMS
process.

The key findings were as follows:

     There appears to be sufficient current hospital capacity, in terms of numbers of established
      beds, to cope with an increase in demand as a result of the LIMS process. South Africa has a
      similar total hospital bed to population ratio to comparable societies, such as Brazil, and the
      total hospital bed to population ratio in SA is 2.82 per 1000 (public and private). There is
      substantial regional variation about this average.

     There is clear evidence of a current shortage of nurses despite the fact that total registered
      nurse numbers appear to be adequate. Explanations for this discrepancy appear to be related
      to inaccuracy of official records of nursing numbers, and the fact that nurses maintain
      professional registration (which is the source of estimated numbers) even when not working
      as nurses within the healthcare system.

     The key shortage appears to be in the area of specialized nurses, in particular those qualified
      in critical care. This has led to ongoing shortages in both the public and private sectors at
      present, and to increases in salary inflation as a result of these shortages.




                                                                                                  81
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   These trends are aggravated by the global shortage of skilled nursing staff, and the
    „marketability‟ of SA trained staff, who appear to take up overseas posts in large numbers
    due to the relatively attractive remuneration on offer.

   Current estimates suggest that given current supply trends, there will remain a significant gap
    between supply and demand for nurses for at least the next 6 years, with a peak in 2009. The
    authorities are now acting to increase the output of nurses, which should address this gap over
    the medium to long term.

   The supply of community pharmacies appears to be adequate and relatively comprehensive,
    although there is substantial regional variation, and some regions appear to be undersupplied
    with community pharmacies (e.g. parts of Free State, Limpopo, North West and the Eastern
    Cape provinces).

   There are significant shortages of medical specialists in both the public and private healthcare
    sectors, with wide regional variations in the ratio of specialists to population. In SA, the
    public sector specialist to population ratio is 0.03 per 1000, and the private sector ratio is 0.11
    per 1000. The equivalent figures for a comparable country, such as Brazil, are 0.61 per 1000
    in the public sector and 0.41 in the private sector.

   This situation is becoming worse due to a decreasing trend in the supply of specialists.
    Official data suggest that the number of medical specialists has decreased from 3,881 to
    3,499 between 2000 and 2005, despite a substantial increase in the population over this
    period.

   The supply of GPs, dentists, optometrists, psychologists, radiographers and other health
    professionals appears to be adequate, but all show wide regional variation, as well as
    significant discrepancies between supply in the public and private sectors.

   This work stream commissioned Deloitte to develop a simple demand model in order to
    assess the impact of increased demand for hospital beds and healthcare professionals that
    might arise from the implementation of LIMS schemes. The full methodology and results are
    described in Annexure 11. Key findings of this modeling exercise are as follows:

    o   The major shortages identified are in certain categories of nurses and dentists. In
        particular, the model estimates that in order to meet the increased service load of the
        projected additional LIMS scheme members, a further 719 general ward nurses, 301
        theatre nurses, 394 ICU nurses and 46 dentists would be required.




                                                                                                    82
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    o   Beyond these clearly identified shortages, it appears that on current assumptions, there
        would be sufficient hospital beds, GPs and specialists to manage the increased demand
        that might arise from the implementation of LIMS schemes.

2.8.2   Sustainable price and cost reductions within the current healthcare delive ry system

It has been clear to all participating in the LIMS process that extending coverage to low income
families cannot be achieved by regulatory reform alone, but will also require significant reduction
in costs throughout the supply chain. While alternative delivery models may address this to an
extent over time, it cannot be assumed that these models will develop with sufficient scale or at
the pace required. This work stream therefore examined how to achieve cost reductions within the
current delivery system. In particular, it tried to assess whether it might be possible to achieve
some kind of formalized price differentiation or tiered pricing approaches within the demarcated
LIMS environment. In other words, if we can assume a clearly demarcated se t of
options/schemes, would it be possible to persuade suppliers to provide their goods and services to
members of these schemes at a different price to those provided to members of the existing
scheme environment? There are examples of this approach from other countries (e.g. The
Netherlands), and many providers already practice this in an informal way in South Africa.

Attempting to formalize such a system raises many difficult questions, including competition
issues, whether formalization would in fact undermine existing discounts for low income
members from some providers, and in the case of pharmaceuticals, how this would interact with
the current legislative requirement for a single exit price (SEP). There are also obvious and
important differences between pricing of services, such as consultations or hospital bed nights,
and pricing of products, such as medicines, since the latter can give rise to arbitrage opportunities
between differently priced environments.

At an early stage in these discussions, the following concerns and views were articulated by
policy makers in the Council for Medical Schemes and MTT:

   Some forms of formalized tiered pricing might allow market players to capture the full
    economic surplus from consumers by entrenching higher margins in the current medical
    schemes market. For this reason, such approaches could be considered, but perhaps on a time-
    limited basis with sunset clauses attached.

   Some aspects of the market are more amenable to market based cost reductions, through the
    use of contracting, alternative delivery systems etc, and that in these cases (say for health
    professional practices), these market approaches might be more effective and preferable to
    formalized tiered pricing.




                                                                                                  83
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   Other approaches, including reference pricing for surgical products and other non ethical
    products used in hospitals could also be examined as a possibility.

   Tiered pricing within the pharmaceutical industry would be complex to implement, in part
    due to the current Single Exit Price (SEP) environment. However, this was not ruled out as a
    possibility.

   The policy makers encouraged the LIMS process to develop as wide a range of options as
    possible for sustained cost reductions.

These concerns and issues were taken into account in the course of detailed discussions with
various stakeholders representing suppliers of goods and services in the private health sector. The
key results and findings from these discussions are briefly summarised below. Full details are
included in the Report of the Supply Side Task Group in Annexure 11.

Hospital Services

   As discussed in detail in the section on Demarcation of LIMS schemes, below, a key
    recommendation of the LIMS process is that LIMS schemes should not offer cover for
    private hospital care.

   This conclusion was reached during the later stages of the LIMS process, and the Supply side
    task group conducted extensive discussions regarding hospital cost reductions prior to this
    conclusion being developed. As the recommendation to exclude private hospital cover from
    LIMS remains simply a recommendation at this stage, and may change, it is still considered
    useful and important to assess the potential for cost reductions within the private hospital
    sector for the LIMS environment.

   Extensive discussions were held with individual hospital groups on these issues, and no
    collective discussions were held for fear of infringement of applicable competition
    legislation.

   The private hospital sector clearly recognizes that the current structure of private hospital
    costs, and the fee for service billing system in place, is not affordable for the LIMS target
    market. All of the groups have been engaged, to different degrees and in different ways, in
    developing their own approaches to providing lower cost services even prior to the LIMS
    process.

   Hospital groups appear committed to providing lower cost care to the LIMS environment,
    and were keen to entertain discounted reimbursement arrangements for marginal volume, as


                                                                                                84
     Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



well as other mechanisms for cost reduction. However, these groups were at pains to
emphasize that such cost reductions would be subject to a number of important constraints,
including:

o   Discounts should be on new business only, and the LIMS environment should not
    „cannibalise‟ existing business and margins. LIMS schemes would thus need to be clearly
    demarcated, and risks of „buy down‟ mitigated to achieve these requirements. This
    approach from the hospital groups demonstrates one of the key risks of including private
    hospital cover within LIMS schemes – i.e. that the emergence of a large volume of new
    demand for private hospital services would allow hospital groups to maintain the current
    high levels of pricing within the current medical schemes environment, while offering
    discounted charges to LIMS schemes. This could be seen as a form of price
    discrimination which might regarded as anti-competitive, and also as undermining the
    current thrust of the policy environment, namely finding sustainable cost reduction within
    the current medical schemes environment.

o   Cost reductions could emerge from discounts on ward and related tariffs to the extent that
    these costs are fixed. Hospital groups pointed out that a substantial proportion of costs are
    in fact variable, which limits capacity for such reductions. However, there were also
    indications of some spare capacity within the private hospital sector which would
    facilitate some measure of discounting. No hard evidence was produced on the
    relationship between fixed and variable costs within the private hospital environment.

o   Cost reductions could also come from more efficient use of, and lower prices for, ethical
    products and surgical disposables and consumables. Ethical products are already subject
    to SEP legislation (see further discussion below). However there remains a significant
    problem in the pricing and use of rebates in relation to surgical items and consumables,
    and this could benefit from changes to ensure greater transparency and perhaps from SEP
    type regulation as well.

o   A key constraint in reducing costs of hospital services relates to the shortages in supply
    of medical specialists. Any increase in use of private hospitals for LIMS will require
    additional input by specialists. While there may be some capacity for this, the relative
    shortage of specialists may mitigate against significant cost reductions from these
    disciplines in providing services to LIMS members. As specialist costs comprise a
    significant component of current private hospitalisation costs, this may prevent material
    reductions in overall cost of hospitalisation for LIMS members.

o   A further critical constraint will be the supply of nurses, in particular, highly trained ICU
    and theatre nurses.


                                                                                              85
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   Perhaps most importantly, these discussions highlighted the need for the development of
    significantly different models of hospital service delivery in order to materially reduce the
    costs of hospital and related services. Current models of delivery, based on the fee-for-service
    billing system, and with independent specialists also paid on a fee for service basis, are
    inherently costly. Efforts to reduce costs by „tinkering‟ at the margin may achieve some cost
    reductions, but these are likely to be neither substantial nor sustainable. Instead, alternative
    structures of hospital delivery need to be considered.

   Alternative delivery models for hospitals that were discussed included:

    o   Staff model arrangements, in which hospitals are able to employ medical and other staff
        in order to appropriately align incentives and thereby reduce costs.

    o   Development of closed networks of hospitals which would be operated on a low cost
        basis, including changes to design, operations and staffing.

    o   Changes to design and operation of hospitals allowing them to operate on a similar basis
        to mine hospitals.

   Critically, many of these changes are currently entirely constrained by existing legislation,
    much of which is outdated. Legislation and regulations governing employment of and
    revenue sharing with healthcare professionals prevents the emergence of efficient, low cost
    integrated delivery models. Hospitals and other healthcare organizations are prevented from
    entering into employment or other revenue or profit sharing arrangements with healthcare
    professionals. This prevents the alignment of incentives that is essential for the development
    of low cost models. In addition, the current hospital building and related standards may
    constrain the development of alternative, lower cost hospital infrastructure. These issues are
    discussed in further detail below.

   The development of partnerships between private hospital groups and the State would
    provide significant additional potential for the delivery of low cost hospital services for LIMS
    schemes. One example of such partnerships would be the expansion of private wards within
    public hospitals, with drafting in of management expertise from the private hospital sector.
    This would immediately expand capacity for delivery of low cost hospital services, while
    improving quality of services relative to current public sector delivery. This would also
    present the private hospital sector with additional profit opportunities, which should
    encourage active and sustainable participation by the private sector in these initiatives.




                                                                                                 86
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Pharmaceutical products

   Extensive discussions were held with representatives of the pharmaceutical industry,
    including IMSA and the PMA (representing much of the multinational pharmaceutical sector)
    as well as NAPM, representing the local generics industry.

   IMSA and the PMA put forward a detailed and concrete proposal for what they term an
    “Access Subsidy” system which would lead to lower medicine prices for LIMS schemes,
    while remaining compatible with the current SEP legislation. The key features of the Access
    Subsidy proposal can be summarised as follows (full details are provided in the Report of the
    Supply Side Task Group):

    o   Manufacturers would, on an entirely voluntary basis, offer a fixed and published discount
        to the SEP price of specific products for LIMS scheme members. This discount would
        apply for a defined period of time.

    o   The logistics of product delivery and dispensing would remain identical to the present
        system, with LIMS members obtaining their medicines from pharmacies or from
        dispensing doctors.

    o   LIMS members would need to have evidence of their membership of a LIMS scheme at
        the time of obtaining their medicine.

    o   The medicine would be provided to the LIMS patient at the normal SEP price, and this
        price would presumably be paid to the dispensing pharmacy or doctor by the LIMS
        scheme.

    o   The LIMS scheme would be refunded the difference between the SEP and the LIMS
        price retrospectively. This would require a specific set of arrangements whereby LIMS
        schemes would submit regular returns indicating the amount owing to them by each
        manufacturer to an independent body, which would audit these returns, then submit them
        to the manufacturers concerned who would refund the schemes, via the same independent
        body. It is possible that the Risk Equalisation Fund could be used as the mechanism for
        collecting and paying out the amounts owed by manufacturers to LIMS schemes.

   The proponents of this model point out that it would be contingent on the following
    conditions:




                                                                                               87
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    o   Clear demarcation of LIMS schemes from current schemes environment in order to
        ensure the ringfencing of the current SEP environment, and hence protection of pricing
        and profitability within that environment.

    o   Adequate mechanisms to ensure that LIMS members can demonstrate proof of
        membership of LIMS schemes.

    o   Some legislative change would be required to allow for modifications to the SEP price.

   This approach to price reduction for pharmaceutical products clearly has several significant
    advantages, including:

    o   It is consistent with the current SEP legislative framework, and would require minimal
        legislative change.

    o   It maintains the current logistics of medicines delivery and dispensing, and would not
        require the creation of entirely separate distribution channels, which would be highly
        problematic.

    o   It encourages voluntary participation by pharma companies in a process that could lead to
        material price reductions on some and perhaps many medicines required by LIMS
        patients.

   However, this approach also has significant disadvantages and/or problems associated with it,
    and these would need to be ironed out prior to any decision to implement such a system. Key
    problems include:

    o   There may in fact be substantial logistical and systems problems associated with payment
        at full SEP price by LIMS schemes, and then claiming back the access subsidy through a
        third party intermediary. While this system sounds very appealing in theory, the full
        practical implications remain to be worked out.

    o   The proposal assumes that current SEP pricing is appropriate, and places a substantial
        burden on that system to ensure appropriate pricing in the LIMS system. If, for example,
        there are systematic upward biases in the SEP pricing over time (either due to flaws in the
        SEP price setting process or to gaming of that process, or for other reasons), then the use
        a discount to that SEP for LIMS schemes would not be particularly effective in reducing
        the costs of medicines for LIMS schemes.




                                                                                                 88
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    o   Related to the above, this proposal might be seen as an attempt by the pharmaceutical
        industry to protect pricing in the current medical schemes environment and to avoid the
        downward pressure on pricing that would emerge from a dramatic growth in the overall
        size of the private market through the emergence of LIMS schemes. This is particularly
        important in light of the fact that the LIMS process will likely lead to a substantial growth
        in the total private pharmaceutical market, since many LIMS members will previously
        has been obtaining their medicines from the state. This systemic change will be beneficial
        for the pharmaceutical industry, and these benefits should ideally be shared to some
        extent with all medical scheme members.

    o   The voluntary nature of the proposal, while attractive in terms of the ease of
        implementation may not generate substantial participation. This will be particularly
        relevant in the context of the LIMS formularies for the LIMS Minimum Package and
        LIMS CDL conditions which are discussed in more detail below. If, for example,
        manufacturers of key formulary products do not volunteer a discount to the SEP for
        LIMS schemes, then this proposal will provide very limited benefit to LIMS schemes.
        This could also result in a situation where manufacturers benefit from the substantial shift
        in patients away from the state to the LIMS environment, but retain all of this benefit by
        adhering to SEP prices.

   Some other alternative approaches to achieving cost reductions of medicines in the LIMS
    environment were also discussed, although to a limited extent. These included the following:

   The supply of medicines from the State Tender to LIMS patients. While no details of this
    model have been discussed, it does imply that LIMS members would need to access the
    actual medicines sold to the State as part of the State Tender process. The clear advantages of
    this approach would be:

    o   LIMS patients would access medicines at the favourable prices obtained by the state. This
        would give these patients access to the State‟s purchasing power, and would in fact
        enhance (or least protect) the extent of that purchasing power.

    o   As there is likely to be a substantial overlap between the LIMS formulary in the LMP and
        the current State Essential Drug list, LIMS patients would be obtaining a strong price
        advantage across many, if not most, of the drugs they would be purchasing.

   This approach also has obvious, and perhaps fatal disadvantages, including:

    o   There would be a fundamental logistical problem in the distribution of State Tender
        medicines to LIMS patients. One approach would be to require LIMS patients to collect


                                                                                                  89
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



        their medicines from State facilities. However, this would defeat much of the service
        advantages and convenience that LIMS members would be seeking in joining LIMS
        schemes, and should thus be ruled out. An alternative would be for State Tender
        medicines to be distributed to retail pharmacy channels and dispensing doctors. However
        this would also pose major logistical problems, since these outlets would now have to
        keep separate (and separately labelled) stocks of the same medicines for current and
        LIMS scheme members. This would obviously also pose substantial risks of fraud.

    o   The pharmaceutical industry has also raised several additional objections to this
        approach, including the loss of transparency that is inherent in the current SEP system.
        Other objections from the industry are summarised in the Report of the Supply Side
        group.

   As between the alternatives outlined above, the Supply Side Task Team reached a clear
    consensus that some version of the Access Subsidy model would be strongly preferable to the
    distribution of State Tender Medicines to LIMS patients.

   This conclusion leaves open the question of how prices should be set in an Access Subsidy
    model. As outlined above, the view of the organised pharmaceutical industry is that this price
    setting should be done on an entirely voluntary basis, with companies deciding whether or not
    to participate, and if so, what discount to SEP to offer.

   There were several discussions on some alternative price setting mechanisms that could be
    applied to an Access Subsidy model. Some alternatives would be:

    o   A regulated standard discount to the SEP. For example, government could regulate that
        items on the LIMS formulary must be sold at a standard discount of say 30% to the SEP
        price.

    o   Use of the State Tender system for price setting for LIMS. In this system, all products
        which are successful in the State Tender system would have to be priced for LIMS at the
        State Tender price. This would effectively set a unique discount to the SEP for each
        product that is both on the LIMS formulary and on the State Tender. In this system, all
        other products not on the State Tender would continue to be sold to LIMS members at the
        SEP price.

   These alternative approaches to price setting have the advantage of securing better pricing for
    many products required by LIMS patients than the voluntary approach proposed by IMSA and
    PMA. The use of the State Tender for benchmarking would also have the advantage of
    retaining the voluntary approach, since companies are not forced to bid for the State Tender, but


                                                                                                  90
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    if they choose to, would do so in the full knowledge that their products would also be sold to
    LIMS members at the price offered to the State Tender system. The pharmaceutical industry
    has also pointed out other specific problems with this approach, in particular that it would
    introduce some instability into the pricing of medicines for LIMS, since the State Tender is only
    valid for one year at a time. On the other hand, the use of a standard regulated discount to SEP
    prices would have the disadvantage of moving away from voluntarism, and might generate
    significant resistance from the pharmaceutical industry. It would however lead to much greater
    price stability over time.

   While some consensus was achieved on the advantages of the Access Subsidy system as a
    means of achieving lower medicine prices for LIMS schemes, there was no consensus on the
    appropriate mechanism for price setting. These issues will need further debate and discussion
    before policy decisions are made.

   The organised pharmaceutical industry should however be complemented for its pro-active
    and very positive engagement with the LIMS process, and for its role in making detailed and
    concrete proposals aimed at reducing the costs of medicines.

Dentistry

   Discussions were held with representatives from the South African Dental Association
    (SADA) on various issues regarding the provision of services for a LIMS environment.

   SADA indicated that there is currently very limited capacity within the private dental
    environment, as the number of dental professionals has dropped considerably leaving excess
    demand in the market and no excess capacity in terms of hours available. Currently there are
    approximately 40 Prosthodontists, 100 maxillofacial surgeons in South Africa and dentist
    registration has dropped considerably over the past 3 years.

   SADA indicated that 85% of the current 2700 general dentists practice according to NHRPL
    rates. The remaining 15% work primarily in your affluent areas of Northern Johannesburg,
    Southern Suburbs Cape Town and Durban. These practitioners would not be easily convinced
    to accept patients on a reduced tariff basis for a LIMS market.

   These observations suggest that it will be difficult to obtain significantly lower costs for
    dental care for LIMS using the current model of care.

   However, the use of dental therapists may be a key part of the solution to lower cost dental
    care in the LIMS environment. The dental therapist could act as a point of entry in the




                                                                                                  91
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    primary care sector for the LIMS environment, with referral to dentists of those cases
    requiring more specialised treatment.

General practitioners

   A general willingness and commitment to the LIMS project was shown by the representatives
    of GP‟s in South Africa. Various comments and suggestions on how prices could be reduced
    in the sector for LIMS patients were presented. These are detailed below.

   Section 45 of the National Health Act No. 61 of 2003, which speaks to the relationship
    between public and private health establishments, creates a real opportunity for the
    development of innovative models of cooperation between the two sectors, especially as far
    as LIMS is concerned.

   However, there are a number of legislative and regulatory obstacles to the development of
    genuinely low cost delivery models involving GPs for the LIMS environment. These include:

    o   The legislated requirement that every private health establishment should have indemnity
        cover, which increases the cost of delivery of health care, and also creates resistance
        amongst professionals to expand their scope of practice, which is often essential in
        integrated, efficient delivery modes.

    o   Section 90 (1) (v) of Act 61, makes provision for the determination and publication of
        one or more reference price lists, which will not be mandatory. This means that it will
        not be possible to determine and/or publish mandatory tariffs or benefits for LIMS under
        the auspices of the National Health Act and that provision will have to be made for
        specific contracting arrangements with the providers of services to LIMS patients.

    o   The Medicine Pricing Regulations are likely to either directly or indirectly impact on the
        ability of practitioners to hold stock of and to administer medicines and/or injectables
        during treatment and/or consultations. This will inevitably have an impact on the LIMS
        process and the way in which the delivery of care is implemented.

    o   Section 59 (2), of the Medical Schemes Act obliges a medical scheme to pay such an
        account within 30 days of the receipt of such an account, but this payment may be
        reversed under certain circumstances, including bona fide errors. This has the specific
        consequence that in the LIMS environment, there will have to be very effective and
        efficient mechanisms of confirming membership eligibility and benefits, simply because
        it would not be realistic to expect suppliers of services to assume these risks in a low cost
        environment.


                                                                                                  92
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   The plethora of different protocols, guidelines and formularies, for obvious reasons add to the
    administrative burden of suppliers of health care services. In order to, amongst other things,
    manage these costs; it would be worth considering a standardisation of the protocols,
    guidelines and formularies for LIMS.

   Financial incentives should only be used to promote quality and cost-effective care and not to
    encourage the withholding of medically necessary care. Providers should not allow financial
    incentives to influence their judgments of appropriate therapeutic alternatives or deny their
    patients access to appropriate services based on such inducements.

   In summary, while the GP representatives indicated substantial interest in assisting with the
    development of low cost models for LIMS, they indicated that there are no obvious
    approaches that could lead to simple price reductions for LIMS patients. Instead, they made
    the useful points that there are already many low cost models in place, and that to assist in the
    expansion of these, as well as the development of more effective models, certain regulatory
    and administrative changes will be required. These would include:

    o   Changing key regulatory requirements which inhibit the emergence of integrated delivery
        models, and which raise the cost of practice as well as the risk aversion of practitioners.

    o   Easing some of the administrative complexity of current managed care arrangements in
        order to reduce the cost of practice.

    o   Ensuring that arrangements with GPs promote high quality of care and avoid incentives
        to under service.

Specialists

   Discussions with representatives of medical specialists, as well as industry experts, were held
    to determine the potentia l for cost reductions in a LIMS environment as well as integration of
    specialists into low cost delivery models.

   These discussions indicated that the prospects for material cost reductions from the current
    specialist private practice environment are very limited, for the following reasons:

    o   In the current environment, most specialists are close to capacity, and could not
        accommodate substantial additional patients.




                                                                                                  93
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    o   Most specialists already practice at or close to NHRPL rates (with exception of affluent
        suburbs in major cities), and perceive that these rates are too low. There would be
        substantial resistance to agreeing to lower rates.

    o   There is already a shortage of specialists, and this is being aggravated by a lack of young
        people entering the specialities due to lifestyle challenges, as well as remuneration
        incentives elsewhere. This will aggravate the capacity problem identified above.

   Reducing costs in the LIMS environment will therefore require alternative approaches to
    delivery of care. Examples of these might include:

    o   Use of state facilities for specialist care, in order for the economies of scale to be
        leveraged. This will assist to reduce overhead costs as well as provide cross training to
        state employed practitioners.

    o   Development of integrated delivery models, perhaps based on centres of excellence rather
        than numerous individual centres with limited capacity. These centres could be focused
        on primary healthcare and the potential to have specialists as supervisors can be
        considered. The specialists can act on a supervisory level, and have the nurse and GP
        doing the majority of the primary care work and consultation.

Optometrists

   General discussions with representatives of the optometric association proved fruitful and a
    sense of support for the LIMS process was expressed.

   In terms of access and cost reductions, numerous issues were discussed including the method
    of accessibility. One of the key issues raised was the fact that rural optometric services in the
    form of mobile clinics are already being utilised at a cost effective rate and making services
    more accessible. This model could be adapted for use in the LIMS environment. LIMS
    should attempt to capitalise on these successful services by ensuring that patients can
    continue to make use of these services and that the reimbursement from LIMS encourages
    participation.

   In addition to the LIMS minimum package that has been proposed, a set of tariffs for further
    services to be covered should be provided, so that the optometrists can ensure that the
    services are provided in accordance with the minimum package and that the re-imbursement
    will in fact cover the costs incurred by the optometrist.




                                                                                                  94
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   A reduced fee for LIMS patients will be seriously considered by the majority of optometrists,
    as numerous optometrists already do agree to reduced prices and altered payment terms for
    lower income patients. A standardised reduced fee with the guarantee for payment from
    LIMS schemes would allow optometrists to better plan their services. This will assist many
    practices with the current bad debt situation in lower income settings. The stipulated
    reimbursement figure will also allow for the service providers to use a certain amount of
    creativity and capitalise on economies of scale for the provision of services within the
    reimbursement amount.

   Motivation for high focus lenses and exceptional specialist services can still be referred to the
    State.

   The scope of practice of the optometrist will be increased during the course of 2006 and will
    therefore allow for consideration of further services for LIMS members including
    preventative and early detection services.

Medical Devices Industry

   Some discussions were held with the representative body for this industry, although these
    have not been extensive.

   A key problem raised concerns the persistence of requests for rebates by some hospitals and
    hospital groups, which is seen as creating perverse incentives to increase hospital cost with no
    gain in quality of care.

   Some proposals which emerged from these discussions included the possibility of reference
    pricing approaches for medical devices and disposables and surgicals.

   Further detailed discussion will be required prior to any formal proposals along these lines.

Medical Scheme administrators and managed care businesses

Although there has been extensive participation by representatives of the administrators and
managed care companies in various LIMS processes, there have been limited formal discussions
on possible cost reductions for LIMS schemes. Notwithstanding this, it is very likely that the cost
of administration and managed care for LIMS schemes will be substantially reduced for LIMS
schemes, for the following reasons:

   Administration and managed care fees are already carefully scrutinised by the Council for
    Medical Schemes, and are calibrated to total premium costs. With the significant reduction in


                                                                                                    95
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    LIMS premiums that is anticipated, administrators and managed care businesses will have
    little choice but to substantially reduce their charges for their services to LIMS schemes.

   This trend will be actively encouraged by the highly competitive market between open
    medical schemes, and between administration companies and managed care companies. The
    competitive dynamics in these markets will ensure that only players able to offer highly
    competitive pricing will participate in the market. It may well be that some players in fact
    choose not to participate, due to the low pricing that can be achieved.

   It is clear that administrators and managed care companies will need to develop different
    business models in order to profitably serve the LIMS market. Current business models,
    involving complex service offerings, intensive call centre support etc, will simply not be
    sustainable in the anticipated very low price LIMS environment. Instead, administrators and
    schemes will need to focus on benefit design and service offerings which minimise cost in
    order to ensure profitable participation in the LIMS environment.
.
2.8.3   Alternative delivery models

   This work-stream examined evidence, from SA and other countries, relating to the impact of
    alternative models of service delivery on cost effectiveness. Specifically, it looked at the mine
    model and similar models in SA, and assessed whether or not such models might have a role
    within a LIMS environment, and if so, what the obstacles are to the emergence of these
    models on a large scale.

   A significant amount of conceptual work has been carried out, which is described in detail in
    Section IV of the Supply Side Task Group Report. In addition to the conceptual work, several
    proposals have been put forward regarding various delivery models, including pharmacy
    delivery models, as well as models at the primary health care and major medical expenses
    level.

   The full report also includes a detailed review of international experience with alternative
    delivery models in low income settings, including the details of some specific case studies.

   The work of this group remains at an early stage, and it is clear that much additional work
    will be required prior to the emergence of a broad range of effective alternative delivery
    models that can cater for the low income environment.

   It is however critical to note, as has been done elsewhere, that there remain some significant
    legislative obstacles to the emergence of alternative delivery models. The key obstacle
    remains the various regulations in terms of which health professionals are restricted from


                                                                                                  96
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    being employed by corporate entities, and from sharing in revenues with non health
    professionals. These restrictions are discussed in more detail in the Regulatory Review
    section below.

2.8.4   Regulatory and legislative issues

   The Regulatory and Legislative Work stream has carried out an extensive and detailed review
    of all legislation and applicable regulations which impact on the current healthcare
    environment, and which may impact on the implementation of LIMS schemes. The full
    report, which includes detailed and specific recommendations, is provided in Section V of the
    Supply Side Task Group Report. Some of the major observations and recommendations are
    summarised here. Note that the relevant laws pertaining to private hospitals are dealt with in
    some detail in Annexure 13.

   The review highlights the many problems thrown up by the current regulation and legislation
    governing the training and scopes of practice of various healthcare professionals. These are
    fragmented, and interfere with the emergence of more cost effective delivery models for
    healthcare. There is a clear need for a centralized co-ordination of training and scopes of
    practices of the different professions that fall under different Statutory Councils. Provision for
    such centralization is made in Section 50 of the NHA through the establishment of a Forum
    of Statutory Health Professional Councils. It is critical that Section 50 be proclaimed as this
    Forum is in essence charged with the responsibility of coordinating the various Health
    Professional Councils, the development of coherent policies relating to the education,
    training, optimal utilisation and distribution of health care professionals, as well as to advise
    the Minister on the scopes of practice of the various professions.

   National Health Act (No. 61 of 2003): Aspects of the National Health Act (61 of 2003) pose a
    threat to the accessibility of healthcare and thus to the extension of accessibility which the
    LIMS process is aiming to achieve. In particular, the provisions of Section 36, relating to the
    Certificate of Need, are regarded as a possible obstacle to delivery of low cost high quality
    health care, and perhaps most importantly, to the required investment in infrastructure and
    systems to ensure the emergence of integrated delivery models. This section should be re-
    examined in the light of the need to encourage diverse models of care, and to encourage
    investment in such models.

   Competition Act (No. 89 of 1998): A strict interpretation of the Competition Act would limit
    the implementation of many of the proposals contained in this report regarding reduced
    pricing from suppliers of goods and services, alternative delivery models and provider
    networks amongst others. It is likely that the Competition authorities would be sympathetic to
    the socioeconomic and social benefit arguments that underpin many of these proposals, and it


                                                                                                   97
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    is recommended that the Department of Health and/or CMS and MTT approach these
    authorities to discuss some of the LIMS proposals in the light of current legislation.

   Medical Schemes Act (No. 131 of 1998). Several changes to the Medical Schemes Act will be
    required to facilitate the LIMS process. In some cases, cognisance will need to be taken of the
    Act in designing the details of LIMS proposals. Key issues discussed in detail in the
    Annexure include:

    o   Issues pertaining to the demarcation of a new class of LIMS schemes.

    o   Access to and use of provider networks by LIMS schemes, including designated service
        providers, and capitation type arrangements.

    o   The use of formularies and protocols.

   The Medicines and Related Substances Act (No. 101 of 1965). Changes to the Medicines and
    Related Substances Act will be required to facilitate the provision of medicines at a lower
    cost to the LIMS market. This applies particularly to the implementation of a proposed
    Access Subsidy model as discussed above, or any other related model which requires
    deviation from the current SEP legislation and associated regulations.

   Policy Document on Undesirable Business Practices (Health Professions Council of SA),
    Ethical Rules of the HPCSA: Various provisions of the Policy Document on Undesirable
    business practices and the Ethical Rules of the HPCSA constrain the emergence of various
    integrated delivery models that will form an essential part of the delivery of low cost
    healthcare services required by the LIMS environment. Whilst the intention of these
    documents and rules is acknowledged as being to ensure ethical and clinical independence of
    health professionals in order to protect patient interests, it is submitted that the current
    drafting and implementation of these provisions has significant unintended consequences as
    outlined above. Some specific constraints that emerge from these various restrictions are as
    follows:

    o   As currently applied, these various restrictions prevent the emergence of models in which
        health professionals may collaborate freely in different practice models. It thus becomes
        very difficult to establish models of care in which, for example, specialists supervise GPs,
        or other forms of shared practice that encourage appropriate leverage of scarce and
        expensive resources.
    o   Restrictions on corporate ownership also prevent investment by non health professionals
        in infrastructure and systems that could support the emergence of low cost delivery
        systems.


                                                                                                 98
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report




      o   Restrictions on employment of health professionals prevent the employment of health
          professionals in systems that can align incentives for cost effective care. An excellent
          example of such systems is found in South Africa‟s mine healthcare systems. These have
          historically been exempted from restrictions on employment of doctors and other health
          professionals. As a result, specialists and others have been employed in these systems,
          and are able to practice high quality, cost effective healthcare in the absence of the
          perversities of the fee for service reimbursement system. Some hospital groups are keen
          to move towards a model of employing doctor in order to bring down the costs of care,
          but are currently prevented from doing so by the HPCSA Ethical Rules etc.

     These observations lead to the clear conclusion that the current regulatory environment is not
      conducive to Integrated Delivery Models, and does not promote cost efficient healthcare
      delivery as there are no financial incentives for healthcare providers to follow cost effective
      practices or use cost effective products, particularly for the LIMS market. It is therefore
      recommended that the Department of Health and COMS work with the HPCSA to encourage
      a review and modification of these approaches. Any such changes should obviously ensure
      that healthcare providers are not put in a position in which financial incentives to influence
      their judgements of appropriate therapeutic alternatives or deny the patient‟s access to
      appropriate services based on such inducements.



2.9       FINDINGS OF THE BENEFIT DESIGN AND GOVERNANCE TASK
          GROUP

2.9.1     Alternative approach to a prescribed minimum package for Low Income Schemes

     The Benefit Design and Governance Task Group divided into two Sub-Groups. The Benefit
      Design Sub-Group focused on questions of an alternative approach to a prescribed minimum
      package for low income schemes, and the associated benefit design issues.

     Prior to the establishment of the Benefit Design Sub-Group, the broader Task Group began
      by examining the high level question of whether the existing PMBs are an obstacle to
      significant expansion of coverage in the low income market, due to the impact of PMBs on
      affordability. After extensive debate, there was a clear consensus in the group that the current
      PMB scope, and the cost of funding these benefits, does pose a significant affordability
      obstacle to the extension of scheme coverage to low income households. At the same time,
      the group recognised the important role of PMBs in the current scheme market. This led to an
      intensive discussion on the merits and problems of a more flexible approach to PMBs in the
      low income market.


                                                                                                   99
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report




   The strong consensus (but not unanimous) view of the group was in favour of proposing a
    modified approach to prescribed minimum benefits for schemes aimed at low income
    households. This is based on the view that the positive equity benefits of extending risk
    pooling to low income households, and making available to these households the
    opportunities to access employer and tax subsidies far outweigh the negative impact of
    reduced scope of a prescribed minimum package of services that low income schemes would
    be obliged to fund.

   As discussed in detail above, this conclusion is now supported by the evidence emerging
    from the LIMS HH survey, and the various employer surveys. These demonstrate that the
    current cost of medical schemes, even at the lowest end of current pricing, is beyond the
    reach of many low income households. This is true even where one or more household
    members are in formal employment. As a result, many employees are choosing to cover only
    themselves, or only some of their dependents. Many others simply cannot afford to purchase
    medical scheme cover for any household member, even where they would choose to do so,
    and where the employer is offering a subsidy. While there are other drivers of the high cost of
    current medical schemes, the scope of the current PMBs is clearly one of these.

   As noted in the First Interim Report, the conceptual proposal of a revised approach to a
    minimum benefit package for LIMS schemes (LMP) was discussed in depth with key policy
    makers at the Policy Workshop held on 19 August 2005. The various arguments for and
    against a LMP were discussed in depth at this workshop. The conclusions of the workshop on
    this issue can be summarised as follows:

    o   The policy makers indicated willingness to accept the principle of a LMP for a
        demarcated LIMS market.

    o   A critical aspect of this innovative approach in the low income environment would be the
        fact that the residual elements of the basic benefit package that are not covered by the
        LMP would be provided for LIMS members through the public sector, both in terms of
        finance and provision.

    o   The openness of the policy makers to this overall approach was subject to some critical
        conditions and further technical work. The key issues raised included:

        -     A LMP approach must not undermine the current PMB environment within current
              medical schemes. This will depend on the effectiveness of demarcation between
              proposed LIMS schemes and current medical schemes.




                                                                                               100
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



        -     The ultimate design of the revised LMPs should take into account both demand,
              preferences (as expressed in the household survey) and need, which households may
              not necessarily identify accurately.

        -     This support is explicitly based on the premise that the revised LMP does not
              constitute a lower standard of care, since the residual elements of „cover‟ are
              provided for LIMS members by the public sector, whereas members of mainstream
              schemes obtain all of their care within the current PMBs from the private sector.

        -     The LIMS process needs to focus explicitly on the challenges presented by
              HIV/AIDS within the potential LIMS market.

   Subsequent to these important conclusions of the Policy Workshop, the Benefit Design Sub-
    Group discussed alternative approaches to a LMP that would be compatible with the
    principles agreed at the workshop and summarised above.

   Firstly, the question of whether to adopt a single LMP or a flexible, modular approach was
    discussed. In these discussions, there was strong support for a more flexible approach,
    perhaps using a series of modular minimum packages, mainly on the basis that this would
    allow greater diversity of schemes and options on offer, allowing schemes to tailor packages
    more directly to the needs of specific groups of low income households. However, there were
    also several strong arguments in favour of a single revised minimum package for low income
    schemes, including:

    o   The need for simplicity in benefit design for LIMS schemes.

    o   Prevention of anti-selection between LIMS schemes.

    o   Facilitating the implementation of some form of risk equalization mechanism within
        LIMS schemes.

    o   Providing a coherent basis for direct premium subsidies from NT, as argued for above.

   Ultimately, there was clear consensus in the group in favour of developing a single LMP.

   A second question concerned the technical basis for specifying the LMP. The initial approach
    discussed was to utilise the same methodology as adopted in the development of the original
    PMBs, namely a modified Oregon approach. In this approach, all possible diagnosis and
    treatment pairs (DTs) are ranked from highest to lowest on criteria such as urgency and cost
    effectiveness. Once this ranking is completed, a total monetary cap can be applied, and all


                                                                                               101
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    DTs affordable within that cap become part of the PMB. From the LIMS perspective, the
    approach would have involved the following steps:

    o   Include all DTs excluded from the original list (mainly primary care related DTs).

    o   Rerank all DTs on the basis of revised cost effectiveness data as well as expert views on
        urgency etc.

    o   Apply a new, significantly lower Rand cap than was applied in the case of the PMBs.

   While this approach is technically feasible, and consistent with the approach of the current
    PMBs, detailed discussion within the Sub-Group and with key policy makers led to the
    conclusion that this approach is not optimal, and that a revised, simpler approach would be
    preferable. This revised approach is based on specification of a package of services (rather
    than DT pairs) that would have to be funded by LIMS schemes.

   The key reasons for preferring this approach over the modified Oregon approach are as
    follows:

    o   LIMS members will still have the protection of the current PMBs (and the subsequent
        extension to a basic benefit package), with the residual elements not funded by LIMS
        schemes to be funded (and provided in most cases) by the public sector. For this reason, it
        would not be appropriate to specify a sub-set of the current DT based PMBs for the LMP.

    o   Using a modified Oregon approach with a lower monetary cap may well result in an
        unworkable minimum package that would not be consistent with the realities of patient
        care delivery, and the experience of patients of the healthcare system. This might lead to
        potentially disruptive discontinuities in patient care, with some elements of treatment for
        a condition included but others excluded.

   A simpler approach, based on specification of a package of services rather than DT pairs,
    would be more consistent with delivery of patient care and avoid the potent ial discontinuities
    outlined above.

   Such an approach would also be more consistent with the Department of Health‟s view on an
    essential care package, as well as with the expressed preferences of low income households,
    which are likely to be expressed in terms of broad access to certain types of care (e.g. access
    to GPs, medicines, hospitals etc), rather than in the form of specific DT pairs.




                                                                                               102
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   On the basis of these conclusions, a small technical work group, comprising members of the
    Sub-Group and outside actuarial and clinical experts, was established to develop more
    detailed proposals.

   The key conclusions of this group, which have been discussed and agreed by the broader
    Benefit Design Sub-Group, are as follows:

   LIMS schemes should be required to offer a revised minimum package for LIMS schemes,
    the LMP, for the following key reasons:

    o   In order to limit anti-selection between LIMS schemes which would result from some
        offering benefits that others do not.

    o   To provide consumer protection against unscrupulous schemes offering low value and
        inappropriate package.

    o   To facilitate use of a risk equalisation mechanism within LIMS schemes.

    o   To provide a coherent basis for motivating for direct NT subsidies to LIMS schemes for a
        uniform minimum package.

   LIMS members will have the protection of the current PMBs, except that only some of this
    would be funded by LIMS schemes themselves, with the balance of this entitlement funded
    and provided by the public sector. It was noted that some of this entitlement within the public
    sector may be limited to the extent that such members are still required to pay user fees in the
    public hospital system. This problem will be effectively addressed if the proposal to eliminate
    user fees at public hospitals for all earning less than the proposed LIMS threshold is accepted
    (this proposal is discussed in more detail in the section on Demarcation of LIMS schemes
    below).

   The LMP must cost substantially less than the current PMBs within medical schemes in order
    to ensure affordability of LIMS schemes. Without this fundamental plank of the LIMS
    approach, there is no justification for separate LIMS schemes.




                                                                                                103
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   The following principles should govern the components of the LMP:

    o   The package must be consistent with expressed preferences of potential LIMS members
        as demonstrated by the LIMS HH survey.

    o   It should ideally be consistent with Department of Health‟s view on the essential care
        package, which includes primary health care (PHC), maternity care, emergency hospital
        care, and HIV/AIDS care.

   After extensive discussion and detailed research, the following LMP is proposed.
    Note that LIMS schemes would be free to offer additional benefits above the LMP,
    subject to some restrictions which are discussed in more detail below.

A. GP consultations:

A minimum of:
   3 GP visits for Principal Member (M).
   6 GP visits for M + 1 dependent.
   8 GP visits for M + 2 dependents.
   10 GP visits for M + 3 dependents.
   12 GP visits for M + 4 dependents.
   A Max of 12 GP visits per family per annum.
   A formulary comprising a limited set of procedures to be performed in GPs rooms.
   An additional minimum of 3 GP visits per annum per beneficiary who has one or more LIMS
    PMB conditions.
   It is assumed that most LIMS options will contract with Designated Service Provider (DSP)
    GP networks, and are likely to offer unlimited GP consultation benefits.
   GP networks may utilise nurses and other service providers, but the package must provide
    access to GPs where this is required.

B. Pathology and radiology investigations ordered by GP, subject to a defined formulary

C. Dental consultations:

A minimum of:
 2 dental visits per beneficiary per annum for basic conservative and restorative dentistry.
 No cover mandated for advanced dentistry or dentures.




                                                                                                104
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



D. Optometry:

A minimum of:
 One eye test per 24 months.
 One pair of spectacles every 24 months, subject to clinical criteria and a formulary.
 LIMS schemes may elect to impose reasonable financial limits, as well as protocols related to
   lens prescriptions.
 A basic frame should however be covered in full.

E. Acute Medicines

   A minimum of a defined formulary for acute and acute intermittent medications, based on the
    DoH Essential Drug List, with suitable modification.
   Acute medications on the formulary will be allowed for a maximum of one month.
   Acute intermittent medications will be allowed for a maximum of 6 months. Examples of
    acute intermittent conditions to be covered by the formulary for 6 months include:
    o depressive disorders
    o allergic rhinitis
    o pain control of shingles
    o osteoarthritis
    o anaemia
    o pyridoxine deficiency
    o rheumatic fever prophylaxis
    o TB therapy

F. Chronic Medication

   A minimum of a defined formulary, based on clinical protocols for the following 14 LMP
    CDL conditions:
    o Asthma
    o Bronchiectasis
    o Cardiac failure/cardiomyopathy
    o Chronic renal disease
    o COPD
    o Coronary artery disease
    o Diabetes mellitus type 1 and 2
    o Dysrhythmias
    o Epilepsy
    o HIV
    o Hyperlipidaemia



                                                                                            105
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    o   Hypertension
    o   Parkinson‟s disease
    o   Rheumatoid arthritis

   A proposed formulary for discussion is attached as Annexure 12 for acute, acute intermittent
    and the 14 LMP CDL chronic conditions.

G. Emergency transport to a public hospital (or private hospital in cases of life threatening
emergency).



H. Maternity Care Services:

   Not mandated as part of minimum package.

I: Specialist Benefits

   Not mandated as part of minimum package.

2.9.2   Commentary and justification for the proposed LMP package

   This proposal is consistent with all of the principles discussed at the Policy Workshop on
    19 August 2005, as well as with the principles agreed in the Benefit Design Sub-Group and
    described above.

   The LMP proposed is intended to provide a minimum of a reasonably comprehensive
    package of out of hospital, primary healthcare. This is consistent with the preferences
    expressed by low income households in the LIMS HH survey for comprehensive primary
    care benefits in preference to private hospital cover.

   The proposed LMP is also consistent with the objective of ensuring that the LMP be
    delivered at very affordable cost levels, without compromising the level or quality of care
    obtained by LIMS scheme members. The inclusion of a more extensive LMP, particularly
    with a broader CDL list, specialist cover and/or private hospital cover would have
    significantly increased the minimum cost of the LMP, thus undermining t he overall
    objectives of the LIMS process, and also eliminating the need for a separately defined LMP.




                                                                                             106
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   The current list of 25 CDL conditions was reduced to 14 on the basis of consensus amongst
    the clinical experts working in the Technical Task Group as to which CDL conditions were
    absolutely essential to mandate cover for, versus those for which mandatory cover would be
    good to have, but would not be critical.

   It was decided to utilise modified versions of the State EDL as the basis for the LIMS acute
    and chronic formularies, rather than the protocol based approach used in the current PMB
    CDLs. The principal drivers of this recommendation are that the State EDL lists are
    reasonably comprehensive, once key gaps are filled, and also that this approach would be
    much more affordable than using the approach of the current PMB CDLs.

   Serious consideration was given to the inclusion in the LMP of an ante-natal and a hospital
    maternity benefit. Ultimately, it was decided to exclude these from the LMP. One of the main
    reasons for this was the recommended exclusion of all private hospital cover from LIMS
    schemes, along with a proposal to eliminate user fees at public hospitals for those earning
    below the LIMS threshold (see below). A second argument in favour of this approach was the
    concern, based on wide experience in the current low income medical scheme environment,
    that provision of maternity hospital benefits drives a very high caesarean section rate, which
    is highly cost inflationary. In the absence of hospital cover for deliveries, it was felt that
    mandating an antenatal care benefit was inappropriate, since this would lead to discontinuity
    between the antenatal care and the delivery. LIMS schemes would not be prevented from
    offering an out of hospital maternity benefit, including for example home deliveries, should
    they choose to do so.

   Serious consideration was also given to the inclusion in the LMP of cover for emergency
    stabilisation in either private or state hospitals. Inclusion of this would have been compatible
    with the general exclusion of hospital cover for non emergencies. However, after extensive
    discussion internally, and with representatives of the private hospital industry, it was
    concluded that the definitions of emergency stabilisation remain too open to interpretation.
    As a result, inclusion of this as a mandated benefit would in all likelihood lead to numerous
    admissions to hospitals, thus either raising the costs of LIMS schemes significantly, and/or
    increasing the exposure of private hospitals to unfunded patient care.

   LIMS schemes will be able to offer any additional benefits they choose, subject only to the
    exclusion of cover for hospital care in either the private or public hospital sectors. This
    recommendation is explained in further detail in the Demarcation section below.

   Out of hospital specialist visits were excluded from the mandatory LMP due to the exclusion
    of private hospital cover. It was felt that it would be inappropriate to mandate out of hospital




                                                                                                107
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    specialist cover in the absence of any hospital cover within LIMS schemes. This exclusion
    would not prevent LIMS schemes from adding specialist cover benefits.

2.9.3   Cost estimates for the LMP

   The Technical Task Group carried out detailed estimations of the likely cost of the proposed
    LMP, using a combination of utilisation and cost data from existing medical scheme and
    consultant databases, as well as the pricing and experience of some providers in the current
    capitated low income primary care environment.

   The results of the cost estimation exercise are shown in table 2.9.1 below. The cost estimation
    includes a small amount for solvency build up, and makes certain assumptions as to the
    underlying cost and utilisation of each of the mandated benefits, as well as for administration
    and distribution costs.

   As the table indicates, the current estimate of the cost of the LMP in 2006 is approximately
    R108 per beneficiary per month. It is essential to note that these costs should be regarded as
    indicative only. Actual LIMS packages offered in the market are likely to cost somewhat
    more than this, due the inclusion of additional services beyond those mandated in the
    proposed LMP, and due to variance between the assumptions utilised here and actual
    experience in the market. In particular, it is very likely that most schemes will offer much
    more comprehensive GP cover, which in many cases may well include unlimited access to
    GP consultations. This is currently the practice in most low income medical scheme options
    utilising GP networks. The costing provided here does not take this benefit into account, and
    is based explicitly on the assumed utilisation in the LMP.




                                                                                               108
             Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



Table 2.9. 1 Estimated cost of LMP in 2006, per beneficiary per month

Discipline                                               Total estimated
                                                                    cost
Basic dentistry                                                    R5.50
GP CDL v isits                                                     R6.73
GP v isits                                                       R15.00
Radio logy X rays                                                  R0.98
Pathology                                                          R1.89
Optical visits                                                     R1.84
Optical frames and lenses                                          R8.00
HIV/AIDS                                                           R3.00
Primary medication                                               R30.00
Emergency transport                                                R4.00
Procedures in GP‟s inco me                                       R12.00
Admin istration cost                                             R10.00
Reserve build up                                                   R5.89
Distribution cost                                                  R3.00
                                                                R107.83



2.10         DEMARCATION OF LIMS SCHEMES                             FROM       THE      CURRENT
             MEDICAL SCHEMES ENVIRONMENT

    From the outset of the LIMS process, a clear guideline for the LIMS process from the MTT
     and CMS has been that any recommendations regarding modification of PMBs would need to
     apply specifically to some form of demarcated schemes or options open only to those
     qualifying as „low income‟ and should not undermine the current medical schemes market in
     any way. This clearly raises the question of how such LIMS schemes might be demarcated
     from the current market and the Demarcation and Governance Sub-Group has worked
     extensively on this issue.

    The basic conceptual proposal of a demarcated LIMS environment was discussed in depth at
     the Policy Workshop held on 19 August 2005. The policy makers at the workshop gave
     strong in-principle endorsement for the concept, provided that satisfactory proposals are
     made on the following key technical issues:




                                                                                                109
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report




    o   How to demarcate a low income scheme environment:

        -     What is the appropriate income level for entry into a low income schemes
              environment?
        -     Should individual or household income or per capita household income be utilized?
        -     How should income be proved and how to avoid fraudulent entry?

    o   How will movement across the demarcated boundary be regulated in order to prevent
        buy-downs, adverse selection through buy-ups, family splitting and other forms of
        gaming which will undermine the viability of both the current schemes environment and
        the potential new LIMS market?

   In order to address these issues, a small Technical Task Group was set up, comprising
    members of the Sub-Group as well as actuarial and other experts from the industry. The
    proposals from this group are as follows:

2.10.1 Definition of appropriate income level for demarcation of LIMS schemes

   Membership of LIMS schemes will be limited to principal members who fall below a defined
    level of personal income level, and their beneficiaries.

   The proposed LIMS threshold should be a personal income of R6,500 per month. This is in
    2005 Rands, and will need to be inflated to the year of implementation, and then annually, as
    suggested below.

   The income threshold should be inflated annually by CPIX. As the proposed level of R6,500
    per month has been set at 2006 levels, this should be inflated by CPIX between July 2005 and
    October 2007 (or thereabouts) in order to arrive at the appropriate number for January 2007.

   The principal member will need to make a declaration, as part of the application form, that
    he/she is the highest income earner in the household.

   The income qualification will apply only at first entry, and will not be re-applied annually.

   A key principal is that the threshold should be set in such a way that it does not trap too many
    people between LIMS and current schemes. In other words, it should avoid creating a
    situation in which large numbers of people who cannot afford current schemes are excluded
    because they earn above the LIMS threshold income level.




                                                                                                    110
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   CMS should monitor this situation annually, and the level of the LIMS threshold income
    level should be formally reviewed every three years.

   The income level has been set at R6,500 per month for the following reasons:

    o   The initial level of R6,500 per month has been set to achieve the multiple objectives of
        maximising opportunities for low income families to enter LIMS schemes, while also
        avoiding buy-downs from those who can afford to belong to current medical schemes.

    o   The LIMS HH survey shows that approximately 23% of individuals in households with
        gross monthly household income of R4,500-R6,000 are covered by some form of medical
        scheme, and that the balance are not covered by any scheme at all. At income levels
        below this, levels of scheme membership are very low.

    o   It is thought quite likely that the majority of the 23% covered are either in Bargaining
        Council type schemes, or in regular medical schemes but where the package covers
        public hospitals only. Both of these categories would be appropriate LIMS members, and
        therefore the risk of creating a systemic buy down from mainstream medical schemes is
        relatively limited.

    o   The buy down risk will be even further mitigated by the strong benefit differentiation
        between LIMS schemes and current schemes through the exclusion of private hospital
        cover proposed for LIMS schemes (see below). In this context, the threshold level should
        be set to ensure that we err on the side of including more rather than fewer low income
        households.

   This will meet the agreed objectives of administrative simplicity, while at the same time
    encouraging maximum entry into the market, and avoiding buy downs by those in current
    cover.

2.10.2 Administration of the entry criterion

   Membership of LIMS schemes will be open to formal sector employees, as well as self
    employed and informally employed individuals, and their beneficiaries.

   The income of the applying principal member will have to be demonstrated to the scheme
    administrator, either on the basis of official payroll data supplied by the employer, or by
    furnishing a SARS tax return. Applications will not be accepted in the absence of one of these
    forms of proof of income. However, no other information will be required as proof of
    income.


                                                                                              111
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report




   At a later stage, it may be possible to open up membership to defined associations or groups
    (e.g. taxi associations, burial societies etc), where the overall profile of the group could serve
    as sufficient proxy for income of members. It was however agreed that it would be premature
    to allow this form of entry at the initial stages of LIMS. The main reason for this is that these
    groups are not currently formally established, and it would therefore be difficult to develop
    criteria that such groups would have to meet at this stage.

   It is recognised that such groups could provide useful intermediation for informal sector
    employees, and it may well be that brokers will try to encourage the formation of such
    groups. This development should be encouraged, and work will need to be done to define
    appropriate entry criteria for such groups.

   The LIMS HH survey demonstrated that over 80% of households in the highest income
    bracket surveyed have one or more members in formal employment, and that for the sample
    as a whole; over 50% of households have one more members in formal employment. This
    suggests that initially limiting entry to formal and informal sector employees with proof of
    income should be more than adequate to facilitate entry of the key target markets for LIMS
    schemes in the first instance.

2.10.3 Benefit differentiation between LIMS schemes and PMB regulated sche mes

   LIMS schemes should be restricted from providing any form of cover for any private hospital
    care. This exclusion would also apply to emergency stabilisation in private hospitals.

   Emergency transport costs will be however be covered as part of the mandatory LMP.

   It is critical that this recommendation be implemented hand-in-hand with a change in the
    policy of means tested fees imposed at public hospitals, such that all users of public hospitals
    falling below the same income threshold as LIMS should be exempted from all public
    hospital fees. Therefore both LIMS members and non LIMS members earning below R6,500
    per month would not be required to pay for public hospital care.

   In the absence of this measure, LIMS schemes will be in the invidious position of having to
    pay for their members‟ usage of public hospital care and to price for this. This will increase
    the overall cost and reduce the affordability of LIMS packages. It will also mean that many
    members will be paying for services they have in effect been obtaining at no charge, due to
    the poor implementation of user fees at many public hospitals.

   The arguments in favour of the exclusion of private hospital cover from LIMS schemes are:


                                                                                                  112
     Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report




o   Perhaps the key risk in the LIMS proposals is the potential for buy downs from current
    medical schemes. If this were to happen to any significant extent, this would undermine
    the stability of the existing risk pool with severely negative long term effects. To the
    extent that policy makers perceive this as a significant risk, this will undermine the
    acceptability of the LIMS proposals.

o   While the implementation of the income threshold for LIMS schemes will provide some
    measure of protection against buy down risk from higher income scheme members, there
    is a material risk that higher income members seeking to buy down will seek to
    manipulate this rule in order to again access to LIMS schemes. For this reason, it is
    essential to create significant benefit differentiation between the current medical scheme
    environment and LIMS schemes. This is regarded as a key principle of the definition and
    regulation of LIMS schemes, and should be consistently maintained.

o   The major buy down risk for LIMS schemes are most likely to be the young and healthy
    members of current medical schemes. These individuals are likely to seek packages with
    good private hospital cover and limited day-to-day cover. LIMS schemes should
    therefore seek to avoid providing cover that will meet the needs of these individuals.

o   Exclusion of private hospital cover would therefore be a highly effective approach to
    preventing the key buy down risks.

o   The HH survey indicates a preference for comprehensive primary health care (GPs,
    specialists, dentists, optometry, and medicines) over private hospital cover amongst all
    segments of the surveyed low income population.

o   It is important to note that while the HH survey shows a preference for primary
    healthcare cover over private hospital cover, respondents still do attach a high value to
    private hospital cover per se. Under this proposal, any individual wishing to purchase
    private hospital cover could still do so through the existing medical scheme environment,
    but will not be able to do so in LIMS schemes.

o   A further advantage of this approach is that it will maintain pressure on the private
    hospital environment to reduce costs to the fullest extent possible in order to increase
    accessibility to private hospital care by members of low cost medical schemes within the
    current environment, and thereby grow volumes. By contrast, allowing LIMS members to
    obtain private hospital cover would create the risk that private hospitals would engage in
    price discrimination by seeking to maintain high prices in the current medical scheme
    environment while reducing costs for LIMS schemes.


                                                                                          113
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report




    o   There have been several consultations with the private hospital industry on this proposal.
        The industry has expressed disappointment with this proposal and the Hospital
        Association of South Africa has submitted a paper on this issue to LIMS. This is
        included as Annexure 13. Some of the key points made in that submission of relevance to
        this issue are:

        -     The decision to exclude private hospital cover from LIMS is a short sighted one, and
              ignores the fact that private hospitals could provide very cost effective care for LIMS
              members.

        -     To do so would require a number of regulatory changes, including changes to R158
              to allow for a „legislative carve-out‟ for LIMS. This would allow, among other
              things, private hospitals to develop special low cost wards for LIMS patients. These
              changes would need to be accompanied by changes to billing codes etc for such
              wards.

        -     Another key change is the need to exempt private hospitals from the provisions of the
              Basic Conditions of Employment Act regarding payment of overtime to nurses. This
              exemption, which the state enjoys, would also facilitate development of low cost
              LIMS services.

        -     Allowing hospitals to employ health professionals would also contribute significantly
              to their ability to deliver low cost hospital care.

   An alternative approach to the exclusion of private hospital cover, which might also allow
    prevention of buy-down risk to some extent, would be the imposition of a regulated
    maximum contribution rate for LIMS schemes. This would ensure that the gap between the
    minimum cost of the LMP and the regulated maximum contribution would not be so wide as
    to allow for benefits that would be attractive to those who pose the greatest buy-down risk.

   After careful consideration, it was decided that the exclusion of private hospital cover would
    be strongly preferable to the regulated maximum contribution approach. The main reason for
    this is that the regulated maximum contribution rate could lead to exploitation of members by
    schemes offering low value, and inappropriate private hospital cover (e.g. with very low
    annual limits), but which still get marketed as private hospital cover. In addition, this
    approach would be harder to administer and regulate, and would also require annual
    adjustment. A final reason is that the combination of the LMP, with no upper limit on
    contribution rate, would allow schemes to innovate in terms of the richness of out of hospital
    benefits.


                                                                                                 114
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report




2.10.4 Underwriting of entrants into LIMS schemes

   LIMS schemes will, at their discretion, be able to impose a 3 month general waiting period,
    as well as a 12 month condition specific waiting period.

   There should be no late joiner penalty for members joining LIMS schemes for an initial
    period after LIMS schemes are launched. This period is recommended to be 3 years, after
    which late joiner penalties should apply.

2.10.5 Prevention of adverse selection through buy-up from LIMS schemes to current
       market

   Once LIMS schemes are established, there will be a systemic risk that LIMS scheme
    members will be incentivised to „buy-up‟ to existing medical schemes when they have a need
    for the coverage offered by those schemes. This could therefore lead to adverse selection
    against existing schemes.

   In order to prevent this form of adverse selection, the following recommendations are
    proposed for movement between LIMS schemes and the current PMB environment:

    o   No late joiner penalties would apply provided that the member had belonged to a LIMS
        scheme for at least a minimum period. This is to avoid gaming in which members may
        join LIMS for a very short period, and then shift to the current schemes, thereby avoiding
        a late joiner penalty. The proposed period for this minimum period of LIMS membership
        would be 3 years. LIMS members who have been in LIMS schemes for at least 3 years
        would therefore qualify for a waiver of the late joiner penalty on shifting to a current
        scheme. All others would still be subject to the late joiner penalty.

    o   In addition, schemes would be entitled, at their discretion, to impose a 3 month general
        waiting period, although this would not be allowable for those benefits covered by the
        LMP.

    o   Schemes would also be entitled to impose a 12 month condition specific waiting period
        on those buying up from LIMS schemes.




                                                                                              115
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



2.10.6 Linkage between LIMS schemes and REF

   In theory, there could be a linkage between the REF and those elements of the LMP that are
    common to PMB packages, with risk equalisation across those elements. Based on the current
    proposal for the LMP, this risk equalisation would apply to the 14 CDL conditions that are
    common to the LMP and the PMBs.

   However, there are several reasons to recommend that risk equalisation is not carried out
    across the PMB and LIMS environments, but rather that these two environments be subjected
    to separate risk equalisation pools. The key reason for these recommendations is the
    substantial and material benefit differences between LIMS and PMB schemes. These include:

    o   The LMP and LIMS schemes in general, will focus exclusively on out of hospital
        benefits, while PMB schemes have historically focussed on hospital benefits.

    o   The quantum of the premiums between the 2 types of schemes will differ so substantially
        that any meaningful REF payment for the PMB environment nature will be
        disproportionate to the total LIMS premium. The current Community Rate proposed by
        the Risk Equalisation Technical Advisory Panel (RETAP) of R235 is likely to be well in
        excess of the total premium for LIMS schemes.

    o   The areas of commonality, namely the 14 CDL conditions, will account for a very small
        proportion of total risk in PMB schemes.

    o   The overall volatility of LIMS schemes is likely to be small compared to PMB schemes,
        due to the nature of benefits and the exclusion of hospital cover. By contrast, there will be
        significant systemic risk to LIMS schemes if they are tied into the REF.

   It will be important to implement a separate REF pool for LIMS schemes as soon as feasible,
    in order to avoid systemic adverse selection. Initial discussions with RETAP have led to the
    following recommendations:

    o   A separate REF system should be established for LIMS schemes as soon as possible, and
        by no later than 2 years after launch of the LIMS schemes.

    o   Given the relatively low volatility expected in these schemes, it should be possible to risk
        equalise on the basis of relatively simple parameters such as age, gender and chronic/non
        chronic status only.




                                                                                                 116
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    o   The LIMS REF system should be governed and operated on the same principles and
        within the same infrastructure as the main REF.



2.11    GOVERNANCE ISSUES IN LIMS SCHEMES

The Demarcation and Governance Sub-Group has worked on a number of issues relating to the
governance of LIMS schemes, or LIMS options within existing schemes. In particular, the group
focused on the following specific questions and issues:

   Should the LIMS environment be open to new medical schemes only or to new options
    within existing schemes as well?

   How can members of LIMS schemes be given appropriate „voice‟ and influence in the
    governance of LIMS schemes or LIMS options within existing schemes?

   Reserves and Solvency Requirements for LIMS schemes and LIMS options within existing
    schemes.

The following proposals have emerged from the work of this sub-group:

2.11.1 LIMS schemes and LIMS options within current schemes

   The Sub-Group discussed in detail the question of whether the proposed new LIMS schemes
    should be separately established medical schemes only, or whether existing medical schemes
    should also be entitled to register new LIMS options within their schemes.

   The clear consensus on this issue is that both current schemes (through the registration of new
    LIMS options) and new LIMS schemes should be encouraged to participate in the LIMS
    environment. The following are the main arguments in favour of this unanimous view within
    the Sub-Group:

    o   This approach would maximise the number and diversity of market participants in the
        LIMS market, which will benefit consumers through competition and diversity of product
        offerings.

    o   Options within existing schemes will be able to draw on the reserves of host schemes,
        either temporarily or permanently. This will significantly reduce initial capital and
        solvency requirements, thus reducing the costs of these options.




                                                                                               117
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    o   The use of options within schemes will facilitate economies of scale, and thus cost
        reductions in administration and other aspects of scheme governance.

    o   Closed schemes may wish to register a LIMS option, and if this were not permitted,
        employers would be forced to register a new LIMS scheme for their eligible employees
        and dependents.

    o   Employers in open schemes are likely to prefer a LIMS option within their current
        scheme rather than being forced to split their workforce between a LIMS scheme/s and
        their current schemes.

    o   Allowing options within existing schemes, as well as new schemes, will limit the
        perception of LIMS schemes as inferior and for the poor, by encouraging maximal
        integration with existing schemes.

    o   New schemes have a higher failure risk that the establishment of options within existing
        schemes.

This strong consensus view was presented to the Policy Workshop on 19 August 2005. The view
from the policy makers was that there is no reason, in principle, to oppose the registration of both
new options and new schemes within a LIMS environment. However, once again several
important issues were raised in this context, including:

   Mechanisms to prevent movement between LIMS options and mainstream options within the
    same scheme. This issue is effectively addressed through the significant benefit
    differentiation between current schemes and LIMS schemes proposed above.

   New governance structures for LIMS options (for e.g. a sub-committee of the Board of
    Trustees) may need to be defined, but care should be taken to ensure that they do not require
    so much regulatory oversight that they become, in essence, new schemes. Also, the powers of
    such a body need to be carefully defined. These issues are addressed further below.

   Host schemes would present the advantage of being able to use reserves to provide a form of
    „internal re-insurance‟ to LIMS options while they build up reserves. Technical work on this
    aspect will be required.

   Issues relating to the minimum size of a LIMS option within an existing scheme will need to
    be discussed and defined.




                                                                                                118
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



It is critical to note that these discussions and recommendations emerged prior to the publication
during February 2006 of Circular 8 from the Council for Medical Schemes. This Discussion
Document proposes an entirely new benefit structure for medical schemes, with a common set of
benefits that would have to be offered across the entire scheme at a uniform price, and with
schemes able to offer supplementary benefit packages above the common set of benefits. The
proposed new benefit structure does complicate the discussion of whether LIMS options could be
registered within existing schemes, or should be registered as separate schemes.

This debate remains at an early stage, and there remains much detail to clarify on how this
significant change to the current medical schemes environment would be implemented in
practice. On the face of it, however, the following comments are relevant in relation to the
discussion regarding LIMS schemes and LIMS options:

   New LIMS schemes would be compatible with these proposals. In this case, the common set
    of benefits across the entire LIMS scheme would comprise at least the LMP, as well as any
    other benefits the scheme wished to add, excluding hospital benefits. LIMS schemes would
    then be able to offer supplementary packages above the common pool of benefits, subject to
    the fact that these too would not be able to provide cover for hospital benefits.

   The development of LIMS options within current schemes becomes more complex in the
    context of the proposed new benefit structure, and the following considerations should be
    noted:

    o   In order to satisfy both these new structural requirements and the requirements of LIMS
        options, there would need to be two separate common benefit pools, one for the main
        body of the scheme, and one for the LIMS option. This would be essential since the
        minimum requirements of current schemes and LIMS schemes are completely different,
        so that a single common benefit pool would not be possible.

    o   The LIMS common benefit pool would need to comply with the agreed provisions of the
        proposed new structure in relation to pricing etc.

    o   Once that separate common benefit pool is in place, the scheme could also offer
        supplementary benefit packages for LIMS members.

    o   Both the LIMS common benefit pool, and any LIMS supplementary packages would
        have to be subject to the legislation and regulations applying to LIMS schemes, including
        all of the critical issues outlined above, such as the income based entry criteria, exclusion
        of hospital benefits etc.




                                                                                                 119
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    o   This more complex approach, with more than one common risk pool in a single scheme,
        might increase the potential for gaming and manipulation by schemes and/or by
        members. However, the substantial benefit differentiation between current schemes and
        LIMS options should significantly mitigate this risk.

   While these issues clearly require further discussion, the proposed new benefit structures do
    not seem to pose material obstacles to the consensus view that current medical scheme should
    be able to register and develop LIMS options within the existing scheme.

2.11.2 Ensuring adequate ‘voice’ and influence for LIMS members in the governance of
       LIMS schemes or LIMS options within existing schemes

The Sub-Group also examined the question of how low income members can be assured of
appropriate and effective „voice‟ in the governance of LIMS schemes. The following concrete
proposals have emerged from these discussions:

   Where a new LIMS scheme is established, this will have a Board of Trustees (BOT) as
    required by the Medical Schemes Act.

   Where a LIMS option is established within an existing scheme, there should be requirement
    for a specific LIMS sub-committee of the BOT (LC), to provide effective voice to members
    of the LIMS option. Note that all comments below pertaining to the LC would also apply to
    the BOT of LIMS specific schemes.

   The constitution of the LC should follow good governance requirements as often applied in
    the case of audit committees of Boards of directors; i.e. the majority of members should be
    drawn from outside of the main BOT of the scheme and the chairperson may not be a
    member of the BOT. In addition, membership of the LC should ensure adequate
    representation of membership and of the necessary technical skills, including legal, provider
    contracting & managed care, medical and actuarial/statistical skills.

   The election mechanisms and the maximum permissible terms of office should be as for
    BOTs.




                                                                                             120
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



2.11.3 Reserves and Solvency Requirements for LIMS schemes and LIMS options

Some initial work has been done on these issues, and the following proposals are advanced for
further discussion:

   In the case of new LIMS schemes, the current provisions of the Medical Schemes Act will be
    adequate to deal with both initial capital requirements and ongoing solvency requirements.

   It could however be argued that the overall volatility of claims in LIMS schemes will be far
    lower than current schemes, and that this suggests that the minimum overall solvency level
    required should be set lower than the current level of 25%.

   Even if the same solvency level is ultimately required, it would be appropriate to allow LIMS
    schemes an extended period, say 5 years, in which to build up to this level, subject to meeting
    a minimum required solvency level (say 5%-10%) within the first 12 months. This gradualist
    approach would assist in the initiation and development of these schemes.

   In the case of LIMS options within existing schemes, the following comments apply:

    o   Schemes applying to register a new LIMS option would obviously have to satisfy the
        Office of the Registrar that current solvency level of the scheme are adequate.

    o   The current reserves of the scheme could be used to provide solvency protection for new
        LIMS options, and/or could be used to provide an internal „re-insurance‟ to the LIMS
        option while it builds up its reserves.

    o   The LIMS options would need to report separately on their reserves.

    o   Should the scheme‟s original reserves have been utilised to fund losses on the LIMS
        options during the phasing in period, then the LIMS option should be required to pay this
        back into the scheme‟s main reserves by the end of the phasing in period.

    o   An alternative to this would be to allow a permanent situation in which the LIMS option
        utilises the reserves of the host scheme for solvency purposes. In this case, it will be
        essential to ensure that this effective „subsidy‟ flows only from the host scheme to the
        LIMS option, and not in the other direction. This underscores the importance of separate
        reporting of the LIMS option reserves.




                                                                                               121
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



2.12    CONCLUSIONS

   The LIMS process has generated a significant body of important information on which to
    base sound recommendations for policies that can extend health insurance cover to those low
    income households who currently have only partial cover or no cover at all. This information
    is derived from significant new research commissioned for the LIMS process, as well as
    through analysis of existing research and experience. It has also emerged from the collation
    of the views of numerous industry experts and participants, and through extensive discussions
    on all key issues in the large number of Task Groups, Sub-Groups and Technical Task
    Groups that have been working on various aspects of the LIMS process.

   The conclusions summarised in this section are therefore well grounded in the empirical
    evidence to hand. They are also, in the main, the product of consensus developed after
    extensive discussion and debate by large numbers of industry participants. Where consensus
    could not be reached, this has been identified, as requested by the Terms of Reference of the
    LIMS process.

   The key conclusions emerging from the findings can be summarised as follows:

   Medical scheme coverage levels in low income households are currently very low. In the
    „target market‟ identified at the outset of the LIMS process, i.e. households earning between
    R2,500 and R6,000 per month, 87% of all individuals (9.95 million) are without medical
    scheme cover.

   Unemployment is clearly a major obstacle to extending cover. Approximately 49% of all
    households with either partial cover or no cover at all indicated that no household member
    was formally employed at the time of the survey.

   However, unemployment is not the only obstacle to extending cover. 51% of households with
    partial or no cover reported one or more household members in formal employment. In these
    households, the main obstacle to cover is the lack of affordability of current premiums. In
    some cases, households also cited employers not offering subsidies for medical scheme cover
    as a reason for lack of coverage in the household.

   Despite the low coverage statistics, a significant majority of low income households appear to
    want access to medical scheme cover. Over 60% of respondents in the HH survey showed a
    strong interest in purchasing medical scheme cover in principle, with very few objecting at a
    conceptual level to the notion of risk pooling and pre-payment for medical benefits. This adds
    weight to the observation that financial constraints, and lack of affordability, are the key




                                                                                              122
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    obstacles to extending coverage at present, at least for those households with access to formal
    sector employment.

   The impact of high premium costs on severely budget constrained low income households is
    also seen in the specific coverage patterns identified in the survey. These households are
    choosing to cover mainly children and working age adults, but not to cover young adults and
    the elderly. There is also some evidence of adverse selection, with households choosing to
    provide cover for the sick, who need access to health care more than healthy household
    members. Both of these patterns are rational for households which desire to provide cover,
    but have to ration this due to budget constraints and high premium costs.

   Despite lack of access to cover, many low income households utilise private healthcare
    services, and incur high OOP expenditures as a result. The survey indicated that 12% of the
    households in the target market used a private GP at least once in the previous 3 months, and
    that expenditure on GP services alone averaged R105 per month, equivalent to just below 3%
    of household income. For the lowest income category, monthly expenditure on GP services
    accounted for just below 10% of household income. These data exclude the small but relevant
    OOP expenditures on public clinics as well. The survey also demonstrated evidence of a high
    level of unmet need for healthcare amongst low income households.

   This pattern of high OOP expenditures and high levels of unmet need amongst uninsured low
    income households is typical of many countries with incomplete health insurance cover, as
    pointed out in the literature review. This pattern suggests that there is an unsatisfied demand
    for high quality healthcare amongst a substantial proportion of the population. This demand is
    not currently being met due to the inability of government to mobilize sufficient financial
    resources through some combination of general taxation and/or social health insurance.

   As also highlighted in the literature review, this situation is inequitable since OOP
    expenditure is the most regressive form of healthcare financing. It also means that those
    without cover cannot benefit from tax subsidies, employer subsidies, and medical schemes‟
    collective purchasing power and provider arrangements etc. The DOH has recognised this
    problem and has developed the SHI policy framework to address it. However, the SHI
    framework is taking longer to implement than originally envisaged.

   It is also very likely that these high OOP expenditures and unmet need for healthcare explain,
    at least to some extent, the high levels of interest amongst low income households in
    obtaining access to medical scheme cover. Households are clearly aware of the financial
    impact of their OOP healthcare expenditure, and want to obtain the benefits of risk pooling
    for these expenditures.




                                                                                               123
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   These various observations suggest that finding some mechanism to expand private health
    insurance to more low income households in South Africa would be highly beneficial.
    Expansion of coverage would increase risk pooling, assist in overcoming the current income
    related inequalities in access to healthcare, and reduce the negative financial impact of OOP
    expenditures on these households.

   Relative to many other countries, South Africa is also ideally placed to use an expansion of
    private health insurance as a mechanism to address these problems. It already has a well
    organised and regulated existing private health insurance system, and a well developed
    government plan to develop an SHI system over time, but faces obstacles to extending this in
    the short to medium term. Expanding medical scheme coverage to as many low income
    households as quickly as possible would thus be an ideal way to bridge the gap betwee n the
    current situation and the later development of a fully fledged SHI framework.

   The use of PHI systems to expand voluntary health insurance to low income populations is
    being strongly supported by the Word Health Organisation and other health policy experts as
    an important intermediate step towards achieving full coverage.

   However, as pointed out above, the fundamental obstacle to expanding coverage to low
    income households in South Africa remains affordability. Any solution to these problems will
    have to significantly reduce the net cost of medical scheme premiums to low income
    households.

   As also suggested in the literature review, demand for health insurance cover will be
    maximized if the level of premiums are similar to households‟ expected OOP expenditures on
    healthcare services. The evidence on current household expenditure by low income
    households suggests that current OOP expenditures are in the region of 3%-10% of HH
    income, depending on the income category examined. The Stats SA surveys indicate that
    current low income members of medical schemes spend approximately 7%-9% of HH
    income on medical scheme contributions. These data indicate that current expenditure and
    affordability levels are well below even the lowest of current medical scheme premium
    levels. Premiums will therefore have to be substantially reduced from current levels if they
    are to lead to a material expansion in medical scheme coverage.

   Estimates of the potential demand for medical scheme cover amongst low income households
    using the data obtained from the HH survey allow us to quantify the likely impact of various
    reduced premium levels. Using an assumption that households will allocate 5% of income to
    medical scheme cover, and assuming a 50% employer subsidy, the model suggests that a
    premium of R200 per life per month would bring an additional 1.55 million lives into cover.
    If the premium could be further reduced to R150 per life per month, estimated new lives


                                                                                             124
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    under cover increases to 1.85 million. If we assume that household affordability is at 8% of
    income rather than 5%, a R200 premium would lead to 1.85 million new lives under cover,
    while the R150 premium would now have a dramatic impact, leading to a projected 3.27
    million new lives under cover.

   The LIMS process has examined three broad sets of interventions that could be used to
    materially reduce the net medical scheme premium costs to low income households. These
    are:

    o   Direct subsidies, either from employers, or the State, or both.
    o   Changes to the scope of benefits offered by medical schemes.
    o   Reductions in the costs of healthcare goods and services.

   Employer subsidies: International and local experience indicates that direct subsidies that
    have the effect of reducing the net premium payable by low income households will be
    critical to the objective of extending coverage to these households. These subsidies typically
    come from employers, as has historically been the case in South Africa. Results from the
    demand model show that the absence of employer subsidies (assumed to be 50% of gross
    premium cost) will materially reduce the positive impact of lower premiums on the extension
    of coverage. Using the R200 per life premium as an example, and assuming the 8%
    affordability level, the absence of the 50% employer subsidy would reduce the projected new
    uptake from 1.85 million by more than half, to 678,000. Similar effects are seen using the
    other assumptions. Fortunately, both the HH survey and the employer surveys indicate that
    substantial numbers of employers in South Africa do in fact offer subsidies to the low income
    employees. Hundreds of thousands of their employees do not take advantage of these
    subsidies; however, do to the high net premium cost to their households, even after the
    subsidies. Thus, materially reducing the gross premium cost would have the impact of
    leveraging the already substantial employer premium on offer to many low income
    households.

   Treasury Subsidies: This report proposes a new, direct subsidy from the NT to members of
    proposed new LIMS schemes or options. Demand modeling suggests that, if set at the right
    levels and if targeted appropriately, such subsidies would have a significant impact on
    extending coverage. Assuming a R200 per life premium and the 8% affordability level, a
    direct subsidy of R25 per life per month would increase the new lives under cover from the
    estimated 1.85 million without the subsidy, to 3.17 million. Similar impacts are seen using
    different assumptions as to the NT subsidy and premium levels, although the relationship
    between the scale of the subsidy and its impact is highly dependent on the actual gross
    premium level assumed, as well as on the assumed household affordability level. More
    detailed analysis of the HH survey and other data will be required in order to calibrate the


                                                                                              125
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    most effective level of subsidy. The cost to the fiscus of subsidies of this level and scale are
    projected to be relatively modest – in the region of R2 billion (assuming a R200 premium,
    R50 per life NT subsidy, 8% affordability level). This is similar to the amounts currently
    budgeted for increased take up by low income families as a result of the recently legislated
    tax subsidy reforms. These data therefore suggest that direct premium subsidies from NT
    would represent very cost effective allocation of scare fiscal resources.

   A direct premium subsidy will be far more effective than other tax based instruments, such as
    a tax deduction on contributions as used in the current market. These latter instruments have
    limited impact where tax liabilities are low due to low income and low marginal rates, and
    where the fundamental barrier is the high cost of the premium itself. International experience
    suggests that a direct subsidy will be far more effective if targeted at individual scheme
    members, rather than at employers. This is also more consistent with employment patterns in
    South Africa, where only half or less of potential members of low income schemes are
    formally employed.

   Reducing the costs of healthcare goods and services: In addition to premium subsidies, the
    affordability of medial scheme cover is fundamentally dependent on reducing the cost of
    healthcare goods and services, and the ongoing high rates of inflation in these costs. The
    LIMS process has investigated the prospects for reduction in cost and cost inflation in the
    short, medium and long term. The key conclusions of these investigations can be summarised
    as follows:

    o   Costs of health professional services: There appears to be relatively little scope for any
        systematic reductions in the costs of health professionals‟ services in the short term. This
        is due to a number of factors, including relative shortages of key categories such as
        medical specialists and dentists, as well as the fact that many individual practitioners
        already provide differential pricing on an informal basis to different patients, and do not
        appear willing or able to enter into formalized arrangements of this kind, principally due
        to competition concerns. At the primary care level, GPs are already organized into many
        different networks, which do offer lower cost primary care access, and there seems little
        room to reduce costs below those currently in the market.

    o   Development of integrated delivery systems: A key conclusion of this report is that, in the
        medium to long term, there could be substantial cost reductions achieved through
        reorganization of health care delivery into various forms of integrated delivery systems.
        Such systems would involve optimal use of different levels of health professionals,
        usually collaborating in teams. They would also likely involve corporate entities being
        able to employ and/or share risk with health professionals in order to adequately align
        incentives for cost effective care. However, the current regulatory environment poses a


                                                                                                126
     Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    major obstacle to the emergence of these integrated systems. The key regulatory hurdles
    to these developments arise from the ethical rules and policies of the Health Professions
    Council of South Africa, as well as from other aspects of prevailing legislation, including
    those that define the scopes of practice of various health professions, and aspects of the
    National Health Act pertaining to Certificates of Need. This report highlights the need for
    an urgent review of the various HPCSA rules and policies, in order to preserve the good
    intentions of these, while addressing their many, serious unintended consequences.

o   Hospital Costs: While hospital costs were examined in the LIMS process, they are not
    reported on in detail here due to the proposed exclusion of hospital cover.

o   Costs of medicines: The LIMS process elicited a productive and creative debate on
    various approaches to reducing the costs of medicines for the LIMS environment. Two
    structural alternatives were debated: the use of an Access Subsidy system in which LIMS
    members would obtain medicines via normal distribution channels at the SEP price, but
    LIMS schemes would obtain a discount to the SEP from manufacturers; and the use of
    State Tender medicines for LIMS members. As between these two broad alternatives,
    there was a clear consensus that the Access Subsidy system is preferable for several
    reasons. These relate principally to the logistics of distribution of the medicines and
    related issues. This leaves open the question of how to appropriately price the discounts
    on the SEP for medicines in the LIMS environment. Three broad alternatives for pricing
    were debated during the LIMS process. Some of the major organisations of the
    pharmaceutical industry have proposed that the Access Subsidy is developed on a
    voluntary basis for specified time periods; another alternative would be to have a
    standardized discount to the SEP for all medicines sold to LIMS patients; a third
    alternative would be to use State Tender pricing as the basis for the access subsidy. No
    consensus was developed on which of these pricing systems to adopt, and further debate
    and discussion will be required before this matter can be finalized.

o   Costs of devices, surgical items, disposables etc: Limited discussion was held on this
    issue. Some useful suggestions emerged for addressing the high cost of many of these
    items, including the development of an SEP approach, and/or the use of a reference
    pricing approach in LIMS schemes.

o   Costs of medical scheme administration and managed care: There are several reasons to
    anticipate that the costs of these services will be dramatically reduced in the LIMS
    environment. These include the fact that this is a highly competitive market, and the fact
    that these fees are closely monitored by the regulator, and are closely calibrated to the
    quantum of premiums. To the extent that premiums are materially reduced, this will of
    necessity lead to proportionate reductions in administration and related fees. As the report


                                                                                            127
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



        indicates, these reductions will necessitate significant changes in the business model of
        administration companies.

   Scope of benefits covered by medical schemes: The affordability of cover is also
    fundamentally dependent on the scope of benefits offered by medical schemes. The
    international literature shows that, where minimum benefit packages are regulated, these tend
    to protect equity and comprehensiveness of cover, but at the cost of increasing the minimum
    cost of cover, thereby reducing affordability. Many governments have faced this specific
    dilemma, and some have addressed this by reducing the scope of prescribed minimum
    benefits in order to extend cover to low income populations.

   This report has argued, based on empirical evidence and on a firm consensus amongst a wide
    group of stakeholders, that the high and increasing cost of the current PMB approach is a
    fundamental driver of the high cost of existing medical schemes. This impacts most severely
    on the affordability and hence the uptake of medical schemes aimed at low income
    households, which have shown only limited penetration, despite several years of intense
    marketing efforts. Reducing the scope of prescribed minimum benefits in medical schemes
    will therefore be a critical component of expanding health insurance cover to low income
    households.

   One of the guiding principles of the LIMS process has been that any recommendations made
    would need to be supportive of the current medical schemes environment, and in particular,
    the hard won PMB provisions, which apply to current medical schemes in terms of the
    Medical Schemes Act. In order to fulfill this requirement, while also reducing the scope of
    prescribed minimum benefits in schemes aimed at low income households, the LIMS process
    had led to the development of a number of concrete recommendations regarding the
    establishment of a new class of LIMS schemes or LIMS options within existing medical
    schemes.

   The key principles and recommendations regarding this new class of LIMS schemes or LIMS
    options include:

    o   They must be effectively demarcated from current medical schemes.

    o   Such demarcation should include the use of a defined income threshold, with
        membership open only to those earning below that threshold, and their dependents. This
        income threshold should be set at a level which would maximally encourage new entrants
        from low income households, while at the same time limiting the risk of „buy down‟ by
        current medical scheme members.




                                                                                             128
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    o   As an income threshold on its own will not create sufficient demarcation due to the
        potential for gaming, demarcation should be further enhanced through material benefit
        differentiation between LIMS and current schemes, and this differentiation should be
        regulated. The key element of the proposals in this regard is that LIMS schemes should
        not provide any cover for hospital care. This will almost completely mitigate the risk of
        buy downs from the current schemes environment, since most current members are likely
        to seek private hospital cover and will therefore remain in the current system. This
        exclusion is also compatible with the preferences identified in the HH survey.

    o   Underwriting and late joiner provisions should be fully compatible with the current
        provisions of the Medical Schemes Act, while at the same time avoiding adverse
        selection against LIMS schemes, as well as against current schemes through buy up by
        LIMS members.

   These principles have been captured in detailed and specific recommendations for the
    establishment and demarcation of LIMS schemes or options, including a recommended
    income threshold, principles for administration of this threshold, and rules to prevent adverse
    selection against LIMS schemes and current schemes.

   Perhaps the key feature of the LIMS recommendations concerns the nature of the revised
    PMBs for the new class of LIMS schemes or options – this has been defined as the LIMS
    minimum package (LMP). The following key conclusions were reached in relation to the
    LMP:

    o   An LMP for LIMS schemes will be essential, in order to limit anti-selection between
        LIMS schemes, to provide consumer protection against unscrupulous schemes, to
        facilitate the use of a risk equalisation mechanism within LIMS schemes, and to provide
        a coherent basis for motivating for direct National Treasury subsidies to LIMS.

    o   The LMP is complementary to the current PMB approach, since members of LIMS
        schemes will still enjoy the protection of the PMBs, with the balance of services not
        funded by LIMS schemes being funded and provided by the public sector. This also
        mitigates against the concern that an LMP would have material negative equity impact.

    o   The LMP must cost substantially less than the current PMBs within medical schemes, in
        order to ensure affordability of LIMS schemes. Without this fundamental plank of the
        LIMS approach, there is no justification for separate LIMS schemes.




                                                                                               129
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    o   The LMP package must be consistent with expressed preferences of potential LIMS
        members as demonstrated by the LIMS HH survey, and should ideally be consistent with
        Department of Health‟s view on the essential care package.

    o   For this reason, the approach to defining the LMP differs from the current PMBs. Rather
        than a revised list of diagnosis and treatment pairs, the LMP is based on a defined
        package of services as a minimum benefit. This package of services is designed to be
        consistent with the structure of health service delivery and patient experience.

    o   The HH survey indicated a strong relative preference for comprehensive out of hospital
        day-to-day benefits rather than cover for private hospital care. For this reason, the
        proposed LMP is based on a comprehensive primary healthcare package of services.

    o   Detailed initial cost estimates have been made on the proposed LMP. These suggest that
        this package could be provided for approximately R108 per life per month. The actual
        costs of these schemes will obviously vary, due to choices made by schemes in relation to
        benefits above the LMP, as well as to the underlying costs of services, administration and
        distribution services. However, these estimates do suggest that once employer subsidies
        are taken into account, premiums for LIMS schemes are likely to be substantially lower
        than even the lowest of current scheme premiums. These estimates are also well within
        the range of the premium parameters used in the demand estimation models, and should
        therefore result in substantial increases in take up of medical scheme coverage.

   Governance of LIMS schemes: The key debate regarding governance was whether existing
    medical schemes should be able to register LIMS options within the scheme, or whether th is
    should be prevented, thus requiring that new LIMS schemes be established. The firm
    consensus on this issue was that both new LIMS schemes and LIMS options within schemes
    should be permitted. This debate has been complicated by the recent publication of Circular 8
    of 2006 by the Office of the Registrar of Medical Schemes, which proposes substantial
    changes in overall benefit structures of medical schemes. These changes have not been
    discussed within the LIMS process. Initial analysis does however suggest that the
    recommendation of allowing LIMS options within existing schemes would in fact be
    compatible with the proposed new approach to benefit structure.

   Appropriate influence of LIMS members within existing schemes: To the extent that schemes
    are allowed to register LIMS options, it will be important to ensure that LIMS members have
    adequate influence over the affairs of their option. In order to address this, it is recommended
    that schemes with LIMS options establish a sub-committee of the Board to govern the LIMS
    option and represent that option on the Board of Trustees.




                                                                                                130
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   Solvency of LIMS schemes and options: Discussions on solvency requirements for LIMS
    schemes reached the following main conclusions:

    o   LIMS schemes and options will have much lower claims volatility than current schemes
        due to exclusion of hospital benefits, and the greater predictability of claims value in day
        to day benefits. For this reason, consideration should be given to much lower overall
        solvency requirements than the current 25%.

    o   New LIMS schemes should be given an extended period, say 5 years, in which to build
        up to required solvency levels.

    o   Options within existing schemes could draw on the host scheme reserves for solvency
        requirements in various ways, including forms of internal reinsurance, but should report
        separately on solvency.

   Distribution of LIMS schemes: The LIMS process also examined how best to distribute LIMS
    schemes or options. The following key conclusions emerged from this work:

    o   Brokers and employers appear to be the most suitable channels for ongoing and expanded
        distribution of medical scheme products to low income households. At a later stage,
        community based organisations and associations (e.g. stokvels, burial societies, NGOs or
        other civic groupings) might have a role to play, but this will be premature at this stage,
        given the recommendations regarding administration of the income threshold, as well as
        the focus on employment linked distribution of LIMS schemes.

    o   Financial modeling and discussions with brokers suggests that distribution costs could be
        contained within the current regulated maximum of 3% of contributions, even at the
        much lower contribution rates estimated for LIMS schemes.

    o   Employers will play a vital role in the distribution of LIMS schemes to their employees
        and dependents. Employers‟ active support for the broker and/or trade unions can make a
        significant impact on the success of otherwise of efforts to extend coverage.

    o   The employer surveys, which LIMS has drawn on, suggest that many employers do in
        fact offer both subsidies for medical schemes, as well as advice and support to their
        employees in their choice of schemes. However, there is also evidence that many
        employers do not offer sufficient advice and support to low income workers investigating
        scheme options, and also that large numbers of particularly small and medium sized
        employers do not offer subsidies at all. Successful expansion of coverage to millions of
        additional low income workers will be fundamentally dependent on active support from


                                                                                                131
     Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    employers throughout the economy. Organised business institutions are excellently
    placed to play a very constructive role in this regard, and should be encouraged to do so.

o   Successful distribution of low income medical scheme products will therefore require a
    careful re-look at the roles, division of labour, and relative costs borne by brokers,
    scheme administrators and employers.




                                                                                          132
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



3.        RECOMMENDATIONS

This section provides the full set of recommendations emerging from the LIMS process. As the
full background, arguments and explanations for each recommendation have been discussed in
detail above, these are not repeated here.

3.1       ESTABLISHMENT OF A NEW CLASS OF MEDICAL SCHEME FOR
          LOW INCOME MEMBERS

      a) The Medical Schemes Act should be modified to allow for the establishment of a new
         class of medical scheme (“LIMS schemes”).

      b) These modifications should allow for the registration of new LIMS schemes, as well as
         for the registration by existing schemes of new LIMS options. This specific
         recommendation requires further discussion in the context of the proposed new benefit
         structure now under discussion by the Regulator and industry stakeholders.

3.2       INCOME THRESHOLD FOR MEMBERSHIP OF LIMS SCHEMES

      a) Membership of LIMS schemes should be limited to principal members who fall below a
         defined level of personal income level, and their beneficiaries.

      b) The proposed LIMS threshold is a personal income of R6,500 per month in 2005 Rands.
         This should be inflated by CPIX between July 2005 and a date approximately 4-6 months
         prior to the planned date of implementation of LIMS schemes. The income threshold
         should be inflated annually by CPIX after the initial level is set, and should be reviewed
         every 3 years by the Council for Medical Schemes, in order to ensure that it is still
         appropriate and in keeping with the principal of avoiding buy down risk from current
         schemes, while encouraging maximal participation from low income households without
         coverage.

3.3       ADMINISTRATION OF THE INCOME THRESHOLD

      a) Membership of LIMS schemes will be open to formal sector employees, as well as self
         employed and informally employed individuals, and their beneficiaries.

      b) At a later stage, it may be possible to open up membership to defined associations or
         groups (e.g. taxi associations, burial societies etc), where the overall profile of the group
         could serve as sufficient proxy for income of members. This is considered premature at
         this stage.


                                                                                                  133
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report




    c) The income of the applying principal member will have to be demonstrated to the scheme
       administrator, either on the basis of official payroll data supplied by the employer, or by
       furnishing a SARS tax return. Applications will not be accepted in the absence of one of
       these forms of proof of income. However, no other information will be required as proof
       of income.

3.4 ESTABLISHMENT OF A NEW PRESCRIBED MINIMUM BENEFIT FOR
    LIMS SCHEMES

    a) The Medical Schemes Act should be modified to require that LIMS schemes and LIMS
       options offer a defined minimum benefit package, the LIMS minimum package (LMP).

    b) The proposed LMP is as follows:

A. GP consultations:

A minimum of:
 3 GP visits for M + 0
 6 GP visits for M + 1
 8 GP visits for M + 2
 10 GP visits for M + 3
 12 GP visits for M + 4
 A Max of 12 GP visits per family per annum
 A formulary comprising a limited set of procedures to be performed in GPs rooms
 An additional minimum of 3 GP visits per annum per beneficiary who has one or more LIMS
   PMB conditions
 It is assumed that most LIMS options will contract with DSP GP networks, and are likely to
   offer unlimited GP consultation benefits. However, this
 GP networks may utilise nurses and other service providers, but the package must provide
   access to GPs where this is required

B. Pathology and radiology investigations ordered by GP, subject to a defined formulary

C. Dental consultations:

A minimum of:
 2 dental visits per beneficiary per annum for basic conservative and restorative dentistry
 No cover mandated for advanced dentistry or dentures, with some upper limit on number of
   visits as for GPs


                                                                                              134
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report




D. Optometry:

A minimum of:
 One eye test per 24 months
 One pair of spectacles every 24 months, subject to clinical criteria and a formulary
 LIMS schemes may elect to impose reasonable financial limits, as well as protocols related to
   lens prescriptions
 A basic frame should however be covered in full

E. Acute Medicines

   A minimum of a defined formulary for acute and acute intermittent medications, based on the
    DoH Essential Drug List, with suitable modification.
   Acute medications on the formulary will be allowed for a maximum of one month
   Acute intermittent medications will be allowed for a maximum of 6 months. Examples of
    acute intermittent conditions to be covered by the formulary for 6 months include:
    o depressive disorders
    o allergic rhinitis
    o pain control of shingles
    o osteoarthritis
    o anaemia
    o pyridoxine deficiency
    o rheumatic fever prophylaxis
    o TB therapy

F. Chronic Medication

   A minimum of a defined formulary, based on clinical protocols for the following 14 LMP
    CDL conditions:
    o Asthma
    o Bronchiectasis
    o Cardiac failure/cardiomyopathy
    o Chronic renal disease
    o COPD
    o Coronary artery disease
    o Diabetes mellitus type 1 and 2
    o Dysrhythmias
    o Epilepsy
    o HIV



                                                                                            135
          Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



    o   Hyperlipidaemia
    o   Hypertension
    o   Parkinson‟s disease
    o   Rheumatoid arthritis

   A proposed formulary for discussion is attached as Annexure 12 for acute, acute intermittent
    and the LMP CDL chronic conditions.

G. Emergency transport to a public hospital (or private hospital in cases of life threatening
emergency)

H. Maternity Care Services:

   Not mandated as part of minimum package

I: Specialist Benefits

   Not mandated as part of minimum package

The full details of the proposed LMP will need to be refined and finalised following further
analysis and consultation. This is particularly important with respect to the defined formularies
for medication, radiology and pathology. The formularies included in this Report should be
considered as preliminary suggestions.

3.5 Benefit differentiation between LIMS schemes and current PMB regulated
    schemes

    a) LIMS schemes should be restricted from providing any form of cover for any private
       hospital care. This exclusion would also apply to emergency stabilisation in private
       hospitals.

    b) Emergency transport costs will be however be covered as part of the mandatory LMP.

    c) It is critical that this recommendation be implemented hand-in-hand with a change in the
       policy of means tested fees imposed at public hospitals, such that all users of public
       hospitals falling below the LIMS income threshold should be exempted from all public
       hospital fees. Therefore both LIMS members and non LIMS members earning below the
       defined LIMS income threshold at any point in time would not be required to pay for
       public hospital care.




                                                                                             136
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



3.6 Unde rwriting of entrants into LIMS schemes

   a) LIMS schemes will, at their discretion, be able to impose a 3 month general waiting
      period, as well as a 12 month condition specific waiting period.

   b) There should be no late joiner penalty for members joining LIMS schemes for an initial
      period after LIMS schemes are launched. This period is recommended to be 3 years, after
      which late joiner penalties should apply.

3.7 Prevention of adverse selection through buy-up from LIMS schemes to current
    market

   a) Where LIMS scheme members wish later to buy up to PMB regulated schemes, no late
      joiner penalties should apply provided that the member had belonged to a LIMS scheme
      for at least a minimum period. The proposed period for this minimum period of LIMS
      membership would be 3 years. LIMS members who have been in LIMS schemes for at
      least 3 years would therefore qualify for a waiver of the late joiner penalty on shifting to
      a current scheme. All others would still be subject to the late joiner penalty.

   b) In addition, schemes should be entitled, at their discretion, to impose a 3 month general
      waiting period on LIMS members wishing to buy up to PMB regulated schemes. These
      waiting periods would not be applicable to those benefits already covered by the LMP.

   c) Schemes would also be entitled to impose a 12 month condition specific waiting period
      on those buying up from LIMS schemes. Once again this should not apply to benefits
      covered by the LMP.

3.8 Linkage between LIMS schemes and REF

   a) A separate REF system should be established for LIMS schemes as soon as possible, and
      by no later than 2 years after launch of the LIMS schemes.

   b) Risk equalisation parameters should be as simple as possible. Initial recommendations for
      this would be the use of age, gender and chronic/non chronic status only.

   c) The LIMS REF system should be governed and operated on the same principles within
      the same infrastructure as the main REF.




                                                                                              137
         Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



3.9 Reducing the costs of healthcare goods and services

   a) No short term policy or regulatory interventions are recommended in relation to the role
      and costs of services provided by healthcare professionals, administrators or brokers and
      consultants involved in the distribution of medical scheme products. In some of these
      sub-sectors (administration, managed care, distribution, and some health professionals –
      e.g. optometrists), market dynamics will ensure significant cost reductions in response to
      the emergence of LIMS schemes. In others (GPs, Specialists, Dentists, Nurses), there are
      either significant shortfalls in supply and/or existing alternative arrangements that can
      deliver lower costs of care. Regulatory intervention will not assist and will likely
      aggravate the achievement of lower cost care in these cases.

   b) No recommendations are made in relation to hospital costs, due to their exclusion from
      cover by LIMS schemes and options.

   c) In relation to medical devices, surgical consumables etc, it is recommended that a system
      of reference pricing, or alternatively, a similar approach to the SEP be adopted in order to
      reduce the negative impact of rebates and other perverse incentives which artificially
      inflate costs. It is recognised that this is ma inly a problem within hospitals, but impacts
      on the out of hospital sector as well.

   d) In relation to medicine pricing for LIMS, it is recommended that the Access Subsidy
      proposal put forward by elements of the pharmaceutical industry be actively investigated
      for its feasibility. This is regarded as a workable proposal by the LIMS stakeholders,
      subject to further analysis and investigation. There remains an open question as to the
      mechanism of pricing (and hence the discount to the SEP) within such a system, and
      LIMS did not reach consensus on the various options for pricing. These include a
      voluntary price reduction offered by individual manufacturers, the implementation of a
      standard discount to the SEP, or the use of state tender prices for LIMS members. It will
      not be essential to complete work on these issues prior to the launch of LIMS. Note that
      the cost estimates for the LMP are based on SEP medicine prices prior to any discount.

   e) It is recommended that the Council for Medical Schemes work actively with the
      Department of Health to review and address several aspects of current legislation and
      regulation which are inhibiting the emergence of cost effective and efficient integrated
      healthcare delivery models. These models hold the only prospect for meaningful and
      sustainable cost reductions in the supply of healthcare services in the medium to long
      term.




                                                                                              138
            Consultative Investigation into Low Income Medical Schemes – Draft Fi nal Report



   f) Specific aspects of legislation and regulation which require urgent review and
      modification include:

       i)     Legislation governing the scope of practice of various health professions. In this
              respect, there is a clear need for a centralized co-ordination of training and scopes of
              practices of the different professions that fall under different Statutory Councils.
              Provision for such centralization is made in Section 50 of the NHA through the
              establishment of a Forum of Statutory Health Professional Councils. It is
              recommended that Section 50 be proclaimed as this Forum is in essence charged with
              the responsibility of coordinating the various Health Professional Councils, the
              development of coherent policies relating to the education, training, optimal
              utilisation and distribution of health care professionals, as well as to advise the
              Minister on the scopes of practice of the various professions.

       ii) Various ethical rules and policies of the Health Professions Council of South Africa,
           particularly those relating to development of group practices, and to the ability of non
           health professionals to employment and/or share financial risk and rewards with
           health professionals.

3.10   Distribution of LIMS schemes

   a) The role of brokers and consultants, in collaboration with employers, will be critical in
      ensuring widespread distribution of the new LIMS schemes. Analysis by LIMS
      stakeholders suggests that distribution costs can be contained within the maximum levels
      prescribed by current legislation, although this will require alterations to the current
      business model. It is recommended that no further changes are made to legislation or
      regulations that might interfere with the ability of distribution agents to effectively
      distribute LIMS schemes to low income employees and their dependents.

   b) It is recommended that the Regulator, and industry stakeholders actively engage with
      both organised business and the trade unions, in order to discuss mechanisms through
      which these key stakeholders can be more actively involved in effective distribution of
      LIMS schemes to low income workers. Particular focus should be given to medium sized
      and especially small employers, where distribution of LIMS schemes is likely to be most
      problematic.




                                                                                                  139

				
DOCUMENT INFO