Documents
Resources
Learning Center
Upload
Plans & pricing Sign in
Sign Out

REACT

VIEWS: 16 PAGES: 25

									                         REACT
                  Work Package 7


       Influence on Emergency Obstetric Care
Research Protocol for Implementation at District Level




           Dominique Dubourg, Vincent De Brouwere


                        March 2006
                                                                                 2
                                                         REACT – WP7 Influence on Emergency Obstetric Care




                                                                                              Table of content
                             4
                             ..
 ................................................................................................................................                    INTRODUCTION
                      .........
                             5
        ................................................................................................................................                 OBJECTIVES
                            ...
                             5
  ................................................................................................................................                   MAIN OBJECTIVE
                            ................................................................................................
                             5
  .............................                                                                                                                   SPECIFIC OBJECTIVES
                      .........
                             5
        ................................................................................................                                   METHODOLOGY AND TOOLS
                   .............
                              5
             ................................................................................................                           STUDY AREA AND POPULATION
                              6
................................................................                                STUDY DESIGN: PART 1. USING THE UON STUDY IN REACT
                         .......
                              7
       ................................................................................................................................              Background
                    ................................................................................................
                              8
            ....................                                                                                                           Methodology and Tools
                              11
                               ................................................................................................
     .............................                                                                                                               OTHER INDICATORS
                              11
                ..................                   PART 2. BASELINE INVENTORY OF ALL SAFE MOTHERHOOD INTERVENTIONS IN DISTRICT
                                      PART 3. DOCUMENTATION OF DECISION MADE BY THE DISTRICT HEALTH MANAGEMENT TEAM IN
                              12
                               ................................................................................................
     .............................                                                                                                               SAFE MOTHERHOOD

                                           RESULTS PRESENTATION. STAGES AND CONTENT FOR DISSEMINATION (SEE
                             12
 ................................................................................................................................ DELIVERABLES)
                                    ETHICAL CONSIDERATIONS (SPECIFICS IN RELATION TO THE OVERALL REACT
                        .........
                             12
          ................................................................................................................................ APPROVAL)
                             13
 ................................................................                                CAPACITY BUILDING POSSIBILITIES AND NEEDS
                             13
     ............................                               LIMITATIONS, LINKS TO OTHER PROJECTS AND APPLIED ASPECTS
                      ................................................................
                             13
            .....................                                                                                     STANDARD OPERATING PROCEDURES
                             13
                             ....
     ................................................................................................                                TRAINING AND CAPACITY BUILDING
                          .......
                             14
        ................................................................................................................................                 WORK PLAN
                             15 .
  ................................................................                                BUDGET WITH UNIT COSTS AND JUSTIFICATION
                             15
                             ................................
     ............................                                                            LIMITATIONS, CHALLENGES AND OPPORTUNITIES
                            .....
                             15
      ................................................................................................................................                  REFERENCES
                          .......
                             17
        ................................................................................................................................                APPENDICES
                             17
                              ...
    ................................                                 APPENDIX 1. MATERNAL INDICATIONS TO CONSIDER FOR INCLUSION IN UON
                             19
                             ................................
     ............................                                                             APPENDIX 2 SPECIMEN QUESTIONNAIRE FOR WOMEN IN UON
                         ................................................................
                             22
         ........................                                                                                        APPENDIX 3 FERTILITY RATES BY COUNTRY
                             24
                .................                 APPENDIX 4. SEVERE MATERNAL MORBIDITY AND UON DATA COLLECTION STEP BY STEP
                             24
                             ................................................................................................
     ............................                                                                                                            Preliminary work
                             24
 ................................................................................................................................              Data collection
                             25
                             ....
     ................................................................................................................................             Coding data
                             25..
   ................................................................................................................................              Data analysis




                                                                                                             2
                                3
        REACT – WP7 Influence on Emergency Obstetric Care


                           Abreviations



A4R     Accountability for reasonableness

ANC     Antenatal Care

BEmOC   Basic Emergency Obstetric Care

EmOC    Emergency Obstetric Care

FGD     Focus group discussion

MOI     Major Obstetric Interventions

UN PI   United Nations Process Indicators

UON     Unmet obstetric need




                                   3
                                       4
               REACT – WP7 Influence on Emergency Obstetric Care




Introduction
In this work programme, maternal health is the domain in which the impact of
the introduction of the accountability for reasonableness (A4R) is assessed.
Maternal health remains a problem area in many developing countries.
International and historic data indicate that, in any population, a proportion of
pregnant women (usually 1-2%) will develop life-threatening obstetric conditions
in childbirth (Loudon 1992; Shorter 1984; Wilmott-Dobie, 1982; Hecht 1980;
Peller 1965; De Brouwere et al., 1998). If these women receive rapid medical
intervention, nearly all will survive. If they fail to receive appropriate assistance,
they will likely die.

The apparent universality of life-threatening obstetric conditions is not matched
by consistency in the ability of national or local health systems to respond to
them. Most pregnant women living in industrialized countries or in major urban
centres in developing countries have good access to medical facilities where
emergency obstetric care is provided. By contrast, women living in under-served
areas of developing countries may be unable to receive the emergency care they
need due to a variety of financial, geographic, logistical and socio-cultural
barriers.

The different abilities of various national and local health systems to respond to
obstetric emergencies are reflected in the wide range of maternal mortality ratios
observed worldwide. In Africa and Asia, where resource constraints and other
problems hamper health system performance, maternal mortality ratios can be
one hundred times greater (Rosenfield 1989). Even within developing countries,
huge disparities exist (Le Bacq & Rietsema, 1997; Fawcus 1997; De Brouwere et
al. 1996).

Process indicators developed in the domain of maternal health, such as unmet
obstetric needs can be used at district level to:
     Raise awareness of maternal health problems among both district health
      professionals and the community;
     Identify sub-district areas with the greatest need for prioritisation of
      investment in maternal health services;
     Evaluate the health system‟s coverage of maternal health needs; and
     Monitor progress of safe motherhood programmes in term of quality of
      care, equity and access to care.

Following the REACT 21-25 November 05 meeting, it was suggested that WP7
addresses quality, equity and trust dimensions of accountability for
reasonableness (A4R) through a few tools.

This protocol defines the method for collecting selected process indicators,
including UON and selected severe maternal morbidity.




                                          4
                                          5
                  REACT – WP7 Influence on Emergency Obstetric Care


Objectives

Main objective
To evaluate the change in access to EmOC on quality, equity and trust as a result
of the application of A4R to the priority setting/planning/decision making process
(Table 1).
    TABLE 1. WORKPACKAGE 7 – EMERGENCY OBSTETRIC CARE IN RELATION TO A4R
                                            DIMENSIONS
                                                         Dimensions of A4R

Registration of activity      Quality                   Equity                   Trust

Unmet Obstetric Need          Perinatal outcome         Geographical
(UON)                                                   differential in
                              Maternal outcome          utilization rates
                                                        Unmet need for MOI

EmOC Process Indicators       Case fatality rate        Coverage in
                                                        functioning BEmOC

Baseline inventory of all     Effectiveness of          Distribution of          Utilization and
Safe motherhood               implemented               interventions            coverage
interventions in district     interventions (ANC,
                              evacuation, malaria
                              protection)

Documentation of decision                               Documentation of shift   Population interviews,
made by the District Health                             in resources (human      FGD
Management Team in safe                                 and equipment)
motherhood




Specific objectives
     To monitor the rates of major obstetric intervention for absolute maternal
      indication in urban and rural areas of the districts unders study.

     To monitor the incidence and case fatality of severe maternal morbidity
      (which required or not a surgical intervention).

     To monitor the early perinatal mortality rates among women with the same
      severe morbidity.


Methodology and tools

Study area and population
The study area is the districts selected for A4R intervention in the three focus
countries: Zambia, Kenya, Tanzania. The population under study is the population
of pregnant women who undergo a major surgical intervention and those
admitted in hospitals for severe morbidity.




                                                    5
                                      6
              REACT – WP7 Influence on Emergency Obstetric Care


Study design: Part 1. Using the UON study in REACT
The unmet obstetric need indicator restricts its scope to maternal life threatening
conditions for which a major obstetric surgery is performed to solve the problem.
It is intended to help health personnel to answer the following questions:
      Are pregnant women receiving the major surgical obstetric interventions
       they need?
      Where are those women whose needs are unmet?
      How many women have unmet needs?
This WP could use the data yielded by the UON study in two ways. First, one
could attempt to assess the outcome of the introduction of the A4R intervention
by carrying out a comparison of results of the UON study at the start and at the
end of the project. Second, the potential of the UON study to „unlock‟ district
health staff by involving them in design, data collection and interpretation could
be tapped. Guindo et al. (2004) described how a UON study in different phases
led to a increased awareness of the problem and a review of the actual practices
and organisation of health services at district level. One could assume that the
local UON study may lead to some advocacy by the staff who carried out the
study in favour of maternal health. How the DHMT deals with this „pressure‟ in its
priority setting process could be used as a tracer to describe the application of
A4R principles. However, attribution of changes in the UON indicator to the A4R
intervention will be difficult, given that maternal health is typically a quality-of-
services sensitive issue. A simple before and after design is therefore unlikely to
give much information or evidence on this link.

The second option would be to have the UON study to be carried out by „local‟
health staff once the A4R intervention has been introduced and to monitor how
the obtained results are appropriated and used by these staff to do something
about it. In other words, the UON study becomes a tracer health issue in the
documentation of the priority setting process by the REACT research staff (WP4?).
Also in this scenario, a follow-up study would be carried out at the end of the
study to assess progress in the domain of maternal health, but because
management decisions in this domain would have been documented, attribution
of changes to interventions could be attempted more easily.

In both cases, technical support to carry out the UON study will be provided by
the ITM UON team. An initial workshop will be organised to introduce the
technique and train the district health staff who will carry out the actual study.
Follow-up support by email has proven to be very effective in assuring a correct
implementation.

The remainder of this protocol describes the basic concepts underlying the unmet
need indicator and the steps required to use it.




                                         6
                                            7
                    REACT – WP7 Influence on Emergency Obstetric Care


Background

THE CONCEPT
The concept of unmet obstetric need (UON) is merely the difference between the
number of cases of obstetric need and the number of such cases which are
covered by health services in reality.
In situations where the incidence of major obstetric indications is known, it is
possible to calculate unmet need directly, by subtracting the number of major
obstetric interventions performed from the total number of women with
indications for such interventions. This is shown in Figure 1, below:
                      FIGURE 1. THE CONCEPT OF UNMET OBSTETRIC NEEDS

            Number of women
                                                        Number of who received
           in a population with
              indications for                -           obstetric intervention         =    Unmet need
          obstetric intervention


The unmet obstetric need indicator is restricted to some of the absolute (life
threatening) obstetric indications, and is calculated using a reference rate or
benchmark.

Absolute maternal indications (AMI) are used since, by definition, women with
such conditions must receive treatment in a medical facility, or risk death.
Women with absolute maternal indications who are not treated at a medical
facility can be presumed to have unmet obstetric needs.1

THE BENCHMARK
The benchmark used to estimate the number of women with absolute maternal
indications was first derived from international literature on maternal morbidity.
The UON exercises carried out in 7 countries between 1999 and 2001 have
resulted in a more robust estimate of the a priori need for major obstetrical
interventions for absolute maternal indications We can now be reasonably
confident that the figure of 1.4% (CI 1.27%-1.52%) is a sensible low-end
estimate of the proportion of deliveries that require a major obstetric intervention
to avoid a maternal death. This benchmark may then be applied to data from
more remote or dispersed populations in which many women experiencing life-
threatening indications die outside the formal health care system.

The benchmark multiplied by the number of births expected in the population
during the study period gives the expected number of women experiencing
absolute maternal indications in the population. However, it is also possible to use
local rates calculated in an area close to the hospital as a benchmark to measure
deficit in remote areas.
The second element of the equation - the number of major obstetric interventions
(MOI) performed for absolute maternal indication (AMI) is the sum of all major
obstetric interventions performed for an absolute maternal indication.


1
  Non-absolute indications are excluded from the UON calculation, since some women with such indications could
have “met” needs even if they never receive treatment at a medical facility. This would be the case, for example, if
their conditions were self-limiting, or were successfully treated at a pharmacy.



                                                         7
                                             8
                     REACT – WP7 Influence on Emergency Obstetric Care


EXPECTED IMPACT
By carrying out this type of study, it is expected that health workers at the
peripheral level, by estimating the extent of the deficits and analysing their
causes, will become aware of the scale of the problem and of potential means of
tackling it, improving the quality and accessibility of emergency obstetric care as
well as increasing equity and trust in hospital care (Guindo et al. 2004).

This is only possible if health workers are motivated, hence the crucial importance
of their involvement in the collection and analysis of data.

Methodology and Tools
To implement the unmet obstetric need approach at district level, the following
steps are required:

STEP 1. MINIMUM CRITERIA FOR IMPLEMENTATION ARE MET
Three minimum criteria must be met. These concern data availability, resources
and population size.
(1) Population Size
The population is higher than 200,000 inhabitants (-year) with a minimum of
5,000 expected births or 25 interventions.
(2) Data Availability
The denominator (expected births) is known with a fair level of reliability.
Expected births, broken down by urban (a defined radius around the general
hospital between 5 and 15km) or rural area could be calculated using age specific
fertility rates, general fertility rate or crude birth rate, depending on the
availability of census data. These rates are applied to the population of reference
projected from the last available census.2

Data should be collected in all health facilities (public and private). Information on
the numerator (type of intervention, type of indication, origin of the mother) can
be collected with a reasonable degree of accuracy and reliability. Hospital records
(operating theatre registers, patient files, etc.) must contain information about
the date and type of each major obstetric intervention performed, the major
indications for the intervention, and the address of the patient
(3) Resources
District health professionals, especially midwives and obstetricians, are willing to
participate in the UON exercise, and are available to participate in initial meetings
to determine the list of indications and interventions to be included.

STEP 2. SELECTING ABSOLUTE INDICATIONS AND INTERVENTIONS FOR INCLUSION
The UON exercise is restricted to a limited number of absolute maternal
indications that are decided in each country by a panel of national experts. It is
necessary to involve all field professionals to initiate a setting for local change.
This approach also allows to aggregate data and to make inter-district
comparisons.

2
    Dates of last census: Kenya 1999, Tanzania 2002, Zambia 2000.



                                                        8
                                             9
                     REACT – WP7 Influence on Emergency Obstetric Care


The indications should meet the following two criteria:
         They are recognized by local specialists as life-threatening conditions. (For
          this purpose “life-threatening” means likely to result in death in the
          absence of a major obstetric intervention.)
         They are defined in a standard way in facility records, or information about
          the indication is recorded in a manner that makes it possible to create a
          standardized definition retroactively.
Next, the expert panel determines which major obstetric interventions should be
included in the survey. The interventions should meet the following criteria:
         The interventions are considered by local professionals to be the
          appropriate response to one or more of the absolute maternal indications
          included in the study, given local resources.
         The interventions are performed only in the health facilities to be included
          in the UON data collection exercise.
Appendix 1 contains a comprehensive list of indications and interventions. The list
differentiates those indications and interventions that have been used successfully
in previous UON exercises (UON Network), and those that were omitted due to
problems related to the criteria for inclusion described here.

STEP 3. MEASURING THE MET AND UNMET NEED IN THE DISTRICTS
Criteria for inclusion
The questionnaire for women covers all women admitted to a health structure
during the year of the study for a major obstetric intervention together with all
women whose death in these structures was related to pregnancy or the
consequences of childbirth.
Collecting data
Prior to data collection, the data collection instrument must be developed and
tested (see example in appendix 2). The instrument should allow the following
essential information to be collected systematically for each intervention:
         file number
         date of intervention
         address of patient (for which the level of precision may be the area of
          residence or health centre)
         type of intervention
         indications for the intervention3
         result for the child
         result for the mother
The questionnaires must, of course, be adapted to the local context. Particular
attention should be paid to the way in which diagnoses are formulated in the
registers, and the items in the questionnaire should be worded in a form as close
as possible to the way they are usually expressed.

3
  It may also be useful, for planning and analysis purposes to record information on the state of mother at discharge
(alive, deceased, complications); state of child after 24 hours (alive, deceased, stillbirth (macerated or fresh); and, if
possible, type of access to hospital (direct, referred by health center, re-admission).



                                                            9
                                           10
                    REACT – WP7 Influence on Emergency Obstetric Care


Data are collected from the routine information system in the hospital services.
The principal data source is generally the operating theatre register where usually
most major obstetric interventions are recorded. Information about the
indications for the interventions and personal data on the patient may be found in
patient files, maternity ward registers, admissions records for maternity or
surgery wards, or delivery files.

When more than one indication for an intervention is recorded on the woman‟s
file, keeping only one indication -the most absolute- makes the analysis easier. A
hierarchical list (ranked from most to least serious has to be established as
presented in Table 2.
    TABLE 2. PRIORITY ORDER OF THE INDICATIONS OF MAJOR OBSTETRIC INTERVENTIONS
1. Ruptured uterus                  7. Prolonged labour for       13. Breech presentation
                                    other causes
2. Obstructed labour due            8. Complications related to   14. Previous C-section
to transverse lie                   cord
3. Obstructed labour due            9. Ante partum                15. Other bad obstetric
to brow presentation                haemorrhage due to            history
                                    placenta praevia
 4. Obstructed labour due           10. Other ante partum         16. Other causes
to cephalo-pelvic                   haemorrhage
disproportion
5. Obstructed labour for            11. Postpartum                17. foetal distress
other presentation                  haemorrhage
6. Prolonged labour for             12. Hypertensive disorders    18. Not recorded cause
dynamic dystocia



A more detailed step by step description of the method is available on the UON
network website (http://www.uonn.org) and summarised in Appendix 4.
Study sites
In each focus country an inventory will be drawn up of all health facilities used by
the district population, including both public and private hospitals in districts
involved in the study and of all hospitals, which are outside the districts but are
possibly used by women living within them.

The population of each district and health centre areas will be projected for the
year of the study, using the last census available and if possible the more
accurate growth rates. The expected births will be calculated by using the more
effective indicator: regional Age-Specific Fertility Rate (ASFR) apply to the female
population by age group or regional Global Fertility Rate (GFR) if only the total
population of females 15-49 is known or the regional Crude Birth Rate (CBR)
where only the total population count is available.

Fertility rates for each country are available in the most recent DHS and
presented in Appendix 3.4




4
    Kenya 2003, Tanzania 1999, Zambia 2001/02



                                                  10
                                           11
                    REACT – WP7 Influence on Emergency Obstetric Care


Analysing data
The unmet obstetric need indicator is calculated as follows for a defined
population:
                         Unmet Obstetric Need = (EB x RR) - (MOI for AMI)


whereby
EB is the expected number of births in the population
RR is the bench mark for the proportion of births associated with absolute
maternal indications for major obstetric intervention

MOI for AMI is the number of women in the population receiving major obstetric
interventions for absolute maternal indications

The deficits are calculated by urban/rural and total areas.5 It is important to
record the residence of the woman for whom the MOI/AMI was performed
because unmet need is calculated for the zone where the woman lives and not for
the catchment area of the hospital.

To appreciate the significance of these proportions of unmet need it is also useful
to translate them into absolute figures as well.

The MOI/AMI deficit ratio is calculated by dividing the deficit by the number of
MOI/AMI expected for urban and rural areas and the district as a whole x 100.

STEP 4. MEASURING THE EARLY PERINATAL MORTALITY
The data collected make it possible to calculate the early perinatal mortality
among the women who underwent a major obstetric intervention. This indicator
can be used to judge the quality of care in hospital or to reveal the consequences
of late referrals and consider their causes.

Other indicators
A panel comprising medical officers working in the district hospitals unders study
and experts will select a list of severe maternal morbidity requiring or not a
surgical procedure. Rigorous criteria for identifying these severe conditions will be
defined and cases will be recorded all along the year on a special form.

The regular compilation of these conditions according to the origin of the women
will allow to calculate progress in case fatality rates, improvement of maternal
and perinatal outcomes.

Part 2. Baseline inventory of all Safe motherhood
interventions in district
As soon as the districts will be selected, the country researchers will make an
inventory of all safe motherhood strategies currently implemented in the districts
and will collect routinely collected information related to this programme (ANC

5
  An urban population is defined for this purpose as a population living within 10 kilometers of a hospital in which
major obstetric interventions are performed. If any other definition is used (for example the administrative definition
of an urban population) this must be clearly stated



                                                         11
                                      12
               REACT – WP7 Influence on Emergency Obstetric Care


coverage, proportion of institutional deliveries, proportion of deliveries with skilled
attendants if relevant).

Part 3. Documentation of decision made by the District Health
Management Team in safe motherhood
In coordination with other relevant WP (WP4), the documentation of decision
taken in the domain of safe motherhood will be made on an ongoing basis. At the
end of the research, focus group discussions will be organised with population
leaders and questions will be inserted in the post-intervention household survey.


Results presentation. Stages and content for dissemination
(see deliverables)
Given that this is in both scenarios a longitudinal study, dissemination of final
findings and results regarding the impact of the A4R intervention in the domain of
maternal health can be foreseen to happen at the end of the study. Effective
dissemination will need relevant stakeholders to be involved early on in the
design and implementation.

However, the results of the initial UON study need to be disseminated to the local
district health management team, and this is usually a „natural‟ process that flows
out of doing the study.


Ethical considerations (specifics in relation to the overall
REACT approval)
The protocol will be sent for ethical clearance to the relevant authorities. For the
UON exercise, the data will be collected with the full participation of the hospital
staff and the formal authorisation of the hospital directors. For the qualitative
data collection in the second scenario, we will seek written consent or verbal
informed consent from the interviewees.

Absolute anonymity (or keeping the identity of participants known only to
research team members) will be strived at, but it may prove difficult to give
absolute guarantees for all staff involved. Ensuring anonymity of staff being
interviewed will be attempted by not discussing their selection with the hospital
team members or any other hospital staff. Given the involvement of the hospital
management team, for them this is not possible. These modalities will be
discussed with all potential interviewees at the start.

To protect confidentiality, no individual-level data collected by this study will be
made public for any reason. Individual identifiers will be kept separately from the
data and all data will be kept strictly confidential. Sources referred to in reports
and papers will be anonymous and not attributed directly, nor by reference to
characteristics that could lead to indirect attribution. in any way.

Staff members will be invited for the interviews by letter. An informed consent
form will be signed by the interviewee and counter-signed by the investigators.
Interviewees will have the possibility to opt out at any time.



                                          12
                                     13
              REACT – WP7 Influence on Emergency Obstetric Care


Capacity building possibilities and needs
To be discussed with partners.


Limitations, links to other projects and applied aspects
To be discussed with partners.


Standard Operating Procedures

Training and capacity building
The training of the researchers and of the personnel wishing to be involved in the
collection of data will be organised during the first visit of WP7 ITM people. This
training is carried out on the spot and needs a preparation of a few weeks to
ensure that the persons and the forms are available.

The analysis of data will be carried out together with national researchers and the
district team since the objective is also to strengthen the district team capacity in
doing this kind of monitoring.




                                         13
                                               14
                        REACT – WP7 Influence on Emergency Obstetric Care




        Work plan
Steps           Activities                          Timing       Outputs        Responsibilities   Synergy
                                                                                                   with

Drafting the    Preparation       Step 1 -          March-May    Tool           Country research
country UON                       Assessment of     2006                        teams
specific                          minimum
protocols                         criteria for
                                  implementation

                                  Step 2 -          June 06      List of AMI    Country research
                                  Selecting                                     teams in
                                  absolute                                      collaboration
                                  indications and                               with district
                                  interventions                                 team, national
                                  for inclusion                                 experts and WP7
                                                                                leader

                Induction and     Step 3.           Around       District-      Country research
                training          Measuring met     December     specific       teams in
                                  and unmet         06           protocol       collaboration
                                  need                                          with district
                                                                 Data           team and WP7
                                                                 collection     leader
                                                                 started

The UON field   Completing        Step 3.           Within 1-4   UON report     Country research
studies         UON field study   Measuring met     weeks                       teams with
                                  and unmet         after the                   district team in
                                  need              training                    collaboration
                                                                                with WP7 leader
                                                    Then,
                                                    every year

EmOC process    Induction and     UN material       Same as      UN             Country research
indicators      training          exists            UON study    indicators     teams with
                                                                 report         district team in
                                                                                collaboration
                                                                                with WP7 leader

Baseline        Preparation       Exchange on       June 06      Agreement      Country research   WP8
inventory of                      potential list                 on list        team in
SM strategies                                                                   collaboration
                                  Interview grid                 And on grid    with WP7 leader

                Implementation    Interview and     July 06      List of        Country research
                                  documentary                    current safe   team in
                                  research                       motherhood     collaboration
                                                                 strategies     with WP7 leader
                                                                 in district

Documentation   Preparation       Development       April-June   Approach       WP7 leader,        WP4
of decision                       of method with    06           described      country research
made                              WP4                            and            team in
                                                                 validated      collaboration
                                                                                with WP4

                Implementation    Regular           Ongoing      Report         Country research   WP4
                                  interviews,       process                     teams in
                                  document          from end                    collaboration
                                  review            of 2006                     with WP7 & WP4
                                                    onwards                     leaders




                                                      14
                                      15
               REACT – WP7 Influence on Emergency Obstetric Care


Budget with unit costs and justification
ACTIVITY 1. PREPARATION OF UON AND UN PI STUDIES:
Country research team to go to the selected district and check if information is
available for studying UON, UN process indicators and severe maternal morbidity
(number of c-sections, quality of routine information, etc.)

        Time of WP7 and country researcher to discuss list of items to collect
        Transport from country research institution to district for 1 researcher

ACTIVITY 2. SELECTING ABSOLUTE MATERNAL INDICATIONS FOR INCLUSION
Country research team to organise a meeting with the district team and one or
two gynaecologists after having read documentation on the topic.

        Transport from research institution to district

        Honorary for external gynaecologists

        Time of WP7 and country researcher to discuss list of indications

ACTIVITY 3. MEASURING UON AND UN PI
Collecting data in district hospitals with hospital staff and analysing them

        Visit from Belgium by WP7 to country teams (1 travel + 3 weeks)

        Transport for 1 researcher from research institution to district in each of
         the 3 countries

        Questionnaires to multiply (depending on the number of MOI)

        Allowance for the hospital staff helping collecting data



ACTIVITY 4. BASELINE INVENTORY OF SAFE MOTHERHOOD STRATEGIES


ACTIVITY 5. DOCUMENTATION OF DECISION MADE IN THE DOMAIN OF SAFE MOTHERHOOD



Limitations, challenges and opportunities
Within the work package
Harmonization/coordination issues to other WP‟s
Links to other projects and application


References
AbouZahr C & Royston E. 1991. Maternal Mortality: a Global Factbook. World Health
Organization, Geneva.
De Brouwere V, Laabid A, & Van Lerberghe W. 1996. Evaluation des besoins en
interventions obstétricales au Maroc; une approche fondée sur l'analyse spatiale des
déficits. Revue d’Epidémiologie et de Santé Publique, 44(2):111-124.



                                            15
                                       16
                REACT – WP7 Influence on Emergency Obstetric Care


De Brouwere V, Tonglet R, Van Lerberghe W. 1998. Strategies for reducing maternal
mortality in developing countries : what can we learn from history of western countries ?
Tropical Medicine and International Health, 3, 771-782.
De Brouwere V & Van Lerberghe W. 1998. Les besoins obstétricaux non couverts.
L'Harmattan, Paris.
Fawcus S, McIntyre J, Jewkes RK, Rees H, Katzenellenbogen JM, Shabodien R, Lombard
CJ, Truter H. 1997. South African Medical Journal, 87(4): 438-42.
Filippi V, Brugha R, Browne E, Gohou V, Bacci A, De Brouwere V, Sahel A, Goufodji S,
Alihonou E, Ronsmans C. 2004. Obstetric audit in resource-poor settings: lessons from a
multi-country project auditing 'near miss' obstetrical emergencies Health Policy and
Planning. 19(1):57-66
Guindo G, Dubourg D, Marchal B, Blaise P and De Brouwere V. 2004. Measuring unmet
obstetric need at district level: how an epidemiological tool can affect health service
organization and delivery. Health Policy and Planning, 19(suppl_1): 87-95.
Hecht J. 1980. L'évaluation de la mortalité aux jeunes âges dans la littérature économique
et démographique de l'ancien régime. (In Boulanger MA & Tabutin D (eds), La mortalité
des enfants dans le monde et dans l'histoire. Liège, ORDINA EDITIONS, p. 29-83).
Kaunitz AM, Spence C, Danielson TS, Rochat RW & Grimes DA. 1984. Perinatal and
maternal mortality in a religious group avoiding obstetric care. American Journal of
Obstetrics and Gynecology, 150: 826-831.
Le Bacq F & Rietsema A. 1997. High maternal mortality levels and additional risk from poor
accessibility in two districts of northern province, Zambia. International Journal of
Epidemiology, 26(2): 357-63.
Loudon I. 1992. The transformation of maternal mortality. British Medical Journal; 19-
26;305(6868): 1557-60.
Peller S. 1965. Births and deaths among Europe's ruling families since 1500, Chicago,
Aldine Publishing Company.
Ronsmans C et al. 1997. Women‟s recall of obstetric complications in South Kalimantan,
Indonesia. Studies in Family Planning, 28(3): 203-214.
Ronsmans C, De Brouwere V, Dubourg D, Dieltiens G. 2004. Measuring the need for life-
saving obstetric surgery in developing countries. British Journal of Gynaecology and
Obstetrics. 111, 1027–1030
Rosenfield A. 1989. Maternal mortality in developing countries: an ongoing but neglected
'epidemic'. Journal of the American Medical Association, 262(3): 376-379.
Shorter E. 1984. Le corps des femmes. Paris, Editions du Seuil.
Stewart MK et al. 1996. Issues in measuring maternal morbidity: lessons from the
Philippines Safe Motherhood Survey Project. Studies in Family Planning, 27(1): 29-35.
UNICEF, World Health Organization & UNFPA. 1997. Guidelines for monitoring the
availability and use of obstetric services. UNICEF, New-York.
UON Network Available http://www.uonn.org/uonn/eng/home2b.html#home2b (12 August
2005).
World Health Organization. 1994. Indicators to monitor maternal health goals. Report of a
technical working group, Geneva, 8-12 November 1993. WHO/FHE/MSM/94.14:1-43,
World Health Organization, Geneva.
Wilmott Dobbie BM. 1982. An attempt to estimate the true rate of maternal mortality,
sixteenth to eighteenth centuries. Medical History, 26: 79-90.




                                            16
                                                     17
                              REACT – WP7 Influence on Emergency Obstetric Care




            Appendices

            Appendix 1. Maternal Indications to consider for inclusion in
            UON
Indication                                  Caveats and Comments
severe antepartum haemorrhage (for          Generally recognized as an AMI. Fair reliability.
placenta praevia or abruptio placentae)

severe postpartum haemorrhage               Difficult to standardize. Probably highly specific when a hysterectomy is
                                            performed to stop bleeding (as used in Morocco).
                                            Criterion could be: PPH for which the decision was to perform a hysterectomy.

major foeto-pelvic (mechanic)               Reliability depends very much on the obstetricians‟ culture. In some countries,
disproportion (CPD) due to relatively       it is a specific and reliable indication; in other (e.g. USA), dynamic and
small pelvis or relatively big foetus or    mechanic dystocia are labelled CPD.
hydrocephalus
                                            When it turns to become a uterine rupture, everybody recognizes it as an AMI.

malpresentation (shoulder or                Transverse lie is usually reliable. Brow presentation may be confused with face
transverse lie and brow presentation)       for non-specialists.

hypertensive disorders                      Hypertension and pre-eclampsia are not systematically fatal. The cut-off points
                                            to consider these conditions as AMI are not clear. Treatment may be
                                            decentralized, so that one should collect data from all health centres where
                                            deliveries take place.
                                            Eclampsia, however, should be considered in settings where it is relatively
                                            frequent.

breech, face, cord prolapse presentations   These conditions are not considered as systematically AMI. There is a
                                            probability that some will necessitate a MOI, but this probability is not known
                                            and the decision to perform a C-section depends very much on the obstetrician
                                            behaviour or on the pressure from the family to avoid any risk.

multiple pregnancies                        Is a risk and not a problem as such.

severe anaemia                              It would be possible to consider severe anaemia as a specific need on the basis
                                            of a defined cut-off point (<6g?). But, on one hand many women live with
                                            severe anaemia and yet do not consult for anaemia; therefore it is difficult to
                                            set up a reference rate. On the other hand, prevalence of anaemia varies very
                                            much from place to place and from social class to social class so that it makes
                                            difficult to take the urban population (where there is the best information) as
                                            reference rate. Finally, it does not always require surgery or hospital
                                            management

infection                                   Is supposed to not exist. Therefore there is no reference rate.

abortion                                    Because alternatives exist, it is difficult to calculate a reference rate without a
                                            tricky survey on abortion (and rates vary from year to year).

post-partum haemorrhage (without MOI)       Difficult to standardize the blood loss. Sometimes, confusion between the act of
                                            uterine revision and the problem of incomplete placenta. Management of PPH
                                            may be decentralized (manual removal of placenta, curettage, oxytocyn). If
                                            one wants to take this condition into account, it means that the survey should
                                            be extended to all facilities where PPH is managed.

psychosis                                   Difficult to standardize.

embolism                                    Difficult to diagnose. Relatively rare complication.

cervix or perineal tear                     Results from bad management of delivery. No reference rate.

            Note: indications shown in bold have been used in previous UON exercises.




                                                            17
                                           18
                    REACT – WP7 Influence on Emergency Obstetric Care


                       INTERVENTION TO CONSIDER FOR INCLUSION IN UON
Intervention                           Caveats and Comments
caesarean section                      Is by definition a MOI. Easy to identify in a register. A C-section,
                                       however, can be performed not only to save the mother‟s life but also
                                       to save the newborn‟s life or for many other reasons not related to life
                                       threatening situations.

laparotomy for uterine breach          In case of ruptured uterus, it is sometimes better to repair the breach
                                       using a laparotomy. It is a MOI and the indication is obviously for an
                                       AMI (ruptured uterus).

hysterectomy                           It is a MOI. Performed in case of ruptured uterus impossible to repair or
                                       in case of bleeding uterus.

internal version                       It is a (rare) MOI performed in order to avoid a C-section.

symphysiotomy                          Intervention performed in order to avoid a C-section in case of cephalo-
                                       pelvic disproportion. Usually performed in hospital (so that in case of
                                       failure, it is rapidly possible to perform a C-section). Considered as a
                                       MOI.

craniotomy                             Intervention performed in place of a C-section when the newborn is
                                       dead. Considered as a MOI.

forceps                                The use of forceps is difficult to standardize and depends very much on
                                       the experience of the obstetrician. No reference rate.

vacuum extraction                      Same as for forceps.

manual removal of placenta             Might be considered. But, since it is possible to perform a manual
                                       removal of the placenta outside the hospital, it means that one has to
                                       extend the survey to all facilities where such an intervention is
                                       performed (cost increases). Sometimes, confusion between curettage
                                       and manual removal of pieces of placenta or uterine revision.

blood transfusion                      Might be considered but another criterion should be added: the level of
                                       haemoglobin (e.g. Hb<6g):

curettage of womb                      Might be considered. But, should be performed in case of haemorrhage
                                       due to incomplete placenta (and not as an exploratory exam). If it is
                                       considered, one should include in the survey all the facilities performing
                                       curettage.

perineal tear repair                   No reference rate for perineal tear.

treatment of infection                 No reference rate for perineal tear

Note: interventions shown in bold have been used in previous UON exercises.




                                             18
                                        19
                 REACT – WP7 Influence on Emergency Obstetric Care



Appendix 2 Specimen questionnaire for women in UON
Q1 – Identification of health facility
        Department ________________
Health district ________________
Category of hospital ______
Name of health facility________
Q2 - Identification of parturient
        Admission number
Q3 –    Date of admission: _____/_____/_____/
        Date of delivery: _____/_____/_____/
Q4 –    Year of birth: ______________
Q5 – Address of parturient
        District: ______________
        Village/city: ___________
        Quarter: ______________
Q6 – Health centre area: __________________________
Q7– Type of area         1= Urban        2= Rural         9= Unknown
Q8 – Place of delivery: 1=At home        2= this health facility       3= another facility
Q8bis Which other? ______________
Q9 – Major Obstetric Intervention
        Date of intervention: _____/_____/_____/
Q10 – Type of intervention MOI
        1= Caesarean
        2= Hysterectomy
        3= Laparotomy for uterine tear / uterine rupture
        4= Version and extraction
        5= Craniotomy / Cranioclasy / Embryotomy
        6= Symphyseotomy
        7= Other (specify)
Q11 – Indication
        1= Uterine rupture
        2= Obstructed labour for transverse presentation
        3= Obstructed labour for frontal presentation
        4= Obstructed labour for foeto-pelvic disproportion
        5= Obstructed labour for other presentation
        7= Obstructed labour for dynamic dystocia
        8= Obstructed labour for other cause
        9= Other cause
        10= Complication connected with cord
        11= Ante-partum haemorrhage for placenta praevia



                                              19
                                       20
                REACT – WP7 Influence on Emergency Obstetric Care


        12= Ante-partum haemorrhage for retro-placental haematoma
        13= Ante-partum haemorrhage for other cause
        14= Post-partum haemorrhage
        15= Toxaemia, pre-eclampsia
        16= Eclampsia
        17= Puerperal infection
        18= Breach presentation
        19= Antecedent of caesarean
        20= Other obstetric antecedent
        21= Foetal distress
        22= Cause not recorded
        99= Other cause (specify) ___________________
Q12 – Results for child
        1= Born living and emerged living
        2= Still-born
        3= Born living and died within 24 hours
        4= Born living and died after 24 h
        9= Not recorded
Q13 – Results for mother
        1= Nothing to report
        2= Complication See Q14
        3= Referred to another health formation
        4= Died See Q15 and Q16
        9= Not recorded
Q14 – Type of complication________________
Q15 – When mother died
        1= Before intervention
        2= During intervention
        3= After intervention
9= Not recorded
Q16 – Cause of mother death
        1= Hypertensive disorder
        2= Haemorrhage
        3= Infection
        4= Other (specify)
        9= Unknown
Q17 – Date of mother's discharge_____/_____/____
Q18 – Name of surveyor
Q19 – Date of completion of questionnaire: _____/_____/____
Q20 – Check: when? ____/_____/____
        By whom: ____________________



                                             20
                                      21
               REACT – WP7 Influence on Emergency Obstetric Care




Observation (about problems during survey or other observation)




                                          21
                                                    22
                             REACT – WP7 Influence on Emergency Obstetric Care



              Appendix 3 Fertility rates by country
                  Age 5-year groups

                                                                                      General        Crude birth
                   15-19     20-24    25-29    30-34    35-39    40-44     45-49    fertility rate      rate

Kenya 2003

Urban                 88.0    162.0    168.0    136.0     78.0     23.0       6.0           123.0          35.3

Rural                123.0    278.0    254.0    217.0    137.0     62.0      17.0           187.0          38.1

Region

Nairobi               68.0    131.0    158.0    126.0     54.0       9.0      0.0           106.0          33.3

Central               84.0    199.0    171.0    121.0     78.0     34.0       0.0           122.0          27.8

Coast                151.0    234.0    224.0    216.0    107.0     45.0      13.0           179.0          39.7

Eastern               84.0    275.0    209.0    223.0    154.0     53.0      22.0           173.0          34.8

Nyanza               141.0    295.0    277.0    206.0    134.0     39.0      20.0           192.0          41.3

Rift Valley          144.0    260.0    264.0    216.0    149.0     97.0      33.0           201.0          42.2

Western              115.0    291.0    306.0    265.0    134.0     51.0       0.0           197.0          40.5

Tanzania 1999

Urban                 95.0    199.0    170.0     93.0     58.0     17.0       0.0           128.0          34.4

Rural                154.0    301.0    269.0    262.0    165.0     96.0      49.0           223.0          44.3

Region

Coastal              136.0    207.0    185.0    179.0    117.0     43.0       3.0           157.0          35.7

Northern
Highlands             77.0    270.0    226.0    163.0    138.0     77.0      63.0           179.0          40.2

Lake                 163.0    317.0    308.0    268.0    173.0    133.0      54.0           238.0          51.1

Central              104.0    332.0    277.0    230.0    119.0       0.0     15.0           203.0          38.5

Southern
Highlands            125.0    264.0    244.0    186.0    111.0     66.0      52.0           185.0          37.1

South                199.0    198.0    151.0    232.0    133.0     38.0      45.0           179.0          37.5

Zambia 2001/02

Urban                127.0    192.0    197.0    159.0    113.0     43.0      25.0           154.0          36.7

Rural                185.0    320.0    288.0    254.0    204.0    101.0      33.0           239.0          47.0

Region

Central              176.0    303.0    278.0    242.0    167.0     58.0      22.0           224.0          45.2

Copperbelt           136.0    201.0    194.0    176.0    111.0     70.0      10.0           160.0          36.5

Eastern              186.0    317.0    273.0    269.0    182.0     87.0      43.0           239.0          46.2

Luapula              196.0    332.0    289.0    272.0    239.0     85.0      50.0           249.0          52.9

Lusaka               130.0    199.0    197.0    141.0    113.0     39.0      50.0           155.0          36.8

Northern             164.0    309.0    319.0    258.0    201.0    116.0      16.0           234.0          48.0

North-Western        166.0    296.0    318.0    254.0    218.0     71.0      27.0           232.0          45.6

Southern             169.0    271.0    279.0    210.0    205.0     71.0      23.0           213.0          42.2

Western              169.0    296.0    234.0    221.0    197.0    116.0      48.0           218.0          45.9

              Source: ORC Macro, 2005. MEASURE DHS STATcompiler. http://www.measuredhs.com, August 10
              2005.




                                                         22
                                     23
              REACT – WP7 Influence on Emergency Obstetric Care


Age-specific and total fertility rates: Age-specific and cumulative fertility
rates for the three years preceding the survey, by urban-rural residence and
region.
General fertility rate: General fertility rates for three years preceding the
survey by urban-rural residence and region.
Note: Rates are for the period 1-36 months preceding the survey. Rates for age
group 45-49 may be slightly biased due to truncation. General fertility rate (births
divided by number of women 15-49), expressed per 1,000 women.
Crude birth rate: Crude birth rates for three years preceding the survey by
urban-rural residence and region.
Note: Rates are for the period 1-36 months preceding the survey. Rates for age
group 45-49 may be slightly biased due to truncation. Crude birth rate expressed
per 1,000 population.




                                        23
                                     24
              REACT – WP7 Influence on Emergency Obstetric Care



Appendix 4. Severe maternal morbidity and UON data
collection step by step


Preliminary work
      Adapt data collection questionnaires in order to take into account country
       particularities and to add specific research questions identified by local
       team members.
      List all facilities (public and private) where women native from the districts
       under study may access in case of surgical intervention.
      List all health centre areas of the study district (and if possible all villages)
       and define the total population for the study year.
      Define the type of area (urban or rural) of each of the health centre areas
       and the distance from the health centre to the hospital.
      Give a unique code to each geographical unit (health areas and villages)
       and to each health facility.
      Visit all the hospitals of the study area in order to check the quality of the
       records and the completeness of the information sources so that we have
       some argument to decide which source it is better to start with.
      Check with physicians working in these hospitals the way diagnoses are
       recorded in the registers and patients‟ files and to adapt data extraction
       forms to the local language.
      Determine, on the basis of hospital activity reports, the number of Major
       Obstetric Interventions and of severe maternal morbidity cases expected
       during a year just to have an idea of the number of questionnaires we
       should print and make the budget accordingly.
      Test questionnaires.



Data collection
From the information recorded in the maternity register (or any other source
defined as primary resource – see point 6 before), fill a UON and/or severe
maternal morbidity questionnaire for every woman who :
      undertook a Major Obstetric Intervention (any indication)
      has been admitted (or developed while in the hospital) as severe maternal
       complication
      died in the hospital during pregnancy, childbirth or in the post partum
For women who were severe maternal morbidity cases and who underwent a
major obstetric intervention, two separated forms (UON and severe maternal
morbidity) will be filled
Keep collecting information asked with the same questionnaire, using other
information sources and taking care to cross check these different sources.
In case of incoherence between sources, try to obtain clarification from the health
care providers.




                                          24
                                      25
               REACT – WP7 Influence on Emergency Obstetric Care


Coding data
Define code for each variable before encoding. (Codes are unique)
Prepare a screening mask for entering data (Epi6, Epi Info 2002, Access…)
Make a double entry of data
Prepare a cleaning procedure and apply it to clean files



Data analysis

UON
Preliminary work
Calculate the expected births for each geographical unit, where possible by using
the Age Specific Fertility Rate or Global Fertility Rate for the region.
Determine the benchmark, either by using data from the previous UON studies
(1,4% of Expected Births), or by calculating a local benchmark based on the MOI
rate for AMI observed in urban areas where all the women of a city are
considered to have an optimal access to hospital in case of obstetrical emergency.
Calculate the expected MOI for AMI : “benchmark” x “Expected Births”


Analysis
(Basic analysis, while other analyses may be carried out depending on the
research questions)
Number of IOM for IMA according to place of residence of the woman and
according her area of residence (urban/rural)
IOM rate for IMA according to her place of residence and according to her area
type of residence.
Deficits per geographical zone and according to the area type of her residence: in
absolute and in relative values
Maternal deaths after MOI according to group of indications (AMI against Non-
AMI) and to the type of area (case fatality rates).
Early perinatal mortality : total, stillbirth, early neonatal mortality according to a
group of indications (AMI against NON-AMI)
Proportion of MOI done for AMI, non-AMI for maternal reason and non-AMI for a
reason relating to the child.
Absolute Maternal Indications according to type of area (urban / rural)




                                          25

								
To top