Cross-National Comparison of Primary
Care Practice in Australia, New Zealand,
and the United States
Christopher B. Forrest, MD, PhD
Associate Professor
Johns Hopkins Bloomberg School of Public Health
Department of Health Policy and Management
624 N. Broadway, Room 689
Baltimore, MD 21205
United States
Voice: +1 (410) 614-1932
E-mail: cforrest@jhsph.edu
Johns Hopkins Bloomberg
School of Public Health 1
Collaborators and Sponsor
• Australia: Helena Britt, University of Sydney
• Great Britain: Azeem Majeed, Imperial College
• New Zealand: Peter Crampton, Wellington
School of Medicine and Health Sciences
• United States: Andrew Bindman, UCSF
• Sponsor: Commonwealth Fund
Why Do Cross-National Studies of
Primary Care Practice?
- Dearth of X-national research on
primary care.
- Primary care challenges are similar
across developed nations.
- Responses vary dramatically. Outcomes
- Research can inform primary care
reform.
Do the primary care responses—
scope-of-practice and referral--to
Services population needs vary by country?
How does the population ―exposure‖ to primary care
differ?
Needs Do the population-based needs of AU, NZ, and US differ?
Primary Care Surveys
Country Survey No. Primary Care No. Office Year(s) of Data
Practitioners Visits Collection
Australia BEACH 983 GPs 79,790 Apr2001-
Mar2002
New Zealand NatMedCa 246 GPs 10,064 2001
United States NAMCS 334 FPs (43%) 25,838 2000-2002
238 GIMs (30%)
211 GPeds (27%)
783 PCPs (100%)
Each survey used a physician self-administered questionnaire that was completed
after the visit. The questionnaires had similar content. Analysis was restricted to
visits that occurred in office practice settings.
Population Needs
• Age—next slide
• Sex: 56-57% female in 3 countries
• Burden of new patients: 7-9% in 3 countries
– Each questionnaire asked if the patient was new to the practice
• Health problems managed: reflect need, but also affected by
diagnostic coding, and scope-of-practice
– Diagnostic codes were assigned to the Johns Hopkins Expanded
Diagnostic Clusters (EDCs); EDCs are organized in 27 Major EDC
categories
Diagnostic Code to EDC Match Rates
• AU: ICPC codes EDCs (100% match)
• NZ: Read codes EDCs (95% match)
• US: ICD-9-CM codes EDCs (100% match)
Patient Age Distribution for Primary Care Visits in
Australia, New Zealand, and United States
16
14
% Visit Sample
12
10
8
6
Australia
4 New Zealand
2 United States
0
0-4 5-14 15-24 25-34 35-44 45-54 55-64 65-74 75+
Age, years
Selected Age-Adjusted Major EDC Category
Visit Rates per 1,000 Visits
Health Problem
Category Australia New Zealand United States
Allergy/Immunology 47 65 60
Cardiovascular 148 129 232
Ears, Nose, Throat 134 165 128
Female Reproductive 61 75 32
General Signs and 32 36 48
Symptoms
General Surgery 44 61 45
Malignancy 15 14 12
Musculoskeletal 151 127 142
Neurologic 63 62 67
Skin 114 138 83
Selected Cardiovascular-related EDC Visit
Rates per 1,000 Visits
EDC Australia New Zealand United States
Obesity 11 10 18
Hypertension 95 74 135
Hyperlipidemia 32 10 59
Ischemic heart 12 26 24
disease
Congestive heart 6 9 10
failure
Diabetes 31 35 64
(Types 1 and 2)
Chest pain 3 7 10
Mean Number of Unique Problems Managed
(EDCs) per Visit in Australia, New Zealand, and
United States
2.5
2
Mean EDCs/Visit
1.5 AU
NZ
1 US
0.5
0
Overall 0-17 18-64 65+
Patient Age
Correlation of MEDC-Specific Visit Ratios per
1,000 Visits (n=27) between the Three
Countries
• Australia v New Zealand: 0.93
• New Zealand v United States: 0.78
• Australia v United States: 0.92
Primary Care Exposure
• Visits Rates Data Source--
– AU: Medicare Benefits data provided by the
Australian Department of Health and Ageing
– NZ: Data used to calculate national capitation
funding formula
– US: Combined NAMCS, NHAMCS, and CHC
Visit Survey Data combined with census
• Visit Duration
– Each questionnaire asked physician to record
duration of patient visit
• PC Exposure = Visits/person*Minutes/visit
Mean Number Annual Primary Care Visits per Person in
Australia, New Zealand, and the United States
12
Australia
10 NZ
US
8
Visits/yr
6
4
2
0
Total Male Female 0-17 18-64 65+
Patient Sex Patient Age
Mean Visit Duration per Primary Care Visit in Australia,
New Zealand, and the United States
20 Australia NZ US
18
Mean Visit Duration
16
14
12
10
8
6
4
2
0
Total Male Female 0-17 18-64 65+
Patient Sex Patient Age
Annual Primary Care Exposure per Person in Australia,
New Zealand, and the United States
160
Minutes with a PCP/Person
Australia
140
NZ
120
US
100
80
60
40
20
0
Total Male Female 0-17 18-64 65+
Patient Sex Patient Age
Scope of Practice (EDC_75)
EDC_75
With Without
Country Administrative Administrative
Problems Problems
Australia 50 52
New Zealand 55 54
United States 43 47
Definition: The EDC_75 is the minimum number of EDCs that account for 75%
of all visits in the study sample. Higher values suggest broader scope of practice.
Percentage of Visits Referred to Physician Specialists by Age-
Sex Category in Australia, New Zealand, and United States
Australia NZ US
12
10
8
Visits/yr
6
4
2
0
Age- 0-17 Male 0-17 Female 18-64 Male 18-64 65+ Male 65+ Female
Adjusted Female
Overall
Logistic Regression Analyses
Impact of Selected Morbidities (Major EDCs) on Chances of
Specialty Referral During a Primary Care Visit
Age-sex adjusted Odds Ratios
3.5
Australia New Zealand United States
3
2.5
2 No impact on
referral chances
1.5
1
0.5
0
gy
y
c
e
l
in
ry
T
VS
ta
c
tiv
i
EN
og
Sk
e
r
an
le
C
lle
rg
c
ke
l
n
du
ro
Su
A
ig
s
eu
ro
lo
al
en
p
N
cu
M
Re
G
us
e
M
al
m
Fe
Summary of Findings
• Commonalities are greater than differences.
• Morbidities are generally managed at similar
rates in primary care. The high levels of CVS
care in the US stands out as an exception.
• Scope-of-practice is lowest in US.
• Specialty referral rate is highest in NZ.
• Impact of morbidity on specialty referral similar
between US and AU. Referral patterns in NZ
are less influenced by mix of morbidity.
• Primary care exposure in the US is less than
50% that of AU and NZ.
Conclusions:
Primary Care Monitoring
1) Need common coding system (ICPC?) or
common morbidity group system (EDCs?) for
cross-national studies on ambulatory care.
2) Develop OECD-type measures for primary
care
• Primary care visit rates per population
• Primary care exposure per population
• Scope-of-practice: EDC_75
• Treated morbidity ratios
• Specialty referral rates
• Move toward person-oriented records
Conclusions:
United States
3) US healthcare systems must strengthen
their primary care sub-systems to
increase the population’s exposure to
primary care.
• Primary care is associated with improved health
• Primary care is associated with lower costs
• Stronger primary care should be associated
with better value.
Conclusions:
New Zealand
4) High referral rates despite broad scope-of-
practice suggests more work needs to be done
on understanding the determinants of NZ
general practitioner referral thresholds.
5) High proportion of visits for young children
suggests that lowering financial barriers to
primary care for other age groups will increase
the population’s demand for primary care.
Conclusions:
Australia
6) The BEACH annual survey of general
practice is an excellent primary care
monitoring system.
7) Australia should consider developing
patient-level records for assessing
primary care.