Please cite this paper as

:
Devaux, M. and M. de Looper (2012), “Income-Related
Inequalities in Health Service Utilisation in 19 OECD
Countries, 2008-2009”, OECD Health Working Papers,
No. 58, OECD Publishing.
http://dx.doi.org/10.1787/5k95xd6stnxt-en
OECD Health Working Papers No. 58
Income-Related Inequalities
in Health Service Utilisation
in 19 OECD Countries,
2008-2009
Marion Devaux, Michael de Looper
JEL Classification: I14, I18


































































Unclassified DELSA/HEA/WD/HWP(2012)1

Organisation de Coopération et de Développement Économiques

Organisation for Economic Co-operation and Development
10-Jul-2012
___________________________________________________________________________________________
English text only
DIRECTORATE FOR EMPLOYMENT, LABOUR AND SOCIAL AFFAIRS
HEALTH COMMITTEE



Health Working Papers
OECD Health Working Paper No. 58

INCOME-RELATED INEQUALITIES IN HEALTH SERVICE UTILISATION IN 19 OECD
COUNTRIES, 2008-09

Marion Devaux and Michael de Looper


Key words: Inequality, Health care, Private health insurance, Out-of-pocket payments

JEL classification: I14, I18


All Health Working Papers are now available through the OECD's Internet Website at
http://www.oecd.org/els/health/workingpapers

JT03324670
Complete document available on OLIS in its original format
This document and any map included herein are without prejudice to the status of or sovereignty over any territory, to the delimitation of
international frontiers and boundaries and to the name of any territory, city or area.

D
E
L
S
A
/
H
E
A
/
W
D
/
H
W
P
(
2
0
1
2
)
1

U
n
c
l
a
s
s
i
f
i
e
d

E
n
g
l
i
s
h

t
e
x
t

o
n
l
y



DELSA/HEA/WD/HWP(2012)1
2

DIRECTORATE FOR EMPLOYMENT, LABOUR AND SOCIAL AFFAIRS

www.oecd.org/els


OECD HEALTH WORKING PAPERS

http://www.oecd.org/els/health/workingpapers


This series is designed to make available to a wider readership health studies prepared for use within the
OECD. Authorship is usually collective, but principal writers are named. The papers are generally
available only in their original language – English or French – with a summary in the other.

Comment on the series is welcome, and should be sent to the Directorate for Employment, Labour and
Social Affairs, 2, rue André-Pascal, 75775 PARIS CEDEX 16, France.

The opinions expressed and arguments employed here are the responsibility
of the author(s) and do not necessarily reflect those of the OECD.



Applications for permission to reproduce or translate
all or part of this material should be made to:

Head of Publications Service
OECD
2, rue André-Pascal
75775 Paris, CEDEX 16
France

Copyright OECD 2012

DELSA/HEA/WD/HWP(2012)1
3

ACKNOWLEDGEMENTS
1. The authors thank Albane Gourdol from Eurostat for her assistance with the European Health
Interview Survey data analysis, Lien Nguyen from the Finnish National Institute for Health and Welfare for
carrying out the analysis for Finland, Ola Ekholm from the Danish National Institute of Public Health for
assisting in data analysis, and Hideki Hashimoto for his help on Japan. The authors are very grateful to all
for their collaboration in this project.
2. The authors also thank national data providers for supplying individual-level datasets, namely,
Statistics Canada for the Canadian Community Health Survey 2007/08, The Danish National Institute of
Public Health for the National Health Interview Survey 2005, the French Institute for Research and
Information in Health Economics for the Santé et Protection Sociale survey 2008, the Robert Koch
Institute for the Gesundheit in Deutschland aktuell 2009, the Irish Social Science Data Archive for the
Survey of the Lifestyle, Attitudes and Nutrition 2007, the New Zealand Ministry of Health for the National
Health Survey 2006/07, the Spanish National Statistics Institute for the Encuesta Europea de Salud 2009,
the Swiss Federal Statistical Office for the Swiss Health Survey 2007, the Institute for Social and Economic
Research for the British Household Panel Survey 2009, the American Agency for Healthcare Research and
Quality for the Medical Expenditure Panel Survey 2009, and Eurostat for the European Health Interview
Survey. All computations on these micro-data were prepared by the authors, and the responsibility for the
use and interpretation of these data is entirely that of the authors.
3. Finally, the authors also thank Rie Fujisawa, Gaetan Lafortune, Valérie Paris, Mark Pearson, and
Divya Srivastava from the OECD Health Division for their comments and suggestions.
DELSA/HEA/WD/HWP(2012)1
4

ABSTRACT
4. This Working Paper examines income-related inequalities in health care service utilisation in
OECD countries. It extends a previous analysis (Van Doorslaer and Masseria, 2004) to 2008-2009 for 13
countries, and adds new results for 6 countries, for doctor and dentist visits, and cancer screening. Quintile
distributions and concentration indices were used to assess inequalities. For doctor visits, horizontal equity
was assessed, i.e. the extent to which adults in equal need of physician care appear to have equal rates of
utilisation. The paper considers the evolution of inequalities over time by comparing results with the
previous study, as data permit. Health system financing arrangements are examined to see how these might
affect inequalities in health service use.
5. Findings show that, for doctor visits, horizontal inequities in health care utilisation persist across
OECD countries. After adjustment for needs for health care, the better-off are more likely to visit doctors,
especially specialists. With GP contacts, the scenario is different. In most countries, for the same level of
need for health care, the worse-off are as likely as the better-off to contact a GP, and they visit more often.
Inequalities in dental visits and breast and cervical cancer screening appear in numerous countries, with the
better-off making more use of services. The relative position of countries has remained stable for doctor
and GP visits over the two studies. Some discrepancies are found in country ranks for specialist and dentist
visits, but these are attributed to methodological differences.
6. Findings highlight the important effect that the financing of health care services can have on
equity (public and private health insurance coverage, and the share of out-of-pocket payments for different
services), although some of the inequalities in health service use cannot be explained by financial barriers.
DELSA/HEA/WD/HWP(2012)1
5

RESUME
7. Ce document de travail examine les inégalités liées aux revenus dans l’utilisation des services de
santé dans les pays de l’OCDE. Il met à jour une étude précédente (Van Doorslaer and Masseria, 2004)
pour 13 pays, et inclut 6 nouveaux pays, utilisant des données de 2008-2009, portant sur les consultations
de médecins et dentistes, et le dépistage du cancer. Les inégalités sont mesurées à l’aide de distributions
par quintile et d’indices de concentration. Cette étude s’intéresse à l’équité horizontale pour les
consultations de médecins, i.e. dans quelle mesure des adultes ayant un besoin égal de soins médicaux ont
apparemment des taux identiques d’utilisation de soins. Elle examine l’évolution des inégalités en
comparant les résultats avec l’étude précédente lorsque les données le permettent. Le cadre d’analyse
s’intéresse aux caractéristiques de financement des systèmes de santé et à leurs possibles influences sur les
inégalités d’utilisation des services de santé.
8. Les résultats montrent que pour les consultations de médecins, les iniquités horizontales dans
l’utilisation des soins de santé persistent dans les pays de l’OCDE. Après ajustement par les besoins en
soins de santé, les personnes à hauts revenus ont plus de chances de consulter un médecin et notamment un
spécialiste. Le scénario est différent pour les visites de généralistes. Dans la plupart des pays, pour un
même niveau de besoins de soins de santé, les plus démunis ont autant de chances que les riches de
consulter un généraliste, et ils consultent plus fréquemment. Des inégalités dans les consultations de
dentistes et le dépistage des cancers du sein et du col de l’utérus sont apparentes dans plusieurs pays,
favorisant les personnes à hauts revenus. La comparaison avec l’étude précédente montre que la position
relative des pays est restée stable pour les consultations de docteurs et de généralistes. On trouve des
différences dans la position des pays pour les consultations de spécialistes et dentistes, notamment dues à
aux différences de méthodologies d’enquêtes et de questions.
9. Les résultats soulignent l’importance des effets du financement des services des soins de santé
sur l’équité (assurance santé publique et privée, et part des paiements directs à la charge des patients pour
divers services), bien que certaines des inégalités dans l’utilisation des services de santé ne puissent pas
être expliquées par des barrières financières.
DELSA/HEA/WD/HWP(2012)1
6

TABLE OF CONTENTS
ACKNOWLEDGEMENTS ............................................................................................................................ 3
ABSTRACT .................................................................................................................................................... 4
RESUME ........................................................................................................................................................ 5
1. INTRODUCTION ...................................................................................................................................... 9
2. DEFINING, MEASURING AND INTERPRETING HORIZONTAL INEQUITY ................................ 11
3. METHODS ............................................................................................................................................... 13
3.1 Surveys used ................................................................................................................................... 13
3.2 Estimating health care utilisation ................................................................................................... 14
3.3 Adjusting for health care need ........................................................................................................ 14
3.4 Estimating income .......................................................................................................................... 15
3.5 Other explanatory variables ............................................................................................................ 15
4. MEASURING INEQUALITIES IN HEALTH CARE UTILISATION: RESULTS ............................... 17
4.1 Doctor visits.................................................................................................................................... 17
4.2 GP visits ......................................................................................................................................... 18
4.3 Specialist visits ............................................................................................................................... 20
4.4 Dentist visits ................................................................................................................................... 22
4.5 Cancer Screening ............................................................................................................................ 24
4.6 Comparison with earlier results ...................................................................................................... 25
4.7 Country overview ........................................................................................................................... 26
5. INEQUALITIES IN HEALTH CARE UTILISATION RELATED TO HEALTH SYSTEM
FINANCING ................................................................................................................................................ 28
5.1 Public health settings ...................................................................................................................... 28
5.2 Out-of-pocket payments ................................................................................................................. 30
5.3 Private health insurance .................................................................................................................. 32
6. CONCLUSION ......................................................................................................................................... 36
REFERENCES ............................................................................................................................................. 37
ANNEXES .................................................................................................................................................... 40
ANNEX 1. SET OF AVAILABLE INFORMATION IN EACH SURVEY ................................................ 40
ANNEX 2. METHODOLOGICAL ANNEX ............................................................................................... 45
ANNEX 3. QUINTILE DISTRIBUTION OF HEALTH CARE UTILISATION ....................................... 47
ANNEX 4. DETAILS ON HEALTH SYSTEMS CHARACTERISTICS ................................................... 53
Depth of coverage .................................................................................................................................. 53
Degree of private provision in physicians’ services .............................................................................. 55
Regulation of prices billed by providers ................................................................................................ 56

DELSA/HEA/WD/HWP(2012)1
7


Tables
Table 1. Data sources .......................................................................................................................... 13
Table 2. Country overview of probability of a doctor or dentist visit or cancer screening, and level of
inequality .............................................................................................................................................. 27
Table A1.1. Dependent variables .............................................................................................................. 41
Table A1.2. Explanatory variables: Need variables and socio-economic characteristics ......................... 43
Table A3.1: Quintile distribution of the probability and number of doctor visits after need-
standardisation, inequality index (CI before need-standardisation) and inequity index (HI after need-
standardisation) ......................................................................................................................................... 47
Table A3.2: Quintile distribution of the probability and number of GP visits after need-standardisation,
inequality index (CI before need-standardisation) and inequity index (HI after need-standardisation) ... 48
Table A3.3: Quintile distribution of the probability and number of specialist visits after need-
standardisation, inequality index (CI before need-standardisation) and inequity index (HI after need-
standardisation) ......................................................................................................................................... 49
Table A3.4: Dentist visits in the past 12 months, by income quintile, and inequality index (CI) ............. 50
Table A3.5: Breast cancer screening participation in the past 2 years in women aged 50-69, by income
quintile, and inequality index (CI) ............................................................................................................ 51
Table A3.6: Cervical cancer screening participation in the past 3 years in women aged 20-69, by income
quintile, and inequality index (CI) ............................................................................................................ 52
Table A4.1: Scoring of depth of coverage................................................................................................. 54
Table A4.2: Scoring the degree of private provision of physicians’ services ........................................... 56



Figures
Figure 1: Needs-adjusted probability of a doctor visit in last 12 months, by income quintile, 2009 (or
latest year) ................................................................................................................................................. 17
Figure 2: Inequity index for doctor visits in the past 12 months, adjusted for need, 2009 (or latest year) 18
Figure 3: Needs-adjusted probability of a GP visit in last 12 months, by income quintile, 2009 (or latest
year) ........................................................................................................................................................... 19
Figure 4: Inequity index for GP visits in the past 12 months, adjusted for need, 2009 (or latest year) .... 19
Figure 5: Needs-adjusted probability of a specialist visit in last 12 months, by income quintile, 2009 (or
latest year) ................................................................................................................................................. 20
Figure 6: Inequity index for specialist visits in the past 12 months, 2009 (or latest year) ........................ 21
Figure 7: Probability of a dentist visit in last 12 months, by income quintile, 2009 (or latest year) ......... 22
Figure 8: Inequality index for dentist visits in the past 12 months, 2009 (or latest year) ......................... 23
Figure 9: Probability of breast cancer screening in the last two years, women aged 50-69 years, by
income quintile, 2009 (or latest year) ........................................................................................................ 24
Figure 10: Probability of cervical cancer screening in the last three years, women aged 20-69 years, by
income quintile, 2009 (or latest year) ........................................................................................................ 24
Figure 11: Inequality index for breast cancer screening in the past 2 years, 2009 (or latest year)............ 25
Figure 12: Inequality index for cervical cancer screening in the past 3 years, 2009 (or latest year) ........ 25
Figure 13: Comparison of inequity indexes, 2000 and 2009 ..................................................................... 26
Figure 14: Relationship between inequity in doctor visits and the share of public health expenditure .... 29
Figure 15: Relationship between inequity in specialist visits and the degree of private provision of
services ...................................................................................................................................................... 29
Figure 16: Relationship between inequity and the share of out-of-pocket payments ................................ 31
Figure 17: Share of PHI by income quintile, selected OECD countries, 2009 (or latest year) ................. 33
DELSA/HEA/WD/HWP(2012)1
8

Figure 18: Adjusted probabilities of medical visits by PHI status, in selected OECD countries .............. 34
Figure A4.1: Relationship between inequity and depth of coverage ......................................................... 55
Figure A4.2: Relationship between inequity and regulation of prices billed by physicians ..................... 57



DELSA/HEA/WD/HWP(2012)1
9


1. INTRODUCTION
10. The economic crisis has led to increased concerns about access to health care. Several OECD
countries have introduced cuts or slow-downs in public health spending, and to contain costs, increased co-
payments or adjustments to the goods and services available through health care coverage have been made.
The October 2011 WHO-sponsored World Conference on Social Determinants of Health, however, saw
heads of government renew their determination to achieve health equity, through promotion of
accessibility and affordability through universal health care coverage, and health financing that prevents
impoverishment (WHO, 2012).
11. Effective health care coverage provides financial security against expenses due to unexpected or
serious illness, and promotes access to medical goods and services. Most OECD countries have achieved
universal or near-universal coverage of their populations for a core set of health care services. Access to
physician services is ensured at relatively low or no cost for patients. Other services such as dental care and
pharmaceutical drugs are often partially covered, although there are a number of countries where coverage
for these services must be purchased separately (OECD, 2011a). Preventive screening services for certain
cancers such as breast and cervical cancer are generally also available at little or no cost.
12. This coverage results from a variety of public insurance arrangements which aim at providing
equitable access to care. Equity of access is a key element of health system performance in OECD
countries (Hurst and Jee-Hughes, 2001; Kelley and Hurst, 2006). Determining the extent to which OECD
countries have achieved the goal of providing equal access for equal need – or equal utilisation, as an
indicator of access –regardless of individual characteristics such as income, place of residence, occupation
or educational level, has been a focus of previous cross-country comparative work (Van Doorslaer et al.,
2002; Van Doorslaer and Masseria, 2004). This previous work, as well as the current paper, focuses on
equity of access as stated in the principle of horizontal equity, i.e. that people in equal need of health care
should be treated equally, and tests whether there are any deviations from this principle, based on the
income level of the family in which the individual lives.
13. There is already a substantial body of evidence that inequities exist in the use of certain health
care services in many countries. Analysing national health interview surveys, Van Doorslaer and Masseria
(2004) established that, based on individual’s need for care, deviations from the principle of horizontal
inequity existed in a number of OECD countries around the year 2000, with the better-off more likely to
see a medical specialist, and often visiting these specialists more frequently. Dental care was also used
more intensively by the better-off. Other studies using European data also find inequities in health care use.
Or et al., (2008), also using data from around the year 2000, found education-related inequities for
specialist visits in almost all countries studied, whereas the picture was less clear-cut for general
practitioner (GP) visits. Research on European panel data found pro-poor inequity in GP visits in most EU
countries (7 out of 10 countries studied) and pro-rich inequity in specialists in all countries studied (Bago
d’Uva et al., 2008).
14. This paper seeks to update and extend these findings, using the same methodology as earlier
work by Van Doorslaer and colleagues. It analyses inequities in access to health care through measuring
the utilisation of health care services by income level, adjusted for need. This paper extends the analysis to
DELSA/HEA/WD/HWP(2012)1
10

new countries and examines the evolution of inequities over time by comparing results with previous
findings as data permit.
15. The set of countries studied represents a relatively wide geographical spread, as well as a varied
selection of health care systems. The current study covers 19 OECD countries: Austria, Belgium, Canada,
Czech Republic, Denmark, Estonia, Finland, France, Germany, Hungary, Ireland, New Zealand, Poland,
Slovak Republic, Slovenia, Spain, Switzerland, the United Kingdom and the United States. Results for 13
countries are updated to 2008-2009 or the nearest available year, and new results are added for six
countries (Czech Republic, Estonia, New Zealand, Poland, Slovak Republic and Slovenia).
16. The analysis also broadens the selection of health care services covered in previous studies, by
examining whether differences exist in preventive screening for female breast and cervical cancers.
17. A second objective of the paper is to analyse any findings of inequities in health care use by
using selected characteristics of health systems, and to investigate whether these inequities might result
from the ways in which health care services are financed. To do this, a number of other data collections are
utilised, including the OECD Health Data database, and the 2008 OECD Survey of Health System
Characteristics (Paris et al., 2010).
DELSA/HEA/WD/HWP(2012)1
11

2. DEFINING, MEASURING AND INTERPRETING HORIZONTAL INEQUITY
18. All OECD countries aspire to achieve adequate access to health care for all people on the basis of
need. Many countries explicitly endorse equality of access to health care for all people as an objective in
their policy documents (Hurst and Jee-Hughes, 2001; Huber et al., 2008). Most OECD countries offer
options to add to the general public coverage of health costs through complementary or supplementary
private coverage, as is often the case for dental health care, with this additional cover being the sole source
for sizeable shares of the care package.
19. Studies of health care equity examine whether certain population groups systematically receive
different levels of care. Inequality and inequity of care are two key concepts. Here, inequality in health care
utilisation refers to the differences in use that can be observed between individuals or population groups,
whereas inequity refers to those differences remaining after adjustment for need for health care. In this
study, adjustment for need has been made for doctor visits only, with equal need assumed for dental visits
and cancer screening.
20. Horizontal equity is the principle that requires that people in equal need of health care are treated
equally, irrespective of individual characteristics such as income, place of residence or race. A second
principle, vertical equity, means appropriately treating those people who have differing needs for care. It is
this first principle of horizontal equity that the present study uses as a basis for international comparisons.
21. The method used in this paper to describe and measure the degree of horizontal inequity in health
care delivery is identical to that used in Van Doorslaer et al. (2002), and Van Doorslaer and Masseria
(2004). It compares the observed distribution of health care by income with the distribution of need. To
statistically equalize needs for the groups or individuals to be compared, this study uses the average
relationship between need and treatment for the population as a whole as the vertical equity “norm” and
investigates the extent of systematic deviations from this norm by income level.
22. The inequality measure adopted in this paper is the concentration index (CI). The CI quantifies
the degree of socioeconomic-related inequality of actual medical care use, taking full advantage of the
information on all individuals. Unlike studies which only examine lowest and highest quintiles, the CI
reflects the experiences of the entire population, and it is sensitive to changes in distribution of the
population across socioeconomic groups (Wagstaff et al., 1991). The CI is bounded between -1 and 1, with
the sign indicating the direction of inequality - a positive index indicates pro-rich inequality (i.e. a
distribution that favours the rich), a zero value indicates equality, and a negative index indicates pro-poor
inequality (i.e. that favours the poor).
23. The degree of horizontal inequity is measured by the concentration index of the needs-
standardised use, and this is here termed the horizontal inequity index, or HI. When the HI is not
significantly different from zero, it indicates equity. When it is positive, it indicates pro-rich inequity, and
when it is negative, it indicates pro-poor inequity. Horizontal inequity indexes were calculated for doctor
visits only.
24. Interpreting values of the HI is not intuitive. Koolman and van Doorslaer (2004) have shown that
multiplying the value of the HI by 75 gives the proportion of the health variable that would need to be
(linearly) redistributed from the richer half to the poorer half of the population (assuming that health
DELSA/HEA/WD/HWP(2012)1
12

inequity favours the rich) to arrive at equity, i.e. a distribution with an index value of zero. For example, if
pro-rich inequity in doctor visits existed with a HI = 0.05, equalizing doctor visits across the income
distribution requires redistributing 3.75% of all visits made by richer people, increasing visits by poorer
people by the same amount.
DELSA/HEA/WD/HWP(2012)1
13

3. METHODS
3.1 Surveys used
25. Data on health care use were taken from national health interview surveys (see Table 1). For
most European countries, these came from questions proposed in the European Health Interview Survey
(EHIS), which was implemented across countries between 2006 and 2009. EHIS consists of four modules
of questions on health status, health care use, health determinants, and background variables. These
modules may be implemented at a national level either as one specific survey or as elements of existing
surveys (i.e. national health interview surveys or other household surveys). Eight countries have adopted
the same question modules on health care use (Austria, Belgium, Czech Republic, Estonia, Hungary,
Poland, Slovak Republic and Slovenia, called hereafter the EHIS countries).
26. For non-EHIS countries, results from the most recent national health surveys gathering
information on health care visits and socio-economic characteristics were used, as listed in Table 1. A
detailed list of available information in each survey can be consulted in Annex 1.
Table 1. Data sources


Country Survey data
Austria Österreichische Gesundheitsbefragung 2006/07 (EHIS 2006/07)
Belgium Belgium Health Survey 2008 (EHIS 2008)
Canada Canadian Community Health Survey 2007/08
Czech Republic European Health Interview Survey in the Czech Republic 2008 (EHIS 2008)
Denmark National Health Interview Survey 2005
Estonia Estonian Health Interview Survey 2006/07 (EHIS 2006/07)
Finland Welfare and services Survey (HYPA -survey) 2009
France Enquête Santé Protection Sociale 2008
Germany German Telephone Health Interview Survey (GEDA) 2009
Hungary European Health Interview Survey 2009 (EHIS 2009)
Ireland Survey of Lifestyle, Attitudes and Nutrition in Ireland (SLAN 2007)
New Zealand National Health Survey 2006-07
Poland Europejskie Ankietowe Badanie Zdrowia 2009 (EHIS 2009)
Slovak Republic Európsky prieskum zdravia 2009 (EHIS 2009)
Slovenia Anketa o zdravju in zdravstvenem varstvu 2007 (EHIS 2007)
Spain Encuesta Europea de Salud 2009
Switzerland Swiss Health Survey 2007
United Kingdom British Household Panel Survey 2009
United States Medical Expenditure Panel Survey 2008
DELSA/HEA/WD/HWP(2012)1
14

3.2 Estimating health care utilisation
27. Three types of health care services were measured, these being the utilisation of (i) physicians
(and in most countries further information on separate GP and specialist utilisation are also available); (ii)
dentists; and (iii) breast and cervical cancer screening services for women.
28. Both the probability and the frequency of services were measured using sampling weights.
Typically, a first survey question measures the probability of a visit, and is of the form: “In the past 12
months have you visited a GP?”. A second question measures the frequency of visits in the past 12 months.
29. Denmark is an exception since doctor visits were recorded over the past 3 months. Dentist
consultations were measured over the past 12 months with the exception of France (past 2 years) and
Denmark (past 3 months). For the number of visits to doctors and dentists, the recall period was also
generally the past 12 months. However, in EHIS countries, individuals were asked how many visits they
had made in the past 4 weeks
1
. This difference in time scales does not allow for cross-country comparison
of the frequency of visits, but inequalities in the probability of a visit can be still measured.
30. Some caution is needed in comparing the magnitude of inequalities across countries and over
time because of these discrepancies in recall periods. Based on previous findings, it is likely that the
probability of the well-off seeing a GP in the past three months is lower than the probability that the worse-
off had seen a GP, since the worse-off visit GPs more often. Thus, measuring inequities in visits over the
past 3 months in Denmark would likely over-estimate pro-poor inequities.
31. National guidelines relating to cancer screening may also differ across countries (OECD, 2011b),
affecting the inclusion age and frequency. To perform international comparisons, the same age range and
frequency was adopted as that used for the OECD Health Data collection. For breast cancer screening, the
focus was on women aged 50-69 years who reported having a mammogram in the past 2 years, and for
cervical cancer screening, women aged 20-69 years who had a Pap smear in the past 3 years. However, the
recall period refers to the past 12 months in Denmark (for breast cancer screening) and in Ireland (for both
indicators).
32. Free nationwide population-based screening mammography programmes operate in many
countries, including (among the 19 studied countries) Belgium, Finland, France, Germany, Ireland, New
Zealand, Spain and the UK
2
. Pap smear tests are available through free nationwide population-based
programmes in Denmark, Finland, Hungary, Ireland, New Zealand, Slovenia and the United Kingdom (but
only England, Scotland and Wales) (OECD, 2011b).
3.3 Adjusting for health care need
33. Since persons in lower socioeconomic groups have higher rates of morbidity, they have greater
needs for health care services (Mackenbach, 2006; Mackenbach et al., 2008; de Looper and Lafortune,
2009; OECD, 2011c). GP and specialist visits thus need to be standardised to remove the effect of differing
needs for care among persons with different income levels, so that the horizontal equity principle can be
tested.

1
It is not possible to extrapolate because of the different time scales (number of visits in the past 4 weeks
versus probability of visiting in the past 12 months).
2
These are the countries offering nationwide population-based free-access screening. There are also other
countries which have national population-based free-access screening but rollout is not nationwide yet.
They include Denmark and Switzerland.
DELSA/HEA/WD/HWP(2012)1
15

34. In this study, the need for doctor visits is proxied by age, gender, and health status variables.
Health status was measured by two variables: self-assessed health and activity limitation. Self-assessed
health is most often based on a five-response variable in which individuals rate their own general health
status as “Very good, Good, Fair, Bad, or Very bad”. In Canada, Ireland, New Zealand, the United
Kingdom and the United States the response options differ slightly: “Excellent, Very Good, Good, Fair,
Poor”. The activity limitation indicator identifies whether individuals are “Limited, Severely limited or Not
limited” in their daily activities. Again, some countries differ in their response scales (see Annex 1).
Although the appropriateness of the measure of health care needs may be questioned, it is worth noting that
self-assessed health is widely regarded as a good predictor of both health care utilisation (DeSalvo et al.,
2005) and mortality (Idler and Benyamini, 1997). Chronic conditions as reported in national health surveys
was not used, since there is high heterogeneity in individual responses, and this may bias the measurement
of socio-economic inequalities in health (Tubeuf et al., 2008).
35. To perform the adjustment for health care needs, an indirect standardisation method was applied
(see Annex 2). Linear regression models were used first, followed by a two-stage model with logistic
regressions for binary outcomes (the probability of a consultation) and 0-truncated negative binomial
models for count variables (the frequency of consultations). However, the estimated rates by income level
and the concentration indexes change only slightly, if at all, between the two methods. Van Doorslaer and
Masseria’s (2004) results from linear and non-linear methods also differ little. Further, Wagstaff and Van
Doorslaer (2000) showed that HI indices are insensitive to the use of non-linear methods rather than least
squares. To simplify comparison with the previous study, only the linear regression estimates are
displayed.
36. Needs standardisation was not performed for dental consultations. In most countries, an annual
dental visit is recommended for all persons. Also, most data sources do not provide information on dental
care needs among OECD countries, this being only available for Canada and France.
37. Neither is needs standardisation performed for breast and cervical cancer screening, with equal
need across all income levels assumed. Again, health authorities recommend periodic screening for all
women in the target age groups. In practice, many countries make these services available for free.
38. In this paper, equity is determined for doctor visits, using the horizontal inequity index HI, which
is adjusted for need. Equality is determined for dentist visits and cancer screening using the concentration
index CI, which is not adjusted for need.
3.4 Estimating income
39. Socio-economic status was measured by household income, equivalised using the OECD-
modified equivalence scale (Hagenaars et al., 1994). For the purpose of the analysis, income was divided
into quintiles. In Canada, only a categorical variable of equivalised income was available. For some
countries, it was not possible to calculate the equivalised income using the OECD modified equivalence
scale, and so household income categories were used for Denmark, New Zealand, and the EHIS countries.
3.5 Other explanatory variables
40. The need-standardisation procedure also uses other explanatory variables, consistent with the
previous study, and for which we do not want to standardise, but to control for, in order to estimate partial
correlations with the need variables. Demographic and socio-economic characteristics are included, such as
ethnicity (available for a few countries only), the size of the household, and marital status. Education level
was categorised into three groups: Low / Medium / High. The categorical variable for activity status
distinguishes Employee / Self-employed / Student / Unemployed / Retired / Homemakers / Others
DELSA/HEA/WD/HWP(2012)1
16

(disabled, military). An indicator to identify whether people have public or private health insurance
coverage is used when available. The region variable is based on NUTS-2 for most countries. The level of
urbanisation of the living area distinguishes between three degrees: Dense / Intermediate / Thin. The
availability of these variables is described in Annex 1.
DELSA/HEA/WD/HWP(2012)1
17

4. MEASURING INEQUALITIES IN HEALTH CARE UTILISATION: RESULTS
41. Inequities in doctor visits (GP and specialists together) are presented first, and then for GP and
specialist visits for countries where the data permit this distinction. Inequalities in dental visits and cancer
screening follow. The first chart summarises the distribution of health care use by lowest and highest
income quintiles, and the average across the population. In the case of doctor visits, the distribution is
need-adjusted. A second chart shows the horizontal inequity index (HI) for both the probability and
frequency of visits. Detailed results on the need-standardised distribution of doctor visits (and the split by
GP and specialist) across all income quintiles is available in Annex 3.
4.1 Doctor visits
42. The distribution of visits to doctors in the last 12 months across income quintiles, adjusted for
need, is shown in Figure 1. The average rates among the total adult population of visiting a doctor in the
past 12 months vary across countries, from 68% in the United States to 91% in France (rates in Denmark
relate to the past three months). Rates of doctor visits also vary by income level. In most countries, for the
same level of need, high-income people are more likely to visit a doctor than low-income people.
However, in the Czech Republic and the United Kingdom, the gradient is not clear-cut, and it appears to be
reversed in Denmark. The surveys in Austria and Ireland do not include a question on doctor visits (only
on GP visits), and so these countries are not presented in Figure 1
3
.
Figure 1: Needs-adjusted probability of a doctor visit in last 12 months, by income quintile, 2009 (or latest
year)

Note: (*) in the past 3 months in Denmark.

3
In addition to these selected OECD countries, similar studies on Asian countries were undertaken although
methodology and data comparability differ. Results show income-related equity for doctor visits in Korea
with HI=-0.009 (Lu et al., 2007), and pro-rich inequity in Japan with HI=0.0135 (Ikegami et al., 2011).
France
Germany
Belgium
Canada
Czech Republic
Hungary
Spain
Slovak Republic
New Zealand
United Kingdom
Switzerland
Slovenia
Poland
Estonia
Finland
United States
Denmark*
0.40 0.50 0.60 0.70 0.80 0.90 1.00
Rates of doctor visits in the past 12 months
Lowest
income
quintile
Average Highest
income
quintile
DELSA/HEA/WD/HWP(2012)1
18

43. Pro-rich inequities in the probability of a doctor visit are observed in most countries, but not at a
high level (Figure 2). Only in the United States was a higher level of inequity apparent; for the same level
of need for health care, people with higher incomes are more likely to visit a doctor than those with lower
incomes.
Figure 2: Inequity index for doctor visits in the past 12 months, adjusted for need, 2009 (or latest year)

Note: (*) visits in the past 3 months in Denmark. (**) counts in the past 4 weeks in EHIS countries.
44. In Figure 2 Panel A, the confidence intervals of the horizontal inequity index for the United
Kingdom, the Czech Republic and Slovenia cross the zero line, meaning that the HI is not statistically
significantly different from 0. In these three OECD countries, given the same need, people with a lower
income were as likely to see a doctor as higher income people. Denmark (for which the recall period may
over-estimate inequities in favour of the worst-off) displays significant pro-poor inequities.
45. The pattern for the frequency of visits is less clear, since most of the inequity indices are not
significantly different from zero. Nonetheless, four OECD countries – Poland, the United States, Spain and
Finland – display a high level of pro-rich inequity in the frequency of doctor’s visits.
46. According to the redistribution interpretation offered by Koolman and van Doorslaer (2004),
equalizing doctor visits across income quintiles would require redistributing less than 3% of all visits from
the richer to the poorer in Finland, and up to 9% in Poland. Since the average number of doctor visits in the
past 4 weeks is about 0.85 in Poland (see Table A3.1 in Annex 3), i.e. 850 visits per 1000 population, this
would require a redistribution of 74 visits per 1000 population from the richest to the poorest half of the
population to achieve equity. The number of visits to be redistributed is related to the average number of
visits. In France, where HI= 0.02 and the average number of doctor visits per person in the past 12 months
was 5.3, i.e. 5300 visits per 1000 population, 87 doctor visits per 1000 population should be redistributed
from the richer to the poorer to achieve equity. In Finland, whereas the inequity index is higher (HI= 0.04),
the average number of doctor visits in the past 12 months is about 3 per person, and thus, the same number
of doctor visits (87) should be redistributed to arrive at equity.
4.2 GP visits
47. The distribution of visits to GPs in the last 12 months across income quintiles, adjusted for need,
is shown in Figure 3. The share of the total adult population seeing a GP in the past 12 months varies
across countries, from 58% in Finland to 86% in France. Variations by income level favouring high-
income groups are apparent in six countries (New Zealand, Canada, Slovak Republic, Poland, Estonia, and
Finland). The gradient appears to favour low-income people in Denmark, Czech Republic and Spain. The
DELSA/HEA/WD/HWP(2012)1
19

surveys in the United States and Germany do not separately identify GP and specialist visits, and so these
countries are excluded.
Figure 3: Needs-adjusted probability of a GP visit in last 12 months, by income quintile, 2009 (or latest year)

Note: (*) visits in the past 3 months in Denmark.
48. Figure 4 shows the inequity indexes for GP consultations. Nine countries show no significant
inequities in the probability of seeing a GP (Czech Republic, Spain, Switzerland, Ireland, Austria,
Belgium, United Kingdom, Hungary, and Slovenia). Seven countries display significant pro-rich inequity
(Estonia, Canada, Finland, Poland, New Zealand, the Slovak Republic, and France) but the degree is quite
small. Only Denmark presents significant pro-poor inequities.
Figure 4: Inequity index for GP visits in the past 12 months, adjusted for need, 2009 (or latest year)

Note: (*) visits in the past 3 months in Denmark. (**) counts in the past 4 weeks in EHIS countries.
49. With respect to the number of visits, the results are quite different, and six countries display
significant pro-poor inequities (Denmark, Belgium, Austria, France, New Zealand, and Canada). For the
same level of health care needs, the poor see a GP more often than the rich. The other nine countries
display no significant inequity.
France
Belgium
New Zealand
Austria
Canada
Slovak Republic
Spain
Hungary
United Kingdom
Ireland
Czech Republic
Poland
Slovenia
Estonia
Switzerland
Finland
Denmark*
0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Rates of GP visits in the past 12 months
Lowest
income
quintile
Average Highest
income
quintile
DELSA/HEA/WD/HWP(2012)1
20

50. Summarising these findings, the most deprived are generally as likely to see a GP in the past 12
months as the rich, but once they engage with a GP, they consult more often.
4.3 Specialist visits
51. The distribution of visits to specialists in the last 12 months across income quintiles, adjusted for
need, is shown in Figure 5. The average share of the total adult population seeing a specialist in the past 12
months varies from 33% in New Zealand to 60% in Hungary. Most countries present large variations in
specialist visit rates across income quintiles, after need-standardisation. In all countries, high-income
people display higher rates of specialist care utilisation. Data from Austria, Germany, Ireland and the
United States do not permit to analyse specialist visits.
Figure 5: Needs-adjusted probability of a specialist visit in last 12 months, by income quintile, 2009 (or latest
year)

Note: (*) visits in the past 3 months in Denmark.
52. Figure 6 shows pro-rich inequities in both the probability and frequency of specialist visits in
almost all countries. The degrees of inequity are much larger than those displayed for GP visits. However,
they are not significant for the probability of visits in the United Kingdom, the Czech Republic and
Slovenia. For the frequency of visits, they are not significant in Slovenia, the Czech Republic, Denmark,
New Zealand, Slovak Republic and Belgium.
53. Hence, after adjusting for need, the well-off are generally more likely to visit a specialist than the
poor (Panel A), and they also do so more often (Panel B).
54. Among the studied countries, France and Spain are the most inequitable in specialist visits for
both probability and frequency concurring with previous findings (Or et al. 2008; Palència et al. 2011). In
Spain, other research has found that inequity in specialist visits are largely related to the private sector,
with the public health system being more equitable (Regidor et al. 2008).
55. Concerning frequency of visits, HIs are significant for eight countries (Poland, Canada, Hungary,
Switzerland, Estonia, Finland, France, and Spain), and they vary substantially (Figure 6 Panel B). The
Hungary
Czech Republic
France
Canada
Slovak Republic
Spain
Switzerland
Belgium
Poland
Estonia
Slovenia
United Kingdom
Finland
New Zealand
Denmark*
0.00 0.20 0.40 0.60 0.80
Rates of specialist visits in the past 12 months
Lowest
income
quintile
Average Highest
income
quintile
DELSA/HEA/WD/HWP(2012)1
21

percentage of specialist visits to be linearly redistributed from the richer to the poorer to achieve equity
varies from 3.4% in Poland to 14.7% in Spain. For France, where this percentage is equal to 8%, the
average number of specialist visits in the past 12 months is about 1.76 (see Table A3.3 in Annex 3), i.e.
1760 visits per 1000 population. For every 1000 population, it would be necessary to redistribute 140 visits
from the richest to the poorest half of the population to achieve equity. The quantity to be redistributed to
achieve equality would be 66 specialist visits in Finland, and 106 in Switzerland.
Figure 6: Inequity index for specialist visits in the past 12 months, 2009 (or latest year)

Note: (*) visits in the past 3 months in Denmark. (**) counts in the past 4 weeks in EHIS countries.
DELSA/HEA/WD/HWP(2012)1
22

4.4 Dentist visits
56. The distribution of visits to dentists in the last 12 months across income quintiles is shown in
Figure 7. The average rates among the total adult population of seeing a dentist in the past 12 months vary
across countries, from 37% in Hungary to 71% in Czech Republic. All countries display large variations in
dentist visit rates across income quintiles, with systematically higher rates among high-income people.
Figure 7: Probability of a dentist visit in last 12 months, by income quintile, 2009 (or latest year)

Note: (*) visits in the past 2 years in France. (**) visits in the past 3 months in Denmark.
57. Figure 8 shows large inequalities in the probability of a dentist consultation in favour of the well-
off. Similarly, large inequalities in the number of dentist visits are apparent, with the exceptions of the
Czech Republic, Switzerland, and Slovenia where the index is not significantly different from zero.
58. The percentage of dentist visits to be linearly redistributed from the richer to the poorer to
achieve equality is 14.7% in the United States and 11.3% in Spain. Equalizing dentist visits across income
quintiles requires redistributing fewer than 10% of all visits in the other countries. In Canada, where the CI
equals 0.106, the average number of dentist visits in the past 12 months is 1.35 (see Table A3.4 in Annex
3). In the United States, where the CI equals 0.196, the average number of dentist visits in the past 12
months is 0.99. Thus, the number of dentist visits to be redistributed for each 1000 population to achieve
equality would be 107 in Canada and 145 in the United States.
Czech Republic
United Kingdom
Slovak Republic
Switzerland
Canada
Austria
Finland
Belgium
Slovenia
Ireland
New Zealand
Estonia
Spain
United States
Poland
Hungary
France*
Denmark**
0.00 0.20 0.40 0.60 0.80 1.00
Rates of dentist visits in the past 12 months
Lowest
income
quintile
Average Highest
income
quintile
DELSA/HEA/WD/HWP(2012)1
23

Figure 8: Inequality index for dentist visits in the past 12 months, 2009 (or latest year)

Note: (*) visits in the past 3 months in Denmark. In panel A: (**) visits in the past 24 months in France. In panel B: (**) counts in the
past 4 weeks in EHIS countries.
59. High-income persons were more likely to visit a dentist within the last 12 months in all countries.
There is, however, wide variation across countries in the concentration index. Inequalities are larger in
countries with a lower probability of a dental visit such as Hungary, Poland, Spain and the United States.
Denmark and France have different recall periods, which impacts on the average probability of dentist
visits, and may also have an effect on the level of inequalities in cross-country comparison, as mentioned
in section 3.2.
60. These results confirm findings from a recent study among Europeans aged 50 years and over
which identifies significant pro-rich inequalities in access to dental treatment. Among 14 European
countries, Poland and Spain display the largest inequality index, followed by Austria and Belgium (Listl,
2011).
61. The distinction between preventive and curative care in dental visits is also relevant. Listl (2011)
finds considerable income-related inequalities in dental service utilisation and attributes these to
inequalities in preventive dental visits, either alone or in combination with operative treatment. Similarly, a
recent study in Canada shows that access to preventive care is the most pro-rich type of dental care
utilisation (Grignon et al., 2010).
62. In several countries, recent policy reforms related to the extension of coverage or organisation of
dental care have affected access. The Finnish government introduced on 1 December 2002 a new policy
offering publicly-funded dental care for the whole population. Before that date, publicly-funded dental care
was limited according to age, and the majority of adult population had to use private dental care. The main
goals were to offer publicly-funded dental care for the whole population and to equalise access to dental
care by socioeconomic groups and by municipalities. Assessments of this policy implementation show that
equity and fairness of the oral health care provision system improved (Niiranen et al., 2008).

DELSA/HEA/WD/HWP(2012)1
24

4.5 Cancer Screening
63. The probabilities of screening for breast cancer in the last two years, and cervical cancer in the
last three years are shown in Figures 9 and 10. Rates of cancer screening participation among women
within the target age group vary widely across countries. Breast cancer screening participation in the past
two years varies from 36% in Estonia to 85% in Spain, and from 30% in Estonia to 85% in the United
States for cervical cancer screening. Some countries present large inequalities in screening rates across
income quintiles.
Figure 9: Probability of breast cancer screening in
the last two years, women aged 50-69 years, by
income quintile, 2009 (or latest year)
Note: (*) visits in the past 12 months in Denmark and Ireland.
Figure 10: Probability of cervical cancer screening in the
last three years, women aged 20-69 years, by income
quintile, 2009 (or latest year)
Note: (*) visits in the past 12 months in Ireland.

64. In the United States, low-income women, women who are uninsured or receiving Medicaid
(health insurance coverage for the poor, disabled or impoverished elderly), or women with lower
educational levels report much lower use of mammography and pap smears (NCHS, 2011). Other studies
in European countries confirm significant social inequalities in the utilisation of early detection and
prevention health care services (Von Wagner et al., 2011). In particular, women from higher
socioeconomic groups are more likely to have mammograms (Sirven and Or, 2010).
65. Figure 11 shows that pro-rich inequalities in breast cancer screening exist in almost all countries,
and are significant in eight countries (Belgium, Canada, Estonia, France, Ireland, New Zealand, Poland,
and the United States). Likewise, income-related inequalities in cervical cancer screening favour the well-
off (Figure 12), and are significant in all countries, except Slovenia. Caution is needed for all above
analyses for the Czech Republic and Slovenia which have large confidence intervals due to small sample
sizes.
Spain
Austria
United States
France
New Zealand
Canada
Belgium
Czech Republic
Hungary
Poland
Slovak Republic
Switzerland
United Kingdom
Slovenia
Estonia
Ireland*
Denmark*
0.00 0.20 0.40 0.60 0.80 1.00
Rates of breast cancer screening in the past 2 years
Lowest
income
quintile
Average Highest
income
quintile
United States
Austria
Spain
Slovenia
Canada
New Zealand
France
Poland
Denmark
Belgium
Czech Republic
Hungary
Slovak Republic
United Kingdom
Switzerland
Estonia
Ireland*
0.00 0.20 0.40 0.60 0.80 1.00
Rates of cervical cancer screening in the past 3 years
Lowest
income
quintile
Average Highest
income
quintile
DELSA/HEA/WD/HWP(2012)1
25

Figure 11: Inequality index for breast cancer
screening in the past 2 years, 2009 (or latest year)
Note: (*) visits in the past 12 months in Denmark and Ireland.
Figure 12: Inequality index for cervical cancer screening
in the past 3 years, 2009 (or latest year)

Note: (*) visits in the past 12 months in Ireland.

66. Analysis of possible links between inequalities in cancer screening and health system features,
such as free screening access and nationwide population-based programmes was undertaken, but no clear
relationship emerged. In another study, Palència et al (2010), examined 22 European countries, and found
that inequalities in use of breast and cervical cancer screening are higher in countries without a population-
based screening programme.
67. Other individual characteristics, such as ethnicity, younger age, higher level of education,
employment status, residential area, marital status, having health insurance, good health status, having a
usual source of care and use of other preventative services, are all recognized as important additional
predictors of participation in screening (Sirven and Or, 2010).

4.6 Comparison with earlier results
68. Comparing the absolute values of indexes obtained in Van Doorslaer and Masseria’s 2004 study
and in this paper should be done with caution. The values of the inequity and inequality indices between
the two studies often result from surveys which differ slightly in their methodology. Indeed, for many
countries, a different data source is used.
69. There are four countries (Canada, Switzerland, the United Kingdom and the United States) where
data come from the same source in both studies. However, in Switzerland questions on health status and
limitations have changed and so it is difficult to use similar information for the “need for care”
standardisation procedure. This matters since the quantity of health information in the need standardisation
procedure may lead to under- or over-estimation of pro-rich utilisation patterns (Van Doorslaer et al. 1993;
Van Doorslaer and Masseria, 2004).
70. In France, the data source is slightly different since the 2004 study used actual medical visits
from sickness fund records whereas the current study examines visits reported by individuals. In Denmark,
doctor visits were assessed for the past 12 months in the 2004 study versus the past three months in the
current study, which may lead to an over-estimation of pro-poor inequity. In Finland, many people do not
make a distinction between GPs and specialists, and so the general question in the 2000 ECHP survey on
-0.10
-0.05
0.00
0.05
0.10
0.15
0.20
0.25
-0.02
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
DELSA/HEA/WD/HWP(2012)1
26

visits to these physicians in the past 12 months might be difficult to understand. In the 2009 Finnish HYPA
survey, questions are more precise, since they ask about visiting a doctor at health centres (public) or at
occupational health clinics (used to define GP visits); visiting a doctor at outpatient departments or in
private practices (used to define specialist visits). These differences may affect comparisons, in particular
for specialist visits. Similarly, for Ireland and Spain, different sources of data were used, which may cause
inconsistencies in results over time, notably for dental care.
71. Results indicate that, with few exceptions, inequities and inequalities have remained stable over
time. Figure 13 shows that the country ranking is reasonably consistent with the 2004 study, and is
particularly stable regarding doctor and GP visits. The country rank has changed slightly in specialist
visits, due to Finland, Denmark and France’s positions, and in dentist visits because of Ireland, Finland and
Spain’s positions.
Figure 13: Comparison of inequity indexes, 2000 and 2009

Note: Difference in data sources in 9 countries (Austria, Belgium, Denmark, Finland, France, Germany, Hungary, Ireland, and Spain).

4.7 Country overview
72. Individual country findings from the analysis are summarised in Table 2. For each type of health
care, an indication is given as to the average probability of a visit in the past 12 months (or 2 years in the
case of breast cancer screening, and 3 years for cervical cancer screening), relative to other countries. An
indication is also given as to the level of equity found in doctor visits after the adjustment for need, and
DELSA/HEA/WD/HWP(2012)1
27

equality in dental visits and breast and cervical cancer screening. If inequalities exist, these favour the
well-off.
Table 2. Country overview of probability of a doctor or dentist visit or cancer screening, and level of
inequality

Note: For Austria and Ireland, the findings for doctors refer to GPs only. n.a. means not available. The words equitable/inequitable
refer to doctor visits after adjustment for need and, equal/unequal refer to dentist visits and cancer screening which are not adjusted
for need.
Country
Doctor (GP and specialist)
visits
Dentist visits
Breast and cervical cancer
screening
Austria
Higher probability of a (GP)
visit; Equitable
Medium probability of a visit;
Unequal
High probability of screening;
Equal (breast), unequal
(cervical)
Belgium
Higher probability of a visit;
Inequitable
Medium probability of a visit;
Unequal
Medium probability of
screening; Unequal
Canada
Higher probability of a visit;
Inequitable
Higher probability of a visit;
Unequal
Higher probability of
screening; Unequal
Czech Republic
Higher probability of a visit;
Equitable
Highest probability of a visit;
Unequal
Medium probability of
screening; Equal (breast),
unequal (cervical)
Denmark n.a.
Higher probability of a visit;
Unequal
Medium probability of
screening; Equal (breast),
unequal (cervical)
Estonia
Lower probability of a visit;
Inequitable
Lower probability of a visit;
Unequal
Lowest probability of
screening; Most unequal
Finland
Lower probability of visit;
Inequitable
Medium probability of a visit;
Unequal
n.a.
France
Highest probability of a visit;
Inequitable
Medium probability of a visit;
Unequal
Higher probability of
screening; Unequal
Germany
Higher probability of a visit;
Inequitable
n.a. n.a.
Hungary
Medium probability of a visit;
Inequitable
Lowest probability of a visit;
Unequal
Medium probability of
screening; Equal (breast),
unequal (cervical)
Ireland
Medium probability of a (GP)
visit; Equitable
Medium probability of a visit;
Unequal
Medium probability of
screening; Unequal
New Zealand
Medium probability of a visit;
Inequitable
Medium probability of a visit;
Unequal
Higher probability of
screening; Unequal
Poland
Lower probability of a visit;
Inequitable
Lower probability of a visit;
Unequal
Medium probability of
screening; Unequal
Slovak Republic
Medium probability of a visit;
Inequitable
Higher probability of a visit;
Unequal
Medium probability of
screening; Equal (breast),
unequal (cervical)
Slovenia
Lower probability of a visit;
Equitable
Medium probability of a visit;
Unequal
Lower probability of
screening; Equal
Spain
Medium probability of a visit;
Inequitable
Lower probability of a visit;
Unequal
High probability of screening;
Equal (breast), unequal
(cervical)
Switzerland
Medium probability of a visit;
Inequitable
Higher probability of a visit;
Unequal
Lower probability of
screening; Equal (breast),
unequal (cervical)
United Kingdom
Medium probability of a visit;
Equitable
Higher probability of a visit;
Unequal
Lower probability of
screening; Equal (breast),
unequal (cervical)
United States
Lowest probability of visit;
Most inequitable
Lower probability of a visit;
Most unequal
High probability of screening;
Unequal
DELSA/HEA/WD/HWP(2012)1
28

5. INEQUALITIES IN HEALTH CARE UTILISATION RELATED TO HEALTH SYSTEM
FINANCING
73. Explaining persisting inequalities in the utilisation of health care services in OECD countries has
followed three general approaches. A first approach focuses on cultural and information barriers, which
hinder individuals obtaining knowledge of care pathways and seeking care, and are most commonly
experienced among the poor and least educated. Secondly, characteristics related to the organisation of
health systems may create barriers; whether GPs act as ‘gatekeepers’ to facilitate access for persons in
lower socioeconomic positions, or whether primary care systems explicitly focus on the needs of the most
deprived are two such characteristics (Or, et al., 2008). Thirdly, the direct cost of care to individuals, and
particularly whether they have secondary private health insurance, has most often been associated with
inequalities in health care use (van Doorslaer and Masseria, 2004).
74. This section provides further evidence on how the cost of care can create inequalities in health
care utilisation. The results for 2008-09 are complemented by information from the OECD Health Data
2011 database, and the 2008 OECD Health Systems Characteristics survey (Paris et al., 2010) to explore
macro-level relationships between inequality in health services use and the financing of health systems, in
the following areas:
• Public health settings
• Out-of-pocket payments
• Private health insurance.

5.1 Public health settings
75. Health care coverage promotes access to medical goods and services, as well as providing
financial security against unexpected or serious illness. Most OECD countries have achieved universal
coverage of health care costs for at least a core set of services, sometimes through combinations of public
and private health insurance.
76. However, health insurance coverage, even if universal, is an imperfect indicator of accessibility,
since the range of services covered and the degree of cost-sharing applied to some services (e.g. dental care
and pharmaceutical drugs) varies substantially across countries (Paris et al., 2010)
4
.
77. The distribution of health costs between public and private funds is an important dimension that
affects health care access. Previous studies have found a relationship between the share of public health
spending in total health expenditure and lower inequity in doctor consultations (Or et al., 2008).
Conversely, private funding is often regressive and negatively impacts on the uptake of needed services, in
particular for vulnerable people at risk of social exclusion (Huber et al., 2008).

4
Annex 4 presents the link between inequity and the indicator “depth of coverage” as defined in Paris et al.
(2010). However, since there was little variation in depth of coverage across countries, findings were not
conclusive.
DELSA/HEA/WD/HWP(2012)1
29

78. Figure 14 highlights the relationship between high public financing and low inequity in doctor
visits. The United States, with its reliance on primary private health insurance, stands out as having a
substantially lower share of public health expenditure; its exclusion leads to a weaker, but still significant
relationship.
Figure 14: Relationship between inequity in doctor visits and the share of public health expenditure

Source: OECD estimates for inequity indexes, and OECD Health Data 2011

79. A recent literature review focusing on health system features supports that “nationalized,
publicly funded health care systems are most effective at reducing inequalities in access and reducing the
effects on health of income distribution” (Gelormino et al., 2011). Or et al. (2008) highlight that National
Health Services where financing and provision of care are handled centrally demonstrate less inequity in
specialist visits.
80. Data analysis using the 2008 OECD Health Systems Characteristics survey
5
lends support to
these results. Figure 15 shows that countries with a higher degree of private provision of care display
higher levels of inequity in specialist consultations.
Figure 15: Relationship between inequity in specialist visits and the degree of private provision of services


5
More details on the degree of public/private provision of care in Annex 4.
BEL
CAN
CZE
DNK
EST FIN
FRA
DEU
HUN NZL
POL
SVK
SVN
ESP
CHE
GBR
USA
R² = 0.5336
R² = 0.2146
-0.04
-0.02
0.00
0.02
0.04
0.06
40 50 60 70 80 90
I
n
e
q
u
i
t
y

i
n

d
o
c
t
o
r

v
i
s
i
t
s

(
H
I
)
Public Health Expenditure as % of Total current health expenditure
Linear (all countries) Linear (except the US)
BEL
CAN
CZE
DNK
EST
FIN
FRA
HUN
NZL
POL
SVK
SVN
ESP
CHE
GBR
R² = 0.2343
0.00
0.02
0.04
0.06
0.08
0.10
-3 0 3 6
I
n
e
q
u
i
t
y

i
n

s
p
e
c
i
a
l
i
s
t

v
i
s
i
t
s

(
H
I
)
Degree of private provision in outpatient specialists services
Low
Medium High
DELSA/HEA/WD/HWP(2012)1
30

Source: OECD estimates for inequity indexes, and OECD Health Systems Characteristics 2008

81. Another financial barrier is related to access at the point of service delivery. Paying up-front costs
that are reimbursed later creates barriers for low-income households. Bureaucratic requirements for
reimbursement or lack of knowledge are further disincentives (Huber et al., 2008).
82. There is a lack of information on free access at the point-of-care across countries. A question in
the 2008 OECD Health System Characteristics survey on “regulation of prices billed by physicians”
assesses whether patients must pay extra fees that are not reimbursed when consulting a doctor, and can
serve as a proxy of free access at point-of-care. Findings indicate that some countries with high price
regulation display lower inequities in specialist care utilisation (e.g. United Kingdom, and Czech
Republic). However, the relationship is not strong – New Zealand has the lowest regulation of physician
pricing among the countries considered, but has average inequity; and Hungary has a medium degree of
regulation but has one of the lowest levels of inequity (see Annex 4).
83. Safety nets assist in delivering equity of access by subsidising coverage or providing services for
low-income or economically disadvantaged groups. In those countries where coverage is not automatically
provided to all residents through national or local health systems, policies have been implemented to
guarantee access to care for people with low-income or high health risks, either through subsidies for the
purchase of insurance, dedicated programmes, or direct provision of health care (Paris et al., 2010).
84. Almost all countries have implemented policies to protect population groups from usual
copayments or excessive out-of-pocket expenses. Exemptions from out-of-pocket payments for low-
income people exist in half of OECD countries (Paris et al., 2010).
85. Safety nets for the elderly in Japan play a major role in ensuring equity in access. Results from a
recent study suggest pro-rich distribution of health care services
6
in the population aged under 65
(HI=0.013) (Ikegami et al., 2011). However, in people aged 65 and over, estimates suggest equity in health
care services utilisation (HI=-0.006
7
), reflecting the reduced copayment systems in place for the elderly in
Japan. These aim to relieve the burden of cost-sharing, since the elderly often have a limited income based
on their pension.
5.2 Out-of-pocket payments
86. An important determinant of the degree of inequality in health care use is the size of household
out-of-pocket payments. Or et al. (2008) used European data to show that social inequalities are stronger in
countries where out-of-pocket payments are higher, and where the share of public health spending in GDP
is smaller. Gelormino et al. (2011) in reviewing the effects of health care reforms in European countries,
underline that out-of-pocket payments increase inequalities in access to care and contribute to
impoverishment. Direct payments for health care penalise the worst-off, creating barriers to access and
potentially further damaging health. Recent increases in the share of health expenditure in the most
deprived households in Slovenia, for example, have jeopardised access to a number of health goods and
services, including medicines, therapeutic appliances, dental services and outpatient specialised health care

6
The question asks about any visit to healthcare services in the past month, including outpatient and
inpatient doctor visits as well as to traditional medicine practitioners. The majority of health care service
utilisation in Japan consists of outpatient clinic visits.
7
Analysis carried out by Prof. Hideki Hashimoto, University of Tokyo School of Public Health.
DELSA/HEA/WD/HWP(2012)1
31

(WHO Europe, 2011). To offset their effect, many countries have introduced safeguards against excessive
out-of-pocket payments.
87. Analysis of out-of-pocket payments in total basic medical and diagnostic services
8
in 2008-2009
and the inequity index in doctor visits found no significant relationship, lending weight to countries’
effective use of safeguards.
88. In contrast, the analysis of out-of-pocket payments as a percentage of total expenditure on
specialist care shows a significant relationship with the degree of inequity in the probability of specialist
visits (Figure 16, panel A). Similarly, a positive association was observed between out-of-pocket payments
in dental care and inequality in dentist’s consultations (Figure 16, panel B).
Figure 16: Relationship between inequity and the share of out-of-pocket payments
Panel A. Specialist visits Panel B. Dentist visits
Source: OECD estimates for inequity indexes, and System of Health Account 2011
89. Poland has one of the largest shares of out-of-pocket expenditure on health care among OECD
countries. Specialist medical services paid out-of-pocket—typically consultations provided at private
medical facilities—is the second largest component of private expenditure. For dental care, about two-
thirds of expenditure is financed privately, with only basic treatment covered by public insurance, leading
to highly unequal access. Long waiting times to access health care has led to a network of informal
payments and other queue-jumping mechanisms (OECD, 2012).
90. In Spain, adult dental care is largely absent from the benefits package, with correspondingly high
out-of-pocket expenses (Garcia-Armesto et al., 2010). A recent initiative proposes a number of policies
and interventions to reduce social inequalities (Commission on the Reduction of Social Inequalities in
Health in Spain, 2011).


8
Out-of-pocket payment in total basic medical and diagnostic services is the closest available measure for
health spending related to doctor consultations.
CZE
FIN
SVN
BEL
ESP
HUN
CHE
POL
R² = 0.2786
0
0.02
0.04
0.06
0.08
0.1
0.12
0 10 20 30 40
I
n
e
q
u
i
t
y


i
n

s
p
e
c
i
a
l
i
s
t

v
i
s
i
t
s

(
H
I
)
OOP payments as % of total expenditure on specialist care
FRA
BEL
SVN
AUT
CAN
CZE
SVK
FIN
EST
NZL
POL HUN
ESP
R² = 0.2717
0
0.02
0.04
0.06
0.08
0.1
0.12
0 20 40 60 80 100
I
n
e
q
u
a
l
i
t
y


i
n

d
e
n
t
a
l

v
i
s
i
t
s

(
C
I
)
OOP payments as % of total dental expenditure
DELSA/HEA/WD/HWP(2012)1
32

5.3 Private health insurance
91. Private health insurance (PHI) plays an important role in facilitating access, most notably for
specialist and dental care. Persons in countries with private health insurance markets have the possibility of
not only buying more or better care, but also of receiving care more quickly. The function of PHI differs
across countries (OECD, 2004). It can be a primary insurance that represents the only available access to
basic health cover because individuals do not have public health insurance. PHI can also act as a secondary
insurance in addition to public coverage: as a complementary insurance which covers any cost-sharing
remaining after basic coverage, a supplementary insurance which provides cover for additional health
services not covered by the public scheme, or a duplicative insurance that covers services which are
already included in the basic benefit package, but provides faster private-sector access to medical services
where there are waiting times in public systems.
92. Multivariate logistic regression models were used to assess the relationship between having PHI
and the probability of seeing a doctor. This analysis was conducted on a set of countries for which national
survey data provided information on PHI (France, Germany, Ireland, New Zealand, Switzerland, the
United Kingdom, and the United Stated). Because of the complex role PHI serves across countries, more
information besides the single dimension of the “share of total PHI” is needed to assess the link with
inequity in utilisation.
93. In France, almost all of the population is covered for basic primary health care by the social
security system; 94% of the French population has complementary health insurance to cover cost-sharing.
The complementary health insurance is privately funded, with the exception of CMU-C (Couverture
Médicale Universelle Complémentaire). The CMU-C provides care free of charge to the most deprived
population. In this analysis, to isolate the impact of complementary insurance, the indicator refers to
“having a complementary health insurance” which captures both PHI and CMU-C vs. no complementary
insurance.
94. In Germany, most of the population (83%) has basic primary coverage through sickness funds,
and others are privately insured (e.g. the self-employed, civil servants, opting-out income ceiling). People
who are publicly covered can purchase private insurance to cover cost-sharing (complementary PHI) or for
additional services (supplementary PHI). In this analysis, the indicator refers to “having a private health
insurance either as a primary coverage or as a secondary source of coverage” vs. “not privately insured”.
95. In Ireland, basic primary health coverage is automatic and financed from taxes. About one third
of the population, under a certain income threshold, can access public health services free of charge
through a Medical Card. About 3% of the population who are not entitled to a Medical Card have a GP
Visit Card whereby GP visits are covered by the public system. About 47% of the Irish population has
private health insurance, mainly to cover costs of inpatient bed use, some out-of-pocket charges in the
primary care sector, and to bypass waiting lists for inpatient services. Private health insurance plays both a
complementary and supplementary role in Ireland.
96. In New Zealand, basic primary health care coverage is supplied by national health services.
However, about 32% of the population choose a private insurance to supplement what they receive from
the public health system with more comprehensive and timely care. Private health insurance also plays
both a complementary and supplementary role in New Zealand.
97. In Switzerland, basic coverage is mandatory and people can choose among four different options
for health insurance plans: 1) ordinary basic policies, 2) policies with choice of deductibles, 3) bonus
insurance, 4) insurance with a limited choice of providers. Ordinary contracts offer the highest level of
financial protection against health care spending but also have the highest level of premiums. Other forms
DELSA/HEA/WD/HWP(2012)1
33

of health insurance contracts offer lower premiums with either higher deductibles or restrictions in the
choice of doctor or hospital (OECD, 2011d). Supplementary insurance for additional comfort or treatment
is purchased by about 30% of the population. Contracts typically cover one or several of the following
benefits: private rooms in hospitals, dental care, alternative medicines and cash benefits for sickness
absence.
98. In the United Kingdom, the NHS provides preventive medicine, primary care and hospital
services to all residents. Around 12% of the population is covered by a private medical insurance, which
mainly provides access to acute elective care in the private sector. PHI in the United Kingdom is mainly a
supplementary insurance, since it provides cover for enhanced services such as faster access and increased
consumer choice (Boyle, 2011).
99. In the United States, 81% of the population has a basic primary health insurance. Most people
(55% of the population) are covered through private insurance plans, usually tied to their employment,
while Medicare (for those aged over 65) and Medicaid (for the poor) provide public coverage. In this
analysis, the indicator refers to “having a private primary health insurance” vs. “having a public insurance
or no insurance”.
100. Despite differences in PHI settings across countries, higher-income people are more likely to
purchase PHI (Figure 17).
Figure 17: Share of PHI by income quintile, selected OECD countries, 2009 (or latest year)

Source: National population-based surveys (see Table 1).
Note: Figures refer to both PHI and CMU-C (public) complementary coverage in France. In Germany, they represent the share of
population “having a private health insurance either as a primary coverage or as a secondary coverage”. In the US, they represent
the share of population “having primary private health insurance”.
8
8 9
4 9
6
9
7
9
7
1
7
2
3
3
1
4
0
5
6
2
3
3
6
5
1
6
0
7
3
1
5
2
1
3
8
5
0
6
2
1
4
2
1
2
9
3
6
4
8
5
8
1
4
2
0
3
4
2
3
4
2
6
3
8
1
8
9
0%
20%
40%
60%
80%
100%
L
o
w
e
r

i
n
c
o
m
e
M
i
d
d
l
e
-
l
o
w

M
i
d
d
l
e
M
i
d
d
l
e
-
h
i
g
h
H
i
g
h
e
r

i
n
c
o
m
e
L
o
w
e
r

i
n
c
o
m
e
M
i
d
d
l
e
-
l
o
w

M
i
d
d
l
e
M
i
d
d
l
e
-
h
i
g
h
H
i
g
h
e
r

i
n
c
o
m
e
L
o
w
e
r

i
n
c
o
m
e
M
i
d
d
l
e
-
l
o
w

M
i
d
d
l
e
M
i
d
d
l
e
-
h
i
g
h
H
i
g
h
e
r

i
n
c
o
m
e
L
o
w
e
r

i
n
c
o
m
e
M
i
d
d
l
e
-
l
o
w

M
i
d
d
l
e
M
i
d
d
l
e
-
h
i
g
h
H
i
g
h
e
r

i
n
c
o
m
e
L
o
w
e
r

i
n
c
o
m
e
M
i
d
d
l
e
-
l
o
w

M
i
d
d
l
e
M
i
d
d
l
e
-
h
i
g
h
H
i
g
h
e
r

i
n
c
o
m
e
L
o
w
e
r

i
n
c
o
m
e
M
i
d
d
l
e
-
l
o
w

M
i
d
d
l
e
M
i
d
d
l
e
-
h
i
g
h
H
i
g
h
e
r

i
n
c
o
m
e
L
o
w
e
r

i
n
c
o
m
e
M
i
d
d
l
e
-
l
o
w

M
i
d
d
l
e
M
i
d
d
l
e
-
h
i
g
h
H
i
g
h
e
r

i
n
c
o
m
e
France Germany Ireland New Zealand Switzerland United Kingdom United States
P
r
i
v
a
t
e

h
e
a
l
t
h

i
n
s
u
r
a
n
c
e
Privately insured CMU (France only)
DELSA/HEA/WD/HWP(2012)1
34

101. PHI enhances access to medical goods and services, in particular to secondary and dental care.
Figure 18 shows that in most countries people with PHI are significantly more likely to consult doctors and
dentists, although with some exceptions.
Figure 18: Adjusted probabilities of medical visits by PHI status, in selected OECD countries
Panel A. Doctor visits Panel B. GP visits

Panel C. Specialist visits Panel D. Dentist visits

Source: National population-based surveys (see Table 1).
Note: Figure 18 shows the adjusted probabilities of medical visits, all other things being equal. The analysis controlled for a range
of covariates: age, gender, ethnicity, health status, marital status, education level, occupation, income level, region, size of the
household. In France, the indicator refers to “having a complementary insurance” (both CMU-C and PHI) vs. “no complementary
insurance”. In Germany, it refers to “having a private health insurance either as a primary basic coverage or as a secondary
coverage”. In the US, it refers to “having a private primary health insurance” vs. “having a public primary insurance or no
insurance”.
102. In four out of six countries (France, New Zealand, Switzerland, and the US) people with PHI are
more likely to consult a doctor. The effect of PHI is not significant in Germany and the United Kingdom
(Panel A). Regarding GP visits, people with PHI are more likely to consult in four out of five countries
(France, Ireland, New Zealand, and Switzerland), the UK displaying no significant difference (Panel B).
For specialist visits, similar results are found
9
(Panel C). Regarding dentist visits, results show in all

9
Irish data does not provide information on specialist visits.
9
4
%
9
1
%
9
0
%
8
7
%
8
8
%
7
4
%
8
8
%
9
0
%
8
5
%
8
4
%
8
7
%
5
8
%
30%
40%
50%
60%
70%
80%
90%
100%
France Germany New
Zealand
Switzerland United
Kingdom
United
States
P
r
o
b
a
b
i
l
i
t
y

o
f

v
i
s
i
s
t
i
n
g

a

d
o
c
t
o
r
Privately insured Not privately insured
8
9
%
8
5
%
8
8
%
7
2
%
8
3
%
8
1
%
8
0
%
8
3
%
7
0
%
8
1
%
30%
40%
50%
60%
70%
80%
90%
100%
France Ireland New Zealand Switzerland United
Kingdom
P
r
o
b
a
b
i
l
i
t
y

o
f

v
i
s
i
s
t
i
n
g

a

G
P
Privately insured Not privately insured
5
8
%
3
8
%
6
1
%
4
2
%
4
4
%
2
9
%
5
1
%
4
1
%
0%
20%
40%
60%
80%
100%
France New Zealand Switzerland United Kingdom
P
r
o
b
a
b
i
l
i
t
y

o
f

v
i
s
i
s
t
i
n
g

a

s
p
e
c
i
a
l
i
s
t
Privately insured Not privately insured
7
8
%
5
6
%
5
7
% 7
0
%
7
6
%
4
1
%
6
7
%
4
7
%
4
6
%
6
5
%
6
9
%
2
5
%
0%
20%
40%
60%
80%
100%
France Ireland New
Zealand
Switzerland United
Kingdom
United
States
P
r
o
b
a
b
i
l
i
t
y

o
f

v
i
s
i
s
t
i
n
g

a

d
e
n
t
i
s
t
Privately insured Not privately insured
DELSA/HEA/WD/HWP(2012)1
35

studied countries that people having PHI are significantly more likely to consult than those without PHI
(Panel D).
103. In France, a recent study analysed the complementary insurance and its relationship with access
to care (Perronin et al., 2011) As expected, the authors found that people with private health insurance or
CMU-C compared to people without complementary insurance consult GPs, specialists and dentists more
often, and report fewer unmet care needs, although the CMU-C-insured do not reach the level of utilisation
of privately insured people.

DELSA/HEA/WD/HWP(2012)1
36

6. CONCLUSION
104. Horizontal inequities and inequalities in health care utilisation persist across the 19 OECD
countries studied. After adjustment for needs for health care, the better-off are more likely to visit doctors -
especially specialists - than those with lower incomes. With GP contacts, the scenario is different. In most
countries, for the same level of need for health care, the worse-off are as likely as the better-off to contact a
GP, and they visit more often. Income-related inequalities in breast cancer screening appear in around half
of all countries, with a higher rate among the better-off. Pro-rich inequalities in dental visits and in cervical
cancer screening are present in almost all countries.
105. A comparison with previous results (Van Doorslaer and Masseria, 2004) shows that the relative
position of countries between 2000 and 2009 has remained stable for inequities in doctor and GP visits.
There is no strong evidence of diminishing inequities although some discrepancies are found in country
ranks for specialist and dentist visits, these being mostly methodological in nature.
106. Findings highlight the important effects of certain health system features on equity. Broader
health insurance coverage improves access. The higher the share of public health expenditure, the lower
the inequity in doctor visits. Similarly, greater inequity in specialist visits accompanies a higher degree of
private provision. A greater share of out-of-pocket payments is associated with inequity in specialist and
dental care. Secondary private health insurance facilitates the use of care, with the privately insured more
likely to visit doctors and dentists.

DELSA/HEA/WD/HWP(2012)1
37

REFERENCES
AHRQ (Agency for Healthcare Research and Quality) (2012), National Healthcare Disparities Report
2011, AHRQ, Rockville, MD.
Allin, S., M. Grignon, Le Grand J. (2010), “Subjective Unmet Need and Utilization of Health Care
Services in Canada: What are the Equity Implications?”, Soc Sci Med., Vol. 70(3), pp. 465-472.
García-Armesto, S. et al. (2010), Spain: Health System Review. Health Systems in Transition, Vol. 12(4),
pp. 1-295.
Bago d’Uva T., A. M. Jones, E. van Doorslaer, (2009), “Measurement of Horizontal Inequity in Health
Care Utilisation Using European Panel Data”, Journal of Health Economics, Vol. 28, pp. 280–289.
Boyle, S. (2011), United Kingdom (England): Health System Review. Health Systems in Transition, Vol.
13(1).
Commission on the Reduction of Social Inequalities in Health in Spain (2011), Moving Forward Equity. A
Proposal of Policies and Interventions to Reduce Social Inequalities in Health in Spain, Directorate
General for Public Health and Foreign Health Ministry of Health, Social Policy and Equality,
Madrid.
De Looper M. and Lafortune G. (2009), “Measuring Disparities in Health Status and in Access and Use of
Health Care in OECD Countries”, OECD Health Working Paper No. 43.
DeSalvo, K. et al. (2005), “Predicting Mortality and Healthcare Utilization With a Single Question”,
Health Services Research, Vol. 40, pp. 1234–1246.
Dourgnon P., S. Guillaume, M. Naiditch and C. Ordonneau (2007), “Introducing Gate Keeping in France:
First Assessment of the Preferred Doctor Scheme Reform”, IRDES Question d’économie de la santé
No.124.
Gelormino, E., C. Bambra, T. Spadea, S. Bellini, and G. Costa (2011), “The Effects of Health Care
Reforms on Health Inequalities: A Review and Analysis of the European Evidence Base,
International Journal of Health Services, Vol. 41(2), pp. 209-230.
Grignon, M, J. Hurley, L. Wang and S. Allin (2010), “Inequity in a Market-Based Health System:
Evidence From Canada’s Dental Sector”, Health Policy, Vol. 98, pp. 81-90.
Habicht J., R. Kiivet, T. Habicht and A. E. Kunst (2009), “Social Inequalities in the Use of Health Care
Services After 8 years of Health Care Reforms – A Comparative Study of the Baltic Countries”, Int
J Public Health, Vol. 54, pp. 250–259.
Hagenaars, A., K. de Vos and M.A. Zaidi (1994), Poverty Statistics in the Late 1980s: Research Based on
Micro-data, Office for Official Publications of the European Communities. Luxembourg.
DELSA/HEA/WD/HWP(2012)1
38

Huber, M., A. Stanicole, J. Bremner and K. Wahlbeck (2008), Quality in and Equality of Access to
Healthcare Services, European Commission Directorate-General for Employment, Social Affairs
and Equal Opportunities, Luxembourg.
Hurst, J. and M. Jee-Hughes (2001), “Performance Measurement and Performance Management in OECD
Health Systems”, Labour Market and Social Policy Series, Occasional Paper No. 47, OECD, Paris.
Idler, E. L. & Y. Benyamini (1997), “Self-rated Health and Mortality: A Review of Twenty-seven
Community Studies”, Journal of Health and Social Behaviour, Vol. 38, pp. 21-37.
Ikegami N., B.-K. Yoo, H. Hashimoto et al. (2011), “Japanese Universal Health Coverage: Evolution,
Achievements, and Challenges”, The Lancet, Vol. 378(9796), pp. 1106-1115.
Kelley, E. and J. Hurst (2006), “Health Care Quality Indicators Project: Conceptual Framework Paper”,
OECD Health Working Papers, No. 23, OECD Publishing, Paris.
Koolman, X. and E. van Doorslaer (2004), “On the Interpretation of a Concentration Index of Inequality”,
Health Economics, Vol. 13, pp. 649-656.
Listl, S. (2011), “Income-Related Inequalities in Dental Service Utilization by Europeans Aged 50+”,
Journal of Dental Research, Vol. 90(6), pp. 717-723.
Lu J. R., Leung G. M., Kwon S., et al. (2007), “Horizontal Equity in Health Care Utilization Evidence
From Three High-Income Asian Economies”, Social Science & Medicine, 64(1), 199-212.
Mackenbach, J. P. (2006), Health Inequalities: Europe in Profile, Department of Health, London.
Mackenbach, J. P. et al. (2008), “Socioeconomic Inequalities in Health in 22 European Countries”, New
England Journal of Medicine, Vol. 358, pp. 2468-2481.
Niiranen, T., E. Widström and T. Niskanen (2008), “Oral Health Care Reform in Finland – Aiming to
Reduce Inequity in Care Provision”, BMC Oral Health, Vol. 8:3, doi:10.1186/1472-6831-8-3.
National Center for Health Statistics (NCHS) (2011), Health, United States, 2010: With Special Feature on
Death and Dying, NCHS, Hyattsville, MD.
O'Donnell, O., E. Van Doorslaer, A. Wagstaff and M. Lingelow (2008), Analyzing Health Equity Using
Household Survey Data, The World Bank, Washington D.C.
OECD (2004), Private Health Insurance in OECD countries, OECD Publishing, Paris.
OECD (2011a), Health at a Glance 2011: OECD Indicators, OECD Publishing, Paris.
OECD (2011b), “What explains different cancer survival rates?” OECD Background Paper DELSA/HEA/
(2011)6.
OECD (2011c), How’s Life? Measuring Well-Being, OECD Publishing, Paris.
OECD (2011d), OECD Reviews of Health Systems – Switzerland, OECD Publishing, Paris.
OECD (2012), OECD Economic Surveys: Poland 2012, OECD Publishing, Paris.
DELSA/HEA/WD/HWP(2012)1
39

Or, Z., F. Jusot and E. Yilmaz (2008), “Impact of Health Care System on Socioeconomic Inequalities in
Doctor Use”, IRDES Working Paper No. 17, IRDES, Paris.
Palència L., S. Espelt, M. Rodriguez-Sanz, et al. (2010). “Socio-economic inequalities in breast and
cervical cancer screening practices in Europe: influence of the type of screening program”,
International Journal of Epidemiology, 39: 757-765.Palència, L., S. Espelt, M. Rodriguez-Sanz, et
al. (2011), “Trends in Social Class Inequalities in the Use of Health Care Services Within the
Spanish National Health System, 1993-2006”, Eur J Health Econ., Published online: 10 November
2011.
Paris, V., M. Devaux and L. Wei (2010), “Health Systems Institutional Characteristics: A Survey of
29 OECD Countries”, OECD Health Working Paper No. 50, OECD Publishing, Paris.
Perronin, M., A. Pierre and T. Rochereau (2011), “Complementary Health Insurance in France: Wide-
Scale Diffusion but Inequalities of Access Persist”, IRDES Questions d’économie de la santé
No.161.
Regidor, E., D. Martínez, M. E. Calle, P. Astasio, P. Ortega and V. Domínguez (2008), “Socioeconomic
Patterns in the Use of Public and Private Health Services and Equity in Health Care”, BMC Health
Services Research, Vol. 8:183.
Sirven, N. and Z. Or (2010), “Disparities in Regular Health Care Utilisation in Europe”, IRDES Working
Paper No. 37, IRDES, Paris.
Tubeuf, S., F. Jusot, M. Devaux and C. Sermet (2008), “Social Heterogeneity in Self-Reported Health
Status and Measurement of Inequalities in Health”, IRDES Working Paper No. 12, IRDES, Paris.
Van Doorslaer, E., A. Wagstaff and F. Rutten (eds) (1993), Equity in the Finance and Delivery of Health
Care: an International Perspective, Oxford University Press, Oxford.
Van Doorslaer, E., X. Koolman and F. Puffer (2002), “Equity in the Use of Physician Visits in OECD
Countries: Has Equal Treatment for Equal Need Been Achieved?”, Measuring Up: Improving
Health Systems Performance in OECD Countries, OECD, Paris, pp. 225-248.
Van Doorslaer, E. and C. Masseria (2004), “Income-related Inequality in the Use of Medical Care in 21
OECD Countries”, OECD Health Working Paper No. 14, OECD, Paris.
Von Wagner, C. et al. (2011), “Inequalities in Participation in an Organized National Colorectal Cancer
Screening Programme: Results from the First 2.6 Million Invitations in England”, International
Journal of Epidemiology, Vol. 40, pp. 712-718.
Wagstaff, A., P. Paci and E. Van Doorslaer (1991), “On the Measurement of Inequalities in Health”, Soc
Sci Med, Vol. 33(5), pp. 545-557.
Wagstaff, A. and E. Van Doorslaer (2000), “Equity in Health Care Finance and Delivery”, In: A. J. Culver
and J. P. Newhouse (eds.), Handbook of Health Economics, Volume 1, Elsevier.
WHO (2012), World Conference on Social Determinants of Health, www.who.int/sdhconference/en/.
WHO Europe (2011), Health Inequalities in Slovenia, National Institute of Public Health, Ljubljana.
DELSA/HEA/WD/HWP(2012)1
40

ANNEXES
ANNEX 1. SET OF AVAILABLE INFORMATION IN EACH SURVEY

DELSA/HEA/WD/HWP(2012)1
41

Table A1.1. Dependent variables


Country GP visits: probability GP visits: frequency Specialists visits: probability Specialists visits: frequency Dentists visits: probability Dentists visits: frequency
Breast cancer screening
(mammography)
Cervical cancer screening
(pap smear)
Canada
2007-08
In the past 12 months, have you
seen, or talked to any of the
f ollowing health prof essionals
about your physical, emotional or
mental health: a f amily doctor
(pediatrician) or general
practitioner?
In the past 12 months, how
many times have you seen, or
talked on the telephone, about
your physical, emotional or
mental health with a f amily
doctor (pediatrician) or general
practitioner?
In the past 12 months, have you
seen, or talked to an eye
specialist, such as an
ophthalmologist or optometrist
(about your physical, emotional
or mental health)?
- any other medical doctor or
specialist such as a surgeon,
allergist, orthopaedist,
[gynaecologist/urologist] or
psychiatrist (about your
physical, emotional or mental
health)?
In the past 12 months, how
many times have you seen, or
talked on the telephone, about
your physical, emotional or
mental health with an eye
specialist (such as an
ophthalmologist or optometrist)?
- any other medical doctor or
specialist such as a surgeon,
allergist, orthopaedist,
[gynaecologist/urologist] or
psychiatrist (about your
physical, emotional or mental
health)?
In the past 12 months, have you
seen, or talked to a dentist,
dental hygienist or
orthodontist (about your
physical, emotional or mental
health)?
In the past 12 months, how
many times have you seen, or
talked on the telephone, about
your physical, emotional or
mental health with a dentist or
orthodontist?
Have you ever had a
mammogram, that is, a breast x-
ray? Yes / No
When was the last time?
1.<6months ago
2. 6months-<1year
3. 1-<2 years
4. 2-<5 years
5.>=5 years
Have you ever had a pap smear
test? Yes/ No
When was the last time?
1.<6months ago
2. 6months-<1year
3. 1-<3 years
4. 3-<5 years
5.>=5 years
Denmark 2005
During the past 3 months, have
you consulted a physician
because of disease, disorder,
illness or injury?
Yes, my general practitioner
Number of times During the past 3 months, have
you consulted a physician
because of disease, disorder,
illness or injury?
Yes, a practising specialist
physician
Number of times Have you consulted other health
care providers during the past 3
months?
Yes, a dentist
Number of times Mammography in the past 12
months
Pap smear in the past 3 years
Finland
2009
in past 12 months have you
visited a doctor at health centers
(public) or at occupational health
clinics ?
Number of visits in past 12
months
in past 12 months, have you
visited a doctor at outpatient
departments or in private
practices?
Number of visits in past 12
months
in past 12 months, have you
visited a dentist at health centers
or in private practices?
Number of visits in past 12
months
missing missing
France
2008
in past 12 months have you
visited a GP?
Number of visits in past 12
months
in past 12 months, have you
visited a specialist?
Number of visits in past 12
months
in past 24 months, have you
visited a dentist?
missing Last mammography:
within past 2 years /
within past 2-3 years /
more than 3 years /
never

Last pap smear:
within past 3 years /
within past 3-5 years /
more than 5 years /
never

Germany 2009
When was the last time you saw
a doctor (except dentist)?
1.past 4 weeks
2.1-3month
3.4-12months
4.1-5years
5.>5years
Number of visits with private
practitionners in past 12 months
missing missing missing missing missing missing
Ireland 2007
a. When was the last time you
consulted a GPor f amily
doctor on your own behalf ? [1.
In the last 4 weeks/ 2. Between
1 and 12 mths ago/ 3. 1-2 years
ago/ 4. More than 2 years ago/
5. Never]
missing missing missing When was the last time you
visited a dentist, dental hygienist
or orthodontist on your own
behalf ?
[1. In the last 4 weeks / 2.
Between 1 and 12 mths ago / 3.
1-2 years ago / 4. More than 2
years ago / 5. Never]
missing In the last 12 months, have you
been screened or tested f or
brest cancer - mammogram?
In the last 12 months, have you
been screened or tested f or
cervical cancer?
DELSA/HEA/WD/HWP(2012)1
42


Note : EHIS countries refer to Austria, Belgium, Czech Republic, Estonia, Hungary, Poland, Slovak Republic, and Slovenia.

Country GP visits: probability GP visits: frequency Specialists visits: probability Specialists visits: frequency Dentists visits: probability Dentists visits: frequency
Breast cancer screening
(mammography)
Cervical cancer screening
(pap smear)
New Zealand
2006-07
In the last 12 months, have you
seen a GPabout your own
health?
Number of times In the last 12 months, have you
seen any medical specialists
about your own health?
Number of times When was the last time you
visited? [1.less than 12 months
ago/ 2... 4 more than 12 months /
5. never]
missing Mammography in the past 2
years
Pap smear in the past 3 years
Spain
2009
a. When was the last time you
consulted a GP(general
practitioner) or f amily
doctor on your own behalf ?
[1.less than 4 weeks ago/ 2. 12
months ago/ 3. 12months ago or
more/ 4.never]
b.(if a=1) During the past 4
weeks, how many times?
a. When was the last time you
consulted a medical or surgical
specialist on your
own behalf ? [1.less than 4
weeks ago/ 2. 12 months ago/ 3.
12months ago or more/ 4.never]
b.(if a=1) During the past 4
weeks, how many times?
a. When was the last time you
visited a dentist or orthodontist
on your own
behalf (that is, not while only
accompanying a child, spouse,
etc.)? [1.less than 4 weeks ago/
2. 12 months ago/ 3. 12months
ago or more/ 4.never]
b.(if a=1) During the past 4
weeks, how many times?
Last mammography:
1. within three years
2. more than three years ago
if answer 1 => which date?
(month and year)
Last pap smear:
1. within three years
2. more than three years ago
Switzerland
2007
In the past 12 months, have you
seen a general practitioner?
Number of visits in past 12
months
In the past 12 months, have you
seen a specialist?
Number of visits in past 12
months
Number of visits in past 12
months
Number of visits in past 12
months
Date of the last visit Date of the last visit
UK
2008
Number of visits in past 12
months:
1. none
2. 1-2
3. 3-5
4. 6-10
5. >10
Number of visits in past 12
months:
1. none
2. 1-2
3. 3-5
4. 6-10
5. >10
Number of of visits in past 12
months:
1. none
2. 1-2
3. 3-5
4. 6-10
5. >10
Number of visits in past 12
months:
1. none
2. 1-2
3. 3-5
4. 6-10
5. >10
Health check in past 12 months:
dental
missing Health check-up in past 12
months: breast screen (Data
collected in 2008, 2007)
Health check-up in past 12
months: cervical smear (Data
collected in 2008, 2007, 2006)
USA
2008
Number of of f ice-based
physician visits in 2008
Number of of f ice-based
physician visits in 2008
missing missing Number of dental care visits in
2008
Number of dental care visits in
2008
Last mammography:
Within past year
Within past 2 years
Within past 3 years
Within past 5 years
more than 5 years
Never
Last pap smear:
Within past year
Within past 2 years
Within past 3 years
Within past 5 years
more than 5 years
Never
EHIS
countries
a. When was the last time you
consulted a GP(general
practitioner) or f amily
doctor on your own behalf ?
[1.less than 12 months ago / 2.
12months ago or more / 3.never]
b.(if a=1) During the past 4
weeks, how many times?
a. When was the last time you
consulted a medical or surgical
specialist on your
own behalf ? [1.less than 12
months ago/ 2. 12 months ago or
more/ 3.never]
b.(if a=1) During the past 4
weeks, how many times?
a. When was the last time you
visited a dentist or orthodontist
on your own
behalf (that is, not while only
accompanying a child, spouse,
etc.)? [1.less than 12 months
ago/ 2. 12months ago or more/
3.never]
b.(if a=1) During the past 4
weeks, how many times?
Last mammography:
1. past 12 months
2.12-24months
3.24-36 months
4. >36months
5. never had mammography
Last pap smear:
1. past 12 months
2.12-24months
3.24-36 months
4. >36months
5. never had mammography
DELSA/HEA/WD/HWP(2012)1
43

Table A1.2. Explanatory variables: Need variables and socio-economic characteristics

Country
Self-
assessed
health
Activity limitation Income Education Activity status Region
Degree of
urbanization
Insurance
coverage
Ethnicity
Size of
household
Marital
status
Canada
2007-08
Excellent /
very good /
good / fair /
poor
Impact of health problems
on three main domains:
home, work or school,
and other activities
(sometimes / often /
never)
5 quintiles based on the
adjusted ratio of total
household income to
the low income cut-off
corresponding to the
household and
community size
low / middle /
high
working status : yes / no
11
provinces
missing missing missing 1 / 2 / 3 / 4 / 5+
single /
married / other
Denmark
2005
Very good /
Good / Fair /
Bad / Very
bad
0.no /
1.limited
5 groups of income
low / middle /
high
1. employee / 2. self-
employed / 3. student / 4.
unemployed / 5. retired /
6. homemakers /
7.disabled, other
5 regions missing
supplementary
insurance 0.yes /
1.no, uninsured
missing 1 / 2 / 3 / 4 / 5+
single /
married / other
Finland 2009
Very good /
Good / Fair /
Bad / Very
bad
0.no / 1.limited equivalised income
low / middle /
high
1. employee / 2. student /
3. unemployed / 4. retired
/ 5. homemakers /
6.disabled, other
4 regions
1.densely
/2.intermediate
/3.thinly
missing missing 1 / 2 / 3 / 4 / 5+
single /
married / other
France
2008
Very good /
Good / Fair /
Bad /Very bad
0.no / 1.limited but not
severely / 2.limited
severely
equivalised income
low / middle /
high
1. employee / 2. self-
employed / 3. student / 4.
unemployed / 5. retired /
6. homemakers /
7.disabled, military, other
22 regions
1.densely
/2.intermediate
/3.thinly
complementary
insurance 0.yes /
1.no, uninsured
missing 1 / 2 / 3 / 4 / 5+
single /
married / other
Germany
2009
Very good /
Good / Fair /
Bad / Very
bad
0.no / 1.limited but not
severely / 2.limited
severely
equivalised income
low / middle /
high
1. employee / 2. student /
3. unemployed / 4. retired
/ 5. homemakers /
6.disabled, other
7 regions
1.densely
/2.intermediate
/3.thinly
0. privately insured /
1. publicly insured
missing 1 / 2 / 3 / 4 / 5+
single /
married / other
Ireland 2007
Excellent /
very good /
good / fair /
poor
0.no /
1.limited
5 groups of income
low / middle /
high
1. employee / 2. self-
employed / 3. student / 4.
unemployed / 5. retired /
6. homemakers /
7.disabled, military, other
8 regions
1.densely
/2.intermediate
/3.thinly
Private Health
Insurance :0.
privately insured / 1.
not insured
Medical card: 0. no
card 1. full card or
GP card
1. White /
2. Black /
3. Asian
1 / 2 / 3 / 4 / 5+
single /
married / other
New Zealand
2006-07
Excellent /
very good /
good / fair /
poor
0.no / 1.limited 5 groups of income
low / middle /
high
1.Working in paid work /
2.Not in paid work and
looking for a job / 3.Not in
paid work and not looking
for a job
missing
1.densely
/2.intermediate
/3.thinly
0. insured / 1.
uninsured
1.European -
Other /
2.Maori /
3.Pacific /
4.Asian
1 / 2 / 3 / 4 / 5+
single /
married /
unknown
Need standardisation variables Socio-economic variables
DELSA/HEA/WD/HWP(2012)1
44



Note : EHIS countries refer to Austria, Belgium, Czech Republic, Estonia, Hungary, Poland, Slovak Republic, and Slovenia.
Country
Self-
assessed
health
Activity limitation Income Education Activity status Region
Degree of
urbanization
Insurance
coverage
Ethnicity
Size of
household
Marital
status
Spain 2009
Very good /
Good / Fair /
Bad / Very
bad
0.no / 1.limited but not
severely / 2.limited
severely
Equivalised income
low / middle /
high
1. employee / 2. student /
3. unemployed / 4. retired
/ 5. homemakers /
6.disabled /7. other
19 regions
1.densely
/2.intermediate
/3.thinly
missing missing 1 / 2 / 3 / 4 / 5+
single /
married / other
Switzerland
2007
Very good /
Good / Fair /
Bad / Very
bad
0.no / 1.limited but not
severely / 2.limited
severely
equivalised income
low / middle /
high
1. employee / 2. self-
employed / 3. student / 4.
unemployed / 5. retired /
6. homemakers /
7.disabled, military, other
7 regions
1.urban
/2.rural
type of insurance;
Supplementary
insurance (indicator
based on type of
services received
when hospitalised);
Subsidies f or
insurance payments
missing 1 / 2 / 3 / 4 / 5+
single /
married / other
UK
2008
Excellent /
very good /
good / fair /
poor
0.no / 1.yes equivalised income
low / middle /
high
1. employee / 2. self-
employed / 3. student / 4.
unemployed / 5. retired /
6. homemakers /
7.disabled, military, other
12 regions missing
private medical
insurance 0.yes /
1.no, uninsured
1. White /
2. Black /
3. Asian /
4. Other
1 / 2 / 3 / 4 / 5+
single /
married / other
USA
2008
Excellent /
very good /
good / fair /
poor
no / yes
eqvincome based on
the percentage of
family income relative to
poverty line (adjusted
for the composition and
size of f amily)
years of
education
1 Employed / 2 Self -
employed / 3 student
4 unemployed / 5 retired
6 homemakers / 7
disabled, others
missing missing
uninsured: 0.no /
1.yes, uninsured
1.Hispatic /
2.Black /
3.Asian /
4.Other
1 / 2 / 3 / 4 / 5+
single /
married / other
EHIS
countries
Very good /
Good / Fair /
Bad / Very
bad
no / limited but not
severely / limited severely
10 deciles of household
income, recoded into 5
quintiles
low / middle /
high
1. employee / 2. self-
employed / 3. student / 4.
unemployed / 5. retired /
6. homemakers /
7.disabled, military, other
NUTS-2
1.densely
/2.intermediate
/3.thinly
missing missing 1 / 2 / 3 / 4 / 5+
single /
married / other
Need standardisation variables Socio-economic variables
DELSA/HEA/WD/HWP(2012)1
45
ANNEX 2. METHODOLOGICAL ANNEX
107. This annex presents a overview of the need-standardisation procedure. We used the indirect
standardisation method as described in the appendix of Van Doorslaer and Masseria (2004). More
information on Stata programs can be found in O’Donnel et al. (2008).
108. The indirect standardisation method estimates a health regression
10
such as the following:
[1] ¥
ì
= o + [ X
ì
+ o Z
ì
+e
ì

where Y denotes the dependent variable (e.g. doctor visits of individual in a given period), X a set of
need indicator variables including demographic and morbidity variables, and Z a set of non-need variables
(variables for which we do not want to standardise, but to control for, in order to estimate partial
correlations with the need variables), α, β and δ are parameters vectors, and ε an error term.
109. Equation 1 can be used to generate need-predicted, or X-expected, values of Y. Y
X
represents the
amount of medical care an individual would have received if she/he had been treated as others with the
same need characteristics, on average :
[2] Y
I
X
= α
¯
+ β
´
X
I
+ δ
´
Z
i

where Z

the sample mean values.
110. Estimates of the indirectly need-standardised utilisation, ¥
ì
IS
, are then obtained as the difference
between actual and x-expected utilisation, plus the sample mean ¥

:
[3] ¥
ì
IS
= ¥
ì

ì
X

111. The distribution of ¥
ì
IS
across income can be interpreted as the distribution of health care
utilisation to be expected if need were equally distributed across income.
112. The concentration index of the actual health care utilisation measures the degree of inequality and
the concentration index of the need-standardised utilisation (which is the horizontal inequity index HI)
measures the degree of horizontal inequity. The concentration index of a variable Y can be computed using
a simple “convenient covariance” formula:
[4] C =
2×co¡
w
(¡ì ,Rì)
µ

where µ is the weighted sample mean of Y, covw denotes the weighted covariance and Ri is the
(representatively positioned) relative fractional rank of the ith individual in the income distribution.

10
Linear OLS regression is preferred since, as discussed in Van Doorslaer and Masseria (2004), the alternative of
using intrinsically non-linear regression models does not change the final results.
DELSA/HEA/WD/HWP(2012)1
46
113. Another option to obtain robust estimates for C and its standard error consists in running the
following “convenient regression” of the transformed Y on relative fractional rank:
[5]
2×c
r
2
µ
¥
ì
= o
1
+[
1
R
ì
+e
1,ì

where σ
r
2
is the variance of Ri and β
1
is equal to C, and the estimated standard error of β
1
provides the
estimated standard error of C.
DELSA/HEA/WD/HWP(2012)1
47
ANNEX 3. QUINTILE DISTRIBUTION OF HEALTH CARE UTILISATION
Table A3.1: Quintile distribution of the probability and number of doctor visits after need-standardisation,
inequality index (CI before need-standardisation) and inequity index (HI after need-standardisation)


Note: Significant HI and CI indices highlighted with a * (p<0.05). The surveys in Austria and Ireland do not provide information on
doctor visits.
Country
Sample size
Austria total visits in past 4 weeks
probability of visit in the past 12 months
Belgium total visits in past 4 weeks 0.76 0.70 0.74 0.66 0.69 0.70 -0.124 * -0.013
4392 probability of visit in the past 12 months 0.84 0.83 0.86 0.86 0.88 0.86 -0.007 * 0.011 *
Canada total visits in past 12 months 4.17 4.12 4.14 4.26 4.41 4.22 -0.052 * 0.012 *
101127 probability of visit in the past 12 months 0.81 0.84 0.86 0.88 0.90 0.86 0.008 * 0.019 *
Czech Republic total visits in past 4 weeks 0.62 0.99 0.94 1.11 0.78 0.91 -0.114 * 0.002
1452 probability of visit in the past 12 months 0.79 0.84 0.83 0.90 0.82 0.84 -0.025 * 0.005
Denmark total visits in past 3 months 0.97 0.97 0.89 0.85 0.88 0.92 -0.109 * -0.028 *
12040 probability of visit in the past 3 months 0.47 0.44 0.42 0.42 0.41 0.43 -0.074 * -0.023 *
Estonia total visits in past 4 weeks 0.45 0.64 0.62 0.62 0.61 0.60 -0.102 * 0.029
5833 probability of visit in the past 12 months 0.66 0.72 0.74 0.74 0.78 0.74 -0.011 * 0.027 *
Finland total visits in past 12 months 2.80 2.66 3.10 3.06 3.19 2.96 -0.012 0.039 *
3916 probability of visit in the past 12 months 0.65 0.66 0.70 0.70 0.73 0.69 0.010 0.026 *
France total visits in past 12 months 5.22 4.90 5.20 5.25 5.68 5.27 -0.032 * 0.022 *
10174 probability of visit in the past 12 months 0.88 0.89 0.90 0.92 0.94 0.91 0.009 * 0.013 *
Germany total visits in past 12 months 5.28 5.14 5.24 5.35 5.58 5.30 -0.048 * 0.014
19765 probability of visit in the past 12 months 0.85 0.88 0.89 0.88 0.90 0.88 0.001 0.010 *
Hungary total visits in past 4 weeks 0.88 0.95 0.94 0.88 0.99 0.93 -0.107 * 0.014
4508 probability of visit in the past 12 months 0.80 0.84 0.83 0.85 0.89 0.84 -0.005 0.018 *
Ireland total visits in past 12 months
probability of visit in the past 12 months
New Zealand total visits in past 12 months 4.07 3.76 3.55 3.71 3.83 3.75 -0.088 * -0.003
10629 probability of visit in the past 12 months 0.79 0.80 0.81 0.82 0.86 0.82 -0.005 * 0.016 *
Poland total visits in past 4 weeks 0.86 0.81 0.83 0.86 0.90 0.85 0.027 * 0.116 *
23181 probability of visit in the past 12 months 0.72 0.74 0.75 0.78 0.82 0.77 -0.006 * 0.024 *
Slovak Republic total visits in past 4 weeks 0.85 1.03 1.01 1.06 1.02 1.01 -0.138 * 0.022
4113 probability of visit in the past 12 months 0.78 0.82 0.82 0.85 0.85 0.83 -0.019 * 0.014 *
Slovenia total visits in past 4 weeks 0.67 0.79 0.73 0.65 0.71 0.72 -0.155 * -0.009
1528 probability of visit in the past 12 months 0.77 0.74 0.79 0.77 0.80 0.77 -0.017 * 0.012
Spain total visits in past 4 weeks 0.99 0.92 1.01 1.13 1.21 1.05 -0.010 0.056 *
17253 probability of visit in the past 12 months 0.81 0.83 0.83 0.84 0.85 0.83 -0.004 0.010 *
Switzerland total visits in past 12 months 4.28 4.06 4.11 4.52 4.65 4.32 -0.029 * 0.022
14491 probability of visit in the past 12 months 0.76 0.79 0.78 0.79 0.82 0.79 0.005 0.013 *
United Kingdom total visits in past 12 months (categorical)
11949 probability of visit in the past 12 months 0.78 0.80 0.78 0.79 0.80 0.79 -0.021 * 0.004
United States total visits in past 12 months 2.79 3.10 3.10 3.82 4.19 3.59 0.020 * 0.084 *
22611 probability of visit in the past 12 months 0.56 0.61 0.63 0.70 0.76 0.68 0.044 * 0.060 *
CI HI Poorest
Quintile
2
Quintile
3
Quintile
4
Richest Total
DELSA/HEA/WD/HWP(2012)1
48
Table A3.2: Quintile distribution of the probability and number of GP visits after need-standardisation,
inequality index (CI before need-standardisation) and inequity index (HI after need-standardisation)


Note: Significant HI and CI indices highlighted with a * (p<0.05). The surveys in Germany and the United States do not separately
identify GP and specialist visits.
Country
Sample size
Austria total visits in past 4 weeks 0.57 0.55 0.57 0.53 0.49 0.54 -0.120 * -0.027 *
14951 probability of visit in the past 12 months 0.77 0.78 0.80 0.78 0.80 0.79 -0.013 * 0.005 *
Belgium total visits in past 4 weeks 0.52 0.46 0.50 0.45 0.44 0.46 -0.157 * -0.030 *
4392 probability of visit in the past 12 months 0.79 0.79 0.83 0.82 0.82 0.82 -0.018 * 0.006
Canada total visits in past 12 months 2.96 2.84 2.79 2.74 2.80 2.83 -0.076 * -0.011 *
101127 probability of visit in the past 12 months 0.73 0.76 0.77 0.79 0.81 0.77 0.005 * 0.020 *
Czech Republic total visits in past 4 weeks 0.33 0.54 0.51 0.59 0.40 0.48 -0.168 * -0.008
1452 probability of visit in the past 12 months 0.70 0.74 0.74 0.78 0.68 0.73 -0.051 * -0.009
Denmark total visits in past 3 months 0.84 0.82 0.71 0.72 0.71 0.77 -0.119 * -0.038 *
12040 probability of visit in the past 3 months 0.45 0.41 0.39 0.38 0.37 0.40 -0.083 * -0.034 *
Estonia total visits in past 4 weeks 0.29 0.42 0.40 0.36 0.35 0.36 -0.141 * -0.002
5833 probability of visit in the past 12 months 0.59 0.66 0.68 0.67 0.70 0.67 -0.026 * 0.024 *
Finland total visits in past 12 months 1.80 1.61 1.96 1.91 1.78 1.81 -0.027 0.016
3916 probability of visit in the past 12 months 0.54 0.57 0.59 0.60 0.59 0.58 -0.001 0.019 *
France total visits in past 12 months 3.90 3.65 3.58 3.54 3.46 3.62 -0.076 * -0.023 *
10174 probability of visit in the past 12 months 0.84 0.85 0.86 0.87 0.87 0.86 0.001 0.008 *
Germany total visits in past 12 months
probability of visit in the past 12 months
Hungary total visits in past 4 weeks 0.58 0.61 0.54 0.52 0.56 0.56 -0.139 * -0.017
4508 probability of visit in the past 12 months 0.75 0.77 0.77 0.76 0.78 0.76 -0.022 * 0.007
Ireland total visits in past 12 months
8569 probability of visit in the past 12 months 0.76 0.73 0.71 0.73 0.76 0.74 -0.021 * 0.005
New Zealand total visits in past 12 months 3.34 2.89 2.65 2.84 2.83 2.85 -0.106 * -0.017 *
10629 probability of visit in the past 12 months 0.76 0.78 0.79 0.80 0.84 0.80 -0.006 * 0.018 *
Poland total visits in past 4 weeks 0.57 0.49 0.51 0.51 0.52 0.51 -0.095 * -0.007
23181 probability of visit in the past 12 months 0.68 0.69 0.71 0.72 0.75 0.71 -0.018 * 0.019 *
Slovak Republic total visits in past 4 weeks 0.50 0.58 0.55 0.59 0.57 0.56 -0.140 * 0.016
4113 probability of visit in the past 12 months 0.71 0.77 0.77 0.78 0.79 0.77 -0.029 * 0.014 *
Slovenia total visits in past 4 weeks 0.37 0.55 0.48 0.45 0.41 0.46 -0.174 * -0.014
1528 probability of visit in the past 12 months 0.71 0.67 0.72 0.72 0.73 0.71 -0.028 * 0.012
Spain total visits in past 4 weeks 0.79 0.78 0.76 0.80 0.75 0.78 -0.081 * -0.007
17253 probability of visit in the past 12 months 0.76 0.77 0.77 0.78 0.75 0.77 -0.021 * -0.003
Switzerland total visits in past 12 months 2.60 2.31 2.34 2.33 2.29 2.37 -0.068 * -0.022
14491 probability of visit in the past 12 months 0.65 0.67 0.67 0.65 0.67 0.66 -0.002 0.003
United Kingdom total visits in past 12 months (categorical)
11949 probability of visit in the past 12 months 0.75 0.76 0.76 0.77 0.78 0.76 -0.020 * 0.006
United States total visits in past 12 months
probability of visit in the past 12 months
CI HI Poorest
Quintile
2
Quintile
3
Quintile
4
Richest Total
DELSA/HEA/WD/HWP(2012)1
49
Table A3.3: Quintile distribution of the probability and number of specialist visits after need-standardisation,
inequality index (CI before need-standardisation) and inequity index (HI after need-standardisation)


Note: Significant HI and CI indices highlighted with a * (p<0.05). The surveys in Austria, Germany, Ireland, and the United States do
not provide information on specialist visits.
Country
Sample size
Austria total visits in past 4 weeks
probability of visit in the past 12 months
Belgium total visits in past 4 weeks 0.25 0.23 0.24 0.21 0.27 0.24 -0.067 * 0.023
4392 probability of visit in the past 12 months 0.43 0.46 0.50 0.46 0.58 0.51 -0.001 0.054 *
Canada total visits in past 12 months 1.21 1.29 1.35 1.51 1.60 1.39 -0.004 0.059 *
101127 probability of visit in the past 12 months 0.49 0.52 0.55 0.58 0.63 0.56 0.022 * 0.051 *
Czech Republic total visits in past 4 weeks 0.29 0.44 0.44 0.53 0.39 0.43 -0.050 0.017
1452 probability of visit in the past 12 months 0.50 0.56 0.60 0.62 0.59 0.58 -0.024 0.019
Denmark total visits in past 3 months 0.14 0.15 0.17 0.13 0.17 0.15 -0.061 * 0.020
12040 probability of visit in the past 3 months 0.06 0.08 0.09 0.08 0.10 0.08 -0.011 0.076 *
Estonia total visits in past 4 weeks 0.16 0.22 0.22 0.26 0.27 0.24 -0.046 0.075 *
5833 probability of visit in the past 12 months 0.40 0.46 0.44 0.45 0.52 0.47 -0.007 0.048 *
Finland total visits in past 12 months 1.01 1.05 1.14 1.15 1.41 1.15 0.010 0.076 *
3916 probability of visit in the past 12 months 0.38 0.37 0.41 0.39 0.46 0.40 0.025 * 0.043 *
France total visits in past 12 months 1.42 1.43 1.72 1.77 2.32 1.76 0.048 * 0.106 *
9432 probability of visit in the past 12 months 0.46 0.52 0.55 0.60 0.68 0.57 0.061 * 0.079 *
Germany total visits in past 12 months
probability of visit in the past 12 months
Hungary total visits in past 4 weeks 0.30 0.34 0.40 0.36 0.42 0.37 -0.060 * 0.060 *
4508 probability of visit in the past 12 months 0.54 0.61 0.60 0.57 0.66 0.60 -0.005 0.029 *
Ireland total visits in past 12 months
probability of visit in the past 12 months
New Zealand total visits in past 12 months 1.10 1.08 0.96 1.13 1.17 1.08 -0.063 * 0.021
10629 probability of visit in the past 12 months 0.31 0.29 0.32 0.33 0.36 0.33 -0.030 * 0.040 *
Poland total visits in past 4 weeks 0.29 0.32 0.31 0.35 0.37 0.33 -0.040 * 0.045 *
23181 probability of visit in the past 12 months 0.41 0.45 0.47 0.50 0.57 0.48 0.007 0.057 *
Slovak Republic total visits in past 4 weeks 0.37 0.44 0.46 0.46 0.44 0.44 -0.135 * 0.022
4113 probability of visit in the past 12 months 0.49 0.54 0.54 0.59 0.58 0.55 -0.052 * 0.028 *
Slovenia total visits in past 4 weeks 0.29 0.24 0.26 0.19 0.30 0.26 -0.119 * 0.005
1528 probability of visit in the past 12 months 0.46 0.43 0.46 0.45 0.51 0.46 -0.030 * 0.027
Spain total visits in past 4 weeks 0.19 0.23 0.29 0.36 0.46 0.30 0.122 * 0.196 *
17253 probability of visit in the past 12 months 0.44 0.50 0.53 0.56 0.63 0.53 0.038 * 0.071 *
Switzerland total visits in past 12 months 1.71 1.74 1.77 2.19 2.36 1.95 0.015 0.072 *
14491 probability of visit in the past 12 months 0.47 0.48 0.50 0.54 0.59 0.52 0.022 * 0.049 *
United Kingdom total visits in past 12 months (categorical)
11949 probability of visit in the past 12 months 0.40 0.42 0.42 0.42 0.42 0.42 -0.055 * 0.008
United States total visits in past 12 months
probability of visit in the past 12 months
CI HI Poorest
Quintile
2
Quintile
3
Quintile
4
Richest Total
DELSA/HEA/WD/HWP(2012)1
50
Table A3.4: Dentist visits in the past 12 months, by income quintile, and inequality index (CI)


Note: Significant CI indices highlighted with a * (p<0.05). The survey in Germany does not provide information on dentist’s visit.
Country
Sample size
Austria total visits in past 4 weeks 0.15 0.16 0.17 0.21 0.22 0.19 0.084 *
14951 probability of visit in the past 12 months 0.52 0.55 0.60 0.65 0.70 0.61 0.062 *
Belgium total visits in past 4 weeks 0.10 0.09 0.10 0.09 0.16 0.12 0.131 *
4392 probability of visit in the past 12 months 0.40 0.44 0.55 0.57 0.69 0.58 0.094 *
Canada total visits in past 12 months 0.95 1.19 1.38 1.55 1.67 1.35 0.106 *
101127 probability of visit in the past 12 months 0.45 0.57 0.67 0.74 0.79 0.65 0.104 *
Czech Republic total visits in past 4 weeks 0.20 0.30 0.33 0.47 0.24 0.32 0.011
1452 probability of visit in the past 12 months 0.50 0.64 0.70 0.77 0.78 0.71 0.061 *
Denmark total visits in past 3 months 0.42 0.51 0.47 0.52 0.53 0.49 0.032 *
12040 probability of visit in the past 3 months 0.28 0.36 0.36 0.38 0.40 0.35 0.049 *
Estonia total visits in past 4 weeks
5833 probability of visit in the past 12 months 0.31 0.42 0.45 0.49 0.56 0.48 0.092 *
Finland total visits in past 12 months 1.07 1.20 1.26 1.25 1.46 1.25 0.059 *
3916 probability of visit in the past 12 months 0.49 0.55 0.59 0.60 0.71 0.59 0.069 *
France total visits in past 24 months
9904 probability of visit in the past 24 months 0.64 0.72 0.75 0.79 0.82 0.75 0.048 *
Germany total visits in past 12 months
probability of visit in the past 12 months
Hungary total visits in past 4 weeks 0.15 0.19 0.20 0.22 0.24 0.20 0.088 *
4508 probability of visit in the past 12 months 0.28 0.30 0.37 0.43 0.50 0.37 0.121 *
Ireland total visits in past 12 months
8515 probability of visit in the past 12 months 0.38 0.45 0.53 0.57 0.61 0.52 0.092 *
New Zealand total visits in past 12 months
9730 probability of visit in the past 12 months 0.44 0.45 0.49 0.52 0.60 0.51 0.060 *
Poland total visits in past 4 weeks 0.20 0.19 0.23 0.26 0.31 0.24 0.094 *
23181 probability of visit in the past 12 months 0.27 0.34 0.40 0.48 0.55 0.42 0.120 *
Slovak Republic total visits in past 4 weeks 0.23 0.35 0.37 0.39 0.38 0.35 0.058 *
4113 probability of visit in the past 12 months 0.48 0.66 0.70 0.76 0.76 0.69 0.063 *
Slovenia total visits in past 4 weeks 0.26 0.26 0.28 0.36 0.27 0.28 0.028
1528 probability of visit in the past 12 months 0.42 0.48 0.65 0.60 0.64 0.56 0.080 *
Spain total visits in past 4 weeks 0.14 0.17 0.19 0.23 0.30 0.20 0.151 *
17253 probability of visit in the past 12 months 0.34 0.39 0.44 0.48 0.58 0.45 0.104 *
Switzerland total visits in past 12 months 1.15 1.10 1.13 1.21 1.17 1.15 0.012
11366 probability of visit in the past 12 months 0.61 0.65 0.66 0.67 0.69 0.65 0.025 *
United Kingdom total visits in past 12 months (categorical)
11949 probability of visit in the past 12 months 0.58 0.63 0.69 0.75 0.75 0.69 0.050 *
United States total visits in past 12 months 0.49 0.61 0.79 1.04 1.40 0.99 0.196 *
22611 probability of visit in the past 12 months 0.24 0.27 0.34 0.45 0.58 0.42 0.180 *
CI Poorest
Quintile
2
Quintile
3
Quintile
4
Richest Total
DELSA/HEA/WD/HWP(2012)1
51
Table A3.5: Breast cancer screening participation in the past 2 years in women aged 50-69, by income quintile,
and inequality index (CI)


Note: Breast cancer screening in the past 12 months in Denmark and Ireland. Significant CI indices highlighted with a * (p<0.05). The
surveys in Finland and Germany do not provide information on breast cancer screening.
Country (Sample size) Poorest Quintile 2 Quintile 3 Quintile 4 Richest Total CI
Austria (2465) 0.77 0.79 0.83 0.81 0.82 0.80 0.013 *
Belgium (691) 0.59 0.60 0.71 0.72 0.89 0.74 0.074 *
Canada (9596) 0.61 0.73 0.77 0.78 0.77 0.74 0.038 *
Czech Republic (262) 0.56 0.64 0.71 0.75 0.66 0.67 0.033
Denmark* (2038) 0.13 0.17 0.15 0.12 0.15 0.15 -0.011
Estonia (947) 0.20 0.36 0.29 0.43 0.51 0.36 0.168 *
Finland
France (1571) 0.64 0.73 0.75 0.81 0.85 0.77 0.053 *
Germany
Hungary (856) 0.62 0.60 0.69 0.64 0.70 0.65 0.024
Ireland* (1289) 0.21 0.31 0.36 0.33 0.43 0.33 0.125 *
New Zealand (1587) 0.72 0.72 0.78 0.78 0.81 0.76 0.023 *
Poland (4584) 0.52 0.55 0.59 0.61 0.65 0.59 0.039 *
Slovak Republic (692) 0.56 0.57 0.59 0.63 0.60 0.58 0.022
Slovenia (253) 0.49 0.51 0.57 0.44 0.50 0.50 -0.001
Spain (2545) 0.83 0.85 0.84 0.87 0.87 0.85 0.012 *
Switzerland (2526) 0.50 0.49 0.53 0.54 0.55 0.52 0.029 *
United Kingdom (1975) 0.51 0.53 0.54 0.47 0.54 0.52 0.003
United States (3152) 0.69 0.66 0.68 0.79 0.87 0.78 0.066 *
DELSA/HEA/WD/HWP(2012)1
52
Table A3.6: Cervical cancer screening participation in the past 3 years in women aged 20-69, by income
quintile, and inequality index (CI)


Note: Cervical cancer screening in the past 12 months in Ireland. Significant CI indices highlighted with a * (p<0.05). The surveys in
Finland and Germany do not provide information on cervical cancer screening.
Country (Sample size) Poorest Quintile 2 Quintile 3 Quintile 4 Richest Total CI
Austria (6339) 0.73 0.80 0.81 0.83 0.87 0.82 0.029 *
Belgium (1771) 0.61 0.50 0.69 0.70 0.80 0.71 0.066 *
Canada (18556) 0.69 0.75 0.79 0.81 0.85 0.78 0.039 *
Czech Republic (565) 0.53 0.55 0.64 0.72 0.80 0.68 0.079 *
Denmark (5040) 0.52 0.66 0.79 0.79 0.80 0.71 0.071 *
Estonia (2156) 0.20 0.25 0.28 0.29 0.35 0.30 0.089 *
Finland
France (4092) 0.59 0.68 0.74 0.79 0.81 0.73 0.064 *
Germany
Hungary (1941) 0.52 0.61 0.62 0.66 0.74 0.63 0.062 *
Ireland* (4030) 0.13 0.13 0.14 0.18 0.19 0.16 0.090 *
New Zealand (5053) 0.60 0.67 0.75 0.77 0.82 0.74 0.051 *
Poland (10040) 0.59 0.67 0.69 0.75 0.80 0.72 0.049 *
Slovak Republic (1822) 0.56 0.58 0.62 0.64 0.67 0.61 0.034 *
Slovenia (651) 0.69 0.76 0.82 0.84 0.79 0.78 0.023
Spain (5875) 0.71 0.76 0.79 0.79 0.87 0.79 0.038 *
Switzerland (6183) 0.29 0.31 0.35 0.39 0.37 0.34 0.054 *
United Kingdom (5168) 0.43 0.45 0.46 0.56 0.56 0.50 0.064 *
United States (9363) 0.81 0.78 0.80 0.87 0.89 0.85 0.028 *
DELSA/HEA/WD/HWP(2012)1
53
ANNEX 4. DETAILS ON HEALTH SYSTEMS CHARACTERISTICS
114. Most of the information on health system features was collected through the OECD Health Data
2011 database or through the 2008 Health Systems Characteristics survey (see the questionnaire in Paris et
al. 2010). Some indicators were constructed in order to summarize information such as the three following
indicators: (a) depth of coverage, (b) degree of private provision of care, and (c) regulation of price billed
by provider
115. These three aggregated indicators are described here. To be able to perform the analysis on the
largest number of countries, missing data were estimated by the Secretariat or imputed by the OECD mean.
Depth of coverage
116. The indicator “Depth of coverage” is based on 3 sub-indicators:
• “scope and depth of basic coverage defined by regulation”
• “actual level of coverage by health insurance’
• “out-of-pocket payments for essential care”.
Scope and depth of basic coverage defined by regulation
117. This sub-indicator reflects the “theoretical” level of coverage of the population, or what people
are in principle entitled to. It is based on country replies to questions 13 and 16 of the survey questionnaire.
First, a score was attributed to each function of care, according to the share of typical costs covered by
basic primary coverage (see table below). Then, a global score was computed as the weighted average of
function-related scores.
118. To account for the exemption from copayments (question 16 in the HSC questionnaire), we
added 0.5 to the total score if exemption mechanisms exist for people with high medical needs (people
with certain medical conditions or disabilities or people who have reached an upper limit for out-of-pocket
payments) and 0.5 if exemption mechanisms exist for socially vulnerable people (people whose income is
under a designated threshold or beneficiaries of social benefits).
119. Scoring rules for scope and depth of basic coverage defined by regulation:
DELSA/HEA/WD/HWP(2012)1
54
Table A4.1: Scoring of depth of coverage

Actual level of coverage by health insurance
120. This sub-indicator is based on the share of total health spending financed by any type of health
insurance (public sector, social insurance and private insurance) This share, which varies from 48.9% to
92.6% in OECD countries, was rescaled on a 0 to 6 range to obtain a normalised score.
121. The actual level of coverage results from several parameters: first, the share of population
covered by any type of insurance; second, the share of individual spending covered by any type of
insurance (basic or secondary, public or private) for covered goods and services; and third, the share of
self-consumption of not covered services in health spending. Therefore, for some countries, scores are very
different from those obtained for “scope and depth of coverage”. For instance, in Mexico, the scope and
depth of coverage is high for people covered by health insurance, but the actual level of coverage of the
whole population is lower, notably because many people remain uninsured. For different reasons, the
Slovak Republic has a high score for the scope and depth of coverage defined by regulation, but one of the
lowest score for the actual level of coverage. In this case, the size of the whole population insured and
entitlements are wide, but in reality, out-of-pocket payments are high.
Out-of-pocket payments for essential care
122. The third sub-indicator, “out-of-pocket payments for essential care”, is based on the share of out-
of-pocket payments for inpatient curative care and for basic medical and diagnostic services. This
indicator was chosen as a proxy to measure “real financial accessibility to health care”. Selecting only two
functions of care rather than total spending does not mean the other functions of care are less important.
Instead, it is based on two principles. First, financial access to the first point of entry in medical care
system guarantees that people are more likely consult in case of problems, which increases their
opportunity to receive the diagnosis and treatments they need. Second, acute inpatient care is generally
costly, and even though it corresponds to a real need, and is confirmed by a medical prescription, high
copayments could be a financial hurdle for many households.
123. Data on the share of out-of-pocket payments for inpatient curative care and for basic medical
and diagnostic services were drawn from the System of Health Account 2009 or estimated by the
Secretariat when not available. Country scores are based on the average of both series, weighted by the
OECD average share of each function in total health spending (0.72 for inpatient curative care and 0.28 for
basic medical and diagnostic services
11
).
124. In contrast to the other sub-indicators, a high score reflects a lower level of coverage. In the
aggregated score of “Depth of coverage”, the sub-score “out-of-pocket payments for essential care” has a
negative sign.

11
These weights were estimated from System of Health Account 2009 series.
Level of coverage Score
Covered, no copayment 6
Covered, copayment 75-99% 4.5
Covered, copayments 51-75% 3
Covered, copayments 1-50% 1.5
Not covered 0

DELSA/HEA/WD/HWP(2012)1
55
125. Figure A4.1 shows the relationship between inequity in specialist services use and the depth of
coverage. There is little variation in the depth of coverage across countries (score from 0 to 6) and results
are inconclusive.
Figure A4.1: Relationship between inequity and depth of coverage

Degree of private provision in physicians’ services
126. The degree of private provision of physicians’ services was measured using the predominant
mode of health care delivery, as well as the second mode of delivery when relevant (i.e. representing more
than 20% of services) for both primary care services and specialist services . Separate sub-scores were first
calculated for primary care services and specialist services (see table below) and further aggregated with a
simple average.
127. The degree of private provision of physicians’ services aggregates information on primary care
and out-patient specialists’ services. Countries were asked to provide information on the predominant
mode of delivery and to indicate whether another significant mode exists, i.e. representing more than 20%
of physicians’ contacts. Consequently, when the score indicates that the provision of physicians’ services is
exclusively public (or private), this does not mean that no other mode of provision exists, but only that it
represents a small proportion of physicians’ services.
BEL
CAN
CZE
DNK
EST
FIN
FRA
HUN
NZL
POL
SVK
SVN
ESP
CHE
GBR
y = -0.0019x + 0.0553
R² = 0.0005
0.00
0.02
0.04
0.06
0.08
0.10
4 5 6
I
n
e
q
u
i
t
y

i
n

s
p
e
c
i
a
l
i
s
t

v
i
s
i
t
s

(
H
I
)
Depth of coverage (score 0-6)
DELSA/HEA/WD/HWP(2012)1
56
Table A4.2: Scoring the degree of private provision of physicians’ services

128. The score for the private provision of physicians’ services is calculated as the simple average of
the two previous sub-scores: private provision of primary care services and private provision of out-patient
specialists’ services.
129. Physicians’ services are predominantly provided by the public sector in five countries: Italy,
Portugal, Spain, Sweden and Turkey. At the other end of the spectrum, the supply is exclusively private in
nine OECD countries: Belgium, Denmark, France, Germany, Japan, Korea, Luxembourg, the Netherlands,
Norway and Switzerland.
Regulation of prices billed by providers
130. The data on “prices billed by providers” approximates a measure of “free care at the point of care
delivery”. The indicator is based on the information collected in Table 25 in Paris et al. (2010) on the
determination of prices/fees for specific services (consultation, exam, procedure). It summarises how
prices are set, for both prices paid by third party payers and prices billed by providers.
131. Figure A4.2 shows the relationship between inequity in specialist services use and regulation of
prices billed by specialists suggesting that higher regulation is associated with lower inequity in some
countries (e.g. United Kingdom, and Czech Republic). However, the relationship, if any, is not strong –
New Zealand has the lowest regulation of physician pricing among the countries considered, but has
average inequity; and Hungary has a medium degree of regulation but has one of the lowest levels of
inequity.
Provision of primary care
services
Q27 (predominant mode) Q27b (second mode) Score
Public health care
centres, only
public centres

0
Mix of public and private
provision
public centres
private clinics, private
group/solo practice
3
private clinics, private group/solo practice public centres
Private provision only
private clinics, private group/solo practice

6
private clinics, private group/solo practice
private clinics, private
group/solo practice
Provision of out-patient
specialist services
Q28 (predominant mode) Q28b (second mode) Score
Public health care
centres, only
public centres, public hospitals

0
public centres, public hospitals public centres, public hospitals
Mix of public and private
provision
public centres, public hospitals
private clinics, private
group/solo practice, private
hospitals 3
private clinics, private group/solo practice,
private hospitals
public centres, public hospitals
Private provision only
private clinics, private group/solo practice,
private hospitals
6
private clinics, private group/solo practice,
private hospitals
private clinics, private
group/solo practice, private
hospitals

DELSA/HEA/WD/HWP(2012)1
57
Figure A4.2: Relationship between inequity and regulation of prices billed by physicians

BEL
CAN
CZE
DNK
FIN
FRA
HUN
NZL
POL
SVK
ESP
CHE
GBR
0.00
0.02
0.04
0.06
0.08
0.10
0 1 2 3
H
I

i
n

s
p
e
c
i
a
l
i
s
t

v
i
s
i
t
s
Degree of regulation of prices billed by specialists
Medium - Prices can
sometimes exceed
prices/fees paid by
third-party payers and
statutory copayments
Low regulation-Prices
can always exceed
prices/fees paid by
third-party payers and
statutory copayments
High regulation- Prices
must be equal to
prices/fees paid by
third-party payers +
statutory copayments
OECD mean
DELSA/HEA/WD/HWP(2012)1
58
OECD HEALTH WORKING PAPERS
A full list of the papers in this series can be found on the OECD website: www.oecd.org/els/health/workingpapers
No. 57 THE IMPACT OF PAY INCREASES ON NURSES’ LABOUR MARKET: A REVIEW OF
EVIDENCE FROM FOUR OECD COUNTRIES (2011) James Buchan and Steven Black
No. 56 DESCRIPTION OF ALTERNATIVE APPROACHES TO MEASURE AND PLACE A VALUE ON
HOSPITAL PRODUCTS IN SEVEN OECD COUNTRIES (2011) Luca Lorenzoni and Mark
Pearson
No. 55 MORTALITY AMENABLE TO HEALTH CARE IN 31 OECD COUNTRIES:ESTIMATES AND
METHODOLOGICAL ISSUES (2011) Juan G. Gay, Valerie Paris, Marion Devaux,
Michael de Looper
No. 54 NURSES IN ADVANCED ROLES: A DESCRIPTION AND EVALUATION OF EXPERIENCES
IN 12 DEVELOPED COUNTRIES (2010) Marie-Laure Delamaire and Gaetan Lafortune
No. 53 COMPARING PRICE LEVELS OF HOSPITAL SERVICE ACROSS COUNTRIES: RESULTS OF
A PILOT STUDY (2010) Luca Lorenzoni
No. 52 GUIDELINES FOR IMPROVING THE COMPARABILITY AND AVAILABILITY OF PRIVATE
HEALTH EXPENDITURES UNDER THE SYSTEM OF HEALTH ACCOUNTS FRAMEWORK
(2010) Ravi P. Rannan-Eliya and Luca Lorenzoni
No. 51 EFFECTIVE WAYS TO REALISE POLICY REFORMS IN HEALTH SYSTEMS (2010)
Jeremy Hurst
No. 50 HEALTH SYSTEMS INSTITUTIONAL CHARACTERISTICS A SURVEY OF 29 OECD
COUNTRIES (2010) Valerie Paris, Marion Devaux and Lihan Wei
No. 49 THE CHALLENGE OF FINANCING HEALTH CARE IN THE CURRENT CRISIS (2010) Peter
Scherer, Marion Devaux
No. 48 IMPROVING LIFESTYLES, TACKLING OBESITY: THE HEALTH AND ECONOMIC IMPACT
OF PREVENTION STRATEGIES (2009) Franco Sassi, Michele Cecchini, Jeremy Lauer and Dan
Chisholm
No. 47 HEALTH CARE QUALITY INDICATORS PROJECT: PATIENT SAFETY INDICATORS
REPORT 2009 (2009) Saskia Drösler, Patrick Romano, Lihan Wei; and
ANNEX Saskia Drösler
No. 46 EDUCATION AND OBESITY IN FOUR OECD COUNTRIES (2009) Franco Sassi, Marion
Devaux, Jody Church, Michele Cecchini and Francesca Borgonovi
No. 45 THE OBESITY EPIDEMIC: ANALYSIS OF PAST AND PROJECTED FUTURE TRENDS IN
SELECTED OECD COUNTRIES (2009) Franco Sassi, Marion Devaux, Michele Cecchini and
Elena Rusticelli
No. 44 THE LONG-TERM CARE WORKFORCE: OVERVIEW AND STRATEGIES TO ADAPT SUPPLY
TO A GROWING DEMAND (2009) Rie Fujisawa and Francesca Colombo
DELSA/HEA/WD/HWP(2012)1
59
No. 43 MEASURING DISPARITIES IN HEALTH STATUS AND IN ACCESS AND USE OF HEALTH
CARE IN OECD COUNTRIES (2009) Michael de Looper and Gaetan Lafortune
No. 42 POLICIES FOR HEALTHY AGEING: AN OVERVIEW (2009) Howard Oxley
No. 41 THE REMUNERATION OF GENERAL PRACTITIONERS AND SPECIALISTS IN 14 OECD
COUNTRIES: WHAT ARE THE FACTORS EXPLAINING VARIATIONS ACROSS
COUNTRIES? (2008) Rie Fujisawa and Gaetan Lafortune
No. 40 INTERNATIONAL MOBILITY OF HEALTH PROFESSIONALS AND HEALTH WORKFORCE
MANAGEMENT IN CANADA: MYTHS AND REALITIES (2008) Jean-Christophe Dumont,
Pascal Zurn, Jody Church and Christine Le Thi
No. 39 PHARMACEUTICAL PRICING & REIMBURSEMENT POLICIES IN GERMANY (2008) Valérie
Paris and Elizabeth Docteur
No. 38 MIGRATION OF HEALTH WORKERS: THE UK PERSPECTIVE TO 2006 (2008) James
Buchan, Susanna Baldwin and Miranda Munro
No. 37 THE US PHYSICIAN WORKFORCE: WHERE DO WE STAND? (2008) Richard A. Cooper
No. 36 MIGRATION POLICIES OF HEALTH PROFESSIONALS IN FRANCE (2008) Roland Cash and
Philippe Ulmann
No. 35 NURSE WORKFORCE CHALLENGES IN THE UNITED STATES: IMPLICATIONS FOR
POLICY (2008) Linda H. Aiken and Robyn Cheung
No. 34 MISMATCHES IN THE FORMAL SECTOR, EXPANSION OF THE INFORMAL SECTOR:
IMMIGRATION OF HEALTH PROFESSIONALS TO ITALY (2008) Jonathan Chaloff
No. 33 HEALTH WORKFORCE AND INTERNATIONAL MIGRATION: CAN NEW ZEALAND
COMPETE? (2008) Pascal Zurn and Jean-Christophe Dumont
No. 32 THE PREVENTION OF LIFESTYLE-RELATED CHRONIC DISEASES: AN ECONOMIC
FRAMEWORK (2008) Franco Sassi and Jeremy Hurst
No. 31 PHARMACEUTICAL PRICING AND REIMBURSEMENT POLICIES IN SLOVAKIA (2008)
Zoltán Kaló, Elizabeth Docteur and Pierre Moïse
No. 30 IMPROVED HEALTH SYSTEM PERFORMANCE THROUGH BETTER CARE
COORDINATION (2007) Maria M. Hofmarcher, Howard Oxley, and Elena Rusticelli
No. 29 HEALTH CARE QUALITY INDICATORS PROJECT 2006 DATA COLLECTION UPDATE
REPORT (2007) Sandra Garcia-Armesto, Maria Luisa Gil Lapetra, Lihan Wei, Edward Kelley
and the Members of the HCQI Expert Group
No. 28 PHARMACEUTICAL PRICING AND REIMBURSEMENT POLICIES IN SWEDEN (2007) Pierre
Moïse and Elizabeth Docteur
DELSA/HEA/WD/HWP(2012)1
60
RECENT RELATED OECD PUBLICATIONS
OECD HEALTH DATA 2012 (Database available from http://www.oecd.org/health/healthdata)
OECD REVIEWS OF HEALTH SYSTEMS - RUSSIAN FEDERATION (2012)
OECD REVIEWS OF HEALTH CARE QUALITY – KOREA (2012)
SICK ON THE JOB? MYTHS AND REALITIES ABOUT MENTAL HEALTH AND WORK (2012)
OECD REVIEWS OF HEALTH SYSTEMS - SWITZERLAND (2011)
HEALTH AT A GLANCE 2011: OECD INDICATORS (2011)
See http://www.oecd.org/health/healthataglance for more information
HEALTH REFORM: MEETING THE CHALLENGE OF AGEING AND MULTIPLE MORBIDITIES
(2011)
A SYSTEM OF HEALTH ACCOUNTS 2011 EDITION
HELP WANTED? PROVIDING AND PAYING FOR LONG-TERM CARE (2011)
IMPROVING VALUE IN HEALTH CARE: MEASURING QUALITY (2010)
IMPROVING HEALTH SECTOR EFFICIENCY – THE ROLE OF INFORMATION AND
COMMUNICATION TECHNOLOGIES (2010)
MAKING REFORM HAPPEN – LESSONS FROM OECD COUNTRIES (2010)
HEALTH AT A GLANCE: ASIA/PACIFIC (2010)

HEALTH AT A GLANCE: EUROPE (2010)

VALUE FOR MONEY IN HEALTH SPENDING (2010)

OBESITY AND THE ECONOMICS OF PREVENTION: FIT NOT FAT (2010)

ACHIEVING BETTER VALUE FOR MONEY IN HEALTH CARE (2009), OECD HEALTH POLICY
STUDIES
OECD REVIEWS OF HEALTH SYSTEMS - TURKEY (2009)

For a full list, consult the OECD On-Line Bookstore at www.oecd.org,
or write for a free written catalogue to the following address:
OECD Publications Service
2, rue André-Pascal, 75775 PARIS CEDEX 16
or to the OECD Distributor in your country

Unclassified
Organisation de Coopération et de Développement Économiques Organisation for Economic Co-operation and Development

DELSA/HEA/WD/HWP(2012)1

10-Jul-2012 ___________________________________________________________________________________________ English text only

DIRECTORATE FOR EMPLOYMENT, LABOUR AND SOCIAL AFFAIRS

HEALTH COMMITTEE

DELSA/HEA/WD/HWP(2012)1 Unclassified
English text only

Health Working Papers

OECD Health Working Paper No. 58 INCOME-RELATED INEQUALITIES IN HEALTH SERVICE UTILISATION IN 19 OECD COUNTRIES, 2008-09

Marion Devaux and Michael de Looper

Key words: Inequality, Health care, Private health insurance, Out-of-pocket payments JEL classification: I14, I18

All Health Working Papers are now available through the OECD's Internet Website at http://www.oecd.org/els/health/workingpapers
JT03324670
Complete document available on OLIS in its original format This document and any map included herein are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area.

DELSA/HEA/WD/HWP(2012)1

DIRECTORATE FOR EMPLOYMENT, LABOUR AND SOCIAL AFFAIRS
www.oecd.org/els

OECD HEALTH WORKING PAPERS
http://www.oecd.org/els/health/workingpapers

This series is designed to make available to a wider readership health studies prepared for use within the OECD. Authorship is usually collective, but principal writers are named. The papers are generally available only in their original language – English or French – with a summary in the other. Comment on the series is welcome, and should be sent to the Directorate for Employment, Labour and Social Affairs, 2, rue André-Pascal, 75775 PARIS CEDEX 16, France. The opinions expressed and arguments employed here are the responsibility of the author(s) and do not necessarily reflect those of the OECD.

Applications for permission to reproduce or translate all or part of this material should be made to: Head of Publications Service OECD 2, rue André-Pascal 75775 Paris, CEDEX 16 France Copyright OECD 2012

2

Lien Nguyen from the Finnish National Institute for Health and Welfare for carrying out the analysis for Finland. The authors are very grateful to all for their collaboration in this project. and Hideki Hashimoto for his help on Japan. Finally. 3. All computations on these micro-data were prepared by the authors. Ola Ekholm from the Danish National Institute of Public Health for assisting in data analysis. the New Zealand Ministry of Health for the National Health Survey 2006/07. 3 . and Eurostat for the European Health Interview Survey. the Spanish National Statistics Institute for the Encuesta Europea de Salud 2009. and the responsibility for the use and interpretation of these data is entirely that of the authors. the Irish Social Science Data Archive for the Survey of the Lifestyle. the authors also thank Rie Fujisawa. the Robert Koch Institute for the Gesundheit in Deutschland aktuell 2009. the Institute for Social and Economic Research for the British Household Panel Survey 2009. The authors also thank national data providers for supplying individual-level datasets. the American Agency for Healthcare Research and Quality for the Medical Expenditure Panel Survey 2009. Valérie Paris.DELSA/HEA/WD/HWP(2012)1 ACKNOWLEDGEMENTS 1. Attitudes and Nutrition 2007. namely. The authors thank Albane Gourdol from Eurostat for her assistance with the European Health Interview Survey data analysis. the Swiss Federal Statistical Office for the Swiss Health Survey 2007. the French Institute for Research and Information in Health Economics for the Santé et Protection Sociale survey 2008. 2. The Danish National Institute of Public Health for the National Health Interview Survey 2005. and Divya Srivastava from the OECD Health Division for their comments and suggestions. Mark Pearson. Statistics Canada for the Canadian Community Health Survey 2007/08. Gaetan Lafortune.

i. and the share of out-of-pocket payments for different services). especially specialists. the better-off are more likely to visit doctors. Inequalities in dental visits and breast and cervical cancer screening appear in numerous countries. The paper considers the evolution of inequalities over time by comparing results with the previous study. For doctor visits. although some of the inequalities in health service use cannot be explained by financial barriers. This Working Paper examines income-related inequalities in health care service utilisation in OECD countries. Findings highlight the important effect that the financing of health care services can have on equity (public and private health insurance coverage. Health system financing arrangements are examined to see how these might affect inequalities in health service use. and they visit more often. Findings show that. and adds new results for 6 countries. 4 . With GP contacts. Some discrepancies are found in country ranks for specialist and dentist visits.e. In most countries. The relative position of countries has remained stable for doctor and GP visits over the two studies. and cancer screening. for doctor visits. with the better-off making more use of services. 2004) to 2008-2009 for 13 countries. 5. but these are attributed to methodological differences. for doctor and dentist visits. Quintile distributions and concentration indices were used to assess inequalities. for the same level of need for health care. the scenario is different. It extends a previous analysis (Van Doorslaer and Masseria. the worse-off are as likely as the better-off to contact a GP. 6. horizontal inequities in health care utilisation persist across OECD countries. the extent to which adults in equal need of physician care appear to have equal rates of utilisation.DELSA/HEA/WD/HWP(2012)1 ABSTRACT 4. horizontal equity was assessed. as data permit. After adjustment for needs for health care.

2004) pour 13 pays. les iniquités horizontales dans l’utilisation des soins de santé persistent dans les pays de l’OCDE. et le dépistage du cancer. bien que certaines des inégalités dans l’utilisation des services de santé ne puissent pas être expliquées par des barrières financières. les personnes à hauts revenus ont plus de chances de consulter un médecin et notamment un spécialiste. les plus démunis ont autant de chances que les riches de consulter un généraliste.DELSA/HEA/WD/HWP(2012)1 RESUME 7. favorisant les personnes à hauts revenus. Le cadre d’analyse s’intéresse aux caractéristiques de financement des systèmes de santé et à leurs possibles influences sur les inégalités d’utilisation des services de santé. Cette étude s’intéresse à l’équité horizontale pour les consultations de médecins. La comparaison avec l’étude précédente montre que la position relative des pays est restée stable pour les consultations de docteurs et de généralistes. i. On trouve des différences dans la position des pays pour les consultations de spécialistes et dentistes. Elle examine l’évolution des inégalités en comparant les résultats avec l’étude précédente lorsque les données le permettent. Le scénario est différent pour les visites de généralistes. 9. et part des paiements directs à la charge des patients pour divers services). Les résultats montrent que pour les consultations de médecins. Ce document de travail examine les inégalités liées aux revenus dans l’utilisation des services de santé dans les pays de l’OCDE. Les inégalités sont mesurées à l’aide de distributions par quintile et d’indices de concentration. Les résultats soulignent l’importance des effets du financement des services des soins de santé sur l’équité (assurance santé publique et privée. et ils consultent plus fréquemment. dans quelle mesure des adultes ayant un besoin égal de soins médicaux ont apparemment des taux identiques d’utilisation de soins. portant sur les consultations de médecins et dentistes.e. et inclut 6 nouveaux pays. 5 . 8. pour un même niveau de besoins de soins de santé. notamment dues à aux différences de méthodologies d’enquêtes et de questions. utilisant des données de 2008-2009. Dans la plupart des pays. Après ajustement par les besoins en soins de santé. Il met à jour une étude précédente (Van Doorslaer and Masseria. Des inégalités dans les consultations de dentistes et le dépistage des cancers du sein et du col de l’utérus sont apparentes dans plusieurs pays.

.............................................. QUINTILE DISTRIBUTION OF HEALTH CARE UTILISATION .......................................................................................................... 15 Doctor visits............................... 17 5.................... INEQUALITIES IN HEALTH CARE UTILISATION RELATED TO HEALTH SYSTEM FINANCING ...................................... 53 Depth of coverage ................................. 17 GP visits ..................................DELSA/HEA/WD/HWP(2012)1 TABLE OF CONTENTS ACKNOWLEDGEMENTS ..............4 4.......................................................................................................................................................................................................................................................................... 15 Other explanatory variables ................................... 36 REFERENCES .............................................................................................................................1 4.......................................................................................... 9 2... METHODOLOGICAL ANNEX . DETAILS ON HEALTH SYSTEMS CHARACTERISTICS ..... 25 Country overview .................................................... 14 Estimating income ............................................................. 45 ANNEX 3............. 40 ANNEX 2............................................................................................ 40 ANNEX 1................................ 4 RESUME ..................... 13 Estimating health care utilisation ............................................................................................... DEFINING.............................................................................................................................................................................. 26 4.............................................................................................................6 4................................................................................................................................................3 Public health settings .. METHODS ...................... 56 6 .......................2 5.............................................................. MEASURING AND INTERPRETING HORIZONTAL INEQUITY ............................................................................. 20 Dentist visits ............................................................... 14 Adjusting for health care need .5 4................................ 11 3............................................................................................................ 18 Specialist visits ................ MEASURING INEQUALITIES IN HEALTH CARE UTILISATION: RESULTS ........................2 3........................................................................................................................................................... 47 ANNEX 4...................1 5............................ INTRODUCTION .. 3 ABSTRACT............................. 37 ANNEXES ............................................................. 55 Regulation of prices billed by providers ..................................................................... 13 3....... 22 Cancer Screening ........................1 3.........................2 4......... 24 Comparison with earlier results ........................................................................................................................................................................................................................................................ 5 1........................................................... CONCLUSION ................................................................... 30 Private health insurance .3 4....................................................... 28 5......................................................7 Surveys used ......................................................................4 3................................................................ 28 Out-of-pocket payments ..........................................................................3 3....................................................... 32 6............................................................................. SET OF AVAILABLE INFORMATION IN EACH SURVEY....... 53 Degree of private provision in physicians’ services ...............................................................................................................................................................................................................................................................5 4...................................................................................................

................. 2009 (or latest year) .............................. adjusted for need. inequality index (CI before need-standardisation) and inequity index (HI after needstandardisation) ............................ selected OECD countries.................. women aged 50-69 years.... by income quintile...... 2009 (or latest year) ................ 49 Table A3........................................................................ and inequality index (CI) ........ 24 Figure 11: Inequality index for breast cancer screening in the past 2 years.. 2009 (or latest year) ......... 2009 (or latest year) ...........5: Breast cancer screening participation in the past 2 years in women aged 50-69......................................... 24 Figure 10: Probability of cervical cancer screening in the last three years............................................................................................................ 2009 (or latest year)....................................................... inequality index (CI before need-standardisation) and inequity index (HI after needstandardisation) .......... by income quintile........................................................................................................... 29 Figure 16: Relationship between inequity and the share of out-of-pocket payments............................................ 54 Table A4..........................2: Quintile distribution of the probability and number of GP visits after need-standardisation.............2........................................ 2009 (or latest year) . by income quintile......... 2009 (or latest year) .....................................DELSA/HEA/WD/HWP(2012)1 Tables Table 1... and level of inequality ...............................................................3: Quintile distribution of the probability and number of specialist visits after needstandardisation............................. 2000 and 2009.................. by income quintile......... women aged 20-69 years. by income quintile............................... 2009 (or latest year)................................................................................... 25 Figure 12: Inequality index for cervical cancer screening in the past 3 years............................. 2009 (or latest year) ...................................... inequality index (CI before need-standardisation) and inequity index (HI after need-standardisation) ....... 47 Table A3........................................................... by income quintile......... 27 Table A1....... 41 Table A1...... 48 Table A3..... 2009 (or latest year) .................... Data sources ............................ 31 Figure 17: Share of PHI by income quintile................... and inequality index (CI) .............................................................. by income quintile...... 52 Table A4.................................................................................................... 19 Figure 4: Inequity index for GP visits in the past 12 months.............................1: Quintile distribution of the probability and number of doctor visits after needstandardisation. 56 Figures Figure 1: Needs-adjusted probability of a doctor visit in last 12 months.......4: Dentist visits in the past 12 months.... 2009 (or latest year).... 20 Figure 6: Inequity index for specialist visits in the past 12 months........... 43 Table A3................................. 50 Table A3......................................... 29 Figure 15: Relationship between inequity in specialist visits and the degree of private provision of services .............................. 23 Figure 9: Probability of breast cancer screening in the last two years........... adjusted for need............................2: Scoring the degree of private provision of physicians’ services . 19 Figure 5: Needs-adjusted probability of a specialist visit in last 12 months............................................ Dependent variables ....................................6: Cervical cancer screening participation in the past 3 years in women aged 20-69............... 25 Figure 13: Comparison of inequity indexes.......... 2009 (or latest year) .......................... by income quintile.................................1....................................................................................................1: Scoring of depth of coverage................ 26 Figure 14: Relationship between inequity in doctor visits and the share of public health expenditure ....... 13 Table 2................. and inequality index (CI) ....................... by income quintile......................... 22 Figure 8: Inequality index for dentist visits in the past 12 months.................................................... Explanatory variables: Need variables and socio-economic characteristics ............................... 17 Figure 2: Inequity index for doctor visits in the past 12 months................................ 33 7 .............................. Country overview of probability of a doctor or dentist visit or cancer screening.... 51 Table A3................ 21 Figure 7: Probability of a dentist visit in last 12 months................................ 2009 (or latest year) 18 Figure 3: Needs-adjusted probability of a GP visit in last 12 months........

...............DELSA/HEA/WD/HWP(2012)1 Figure 18: Adjusted probabilities of medical visits by PHI status.............. in selected OECD countries ............................ 34 Figure A4...2: Relationship between inequity and regulation of prices billed by physicians ...................... 57 8 .... 55 Figure A4..1: Relationship between inequity and depth of coverage ........

e. Other studies using European data also find inequities in health care use. Most OECD countries have achieved universal or near-universal coverage of their populations for a core set of health care services. and often visiting these specialists more frequently. 2001. This coverage results from a variety of public insurance arrangements which aim at providing equitable access to care. Preventive screening services for certain cancers such as breast and cervical cancer are generally also available at little or no cost. however. 12. 2002. 11. There is already a substantial body of evidence that inequities exist in the use of certain health care services in many countries. Other services such as dental care and pharmaceutical drugs are often partially covered. This paper extends the analysis to 9 . focuses on equity of access as stated in the principle of horizontal equity. as well as the current paper. INTRODUCTION 10. Van Doorslaer and Masseria (2004) established that. Research on European panel data found pro-poor inequity in GP visits in most EU countries (7 out of 10 countries studied) and pro-rich inequity in specialists in all countries studied (Bago d’Uva et al. i. whereas the picture was less clear-cut for general practitioner (GP) visits. The October 2011 WHO-sponsored World Conference on Social Determinants of Health. It analyses inequities in access to health care through measuring the utilisation of health care services by income level. based on the income level of the family in which the individual lives. This previous work. found education-related inequities for specialist visits in almost all countries studied. Or et al. 2004). has been a focus of previous cross-country comparative work (Van Doorslaer et al. 2011a). adjusted for need. 2006). and health financing that prevents impoverishment (WHO. and tests whether there are any deviations from this principle. Equity of access is a key element of health system performance in OECD countries (Hurst and Jee-Hughes. This paper seeks to update and extend these findings. although there are a number of countries where coverage for these services must be purchased separately (OECD. Dental care was also used more intensively by the better-off.. occupation or educational level. through promotion of accessibility and affordability through universal health care coverage. that people in equal need of health care should be treated equally. and to contain costs. Determining the extent to which OECD countries have achieved the goal of providing equal access for equal need – or equal utilisation. increased copayments or adjustments to the goods and services available through health care coverage have been made. 2008). Kelley and Hurst. Analysing national health interview surveys. saw heads of government renew their determination to achieve health equity. 2012).. based on individual’s need for care. Effective health care coverage provides financial security against expenses due to unexpected or serious illness. 13. The economic crisis has led to increased concerns about access to health care. (2008). with the better-off more likely to see a medical specialist. Van Doorslaer and Masseria. 14. also using data from around the year 2000. as an indicator of access –regardless of individual characteristics such as income.. place of residence. and promotes access to medical goods and services. deviations from the principle of horizontal inequity existed in a number of OECD countries around the year 2000. Several OECD countries have introduced cuts or slow-downs in public health spending.DELSA/HEA/WD/HWP(2012)1 1. using the same methodology as earlier work by Van Doorslaer and colleagues. Access to physician services is ensured at relatively low or no cost for patients.

10 . Belgium. A second objective of the paper is to analyse any findings of inequities in health care use by using selected characteristics of health systems. Germany. Slovak Republic. Canada.DELSA/HEA/WD/HWP(2012)1 new countries and examines the evolution of inequities over time by comparing results with previous findings as data permit. Finland. Estonia. and the 2008 OECD Survey of Health System Characteristics (Paris et al. and to investigate whether these inequities might result from the ways in which health care services are financed. the United Kingdom and the United States. The analysis also broadens the selection of health care services covered in previous studies. 15. The set of countries studied represents a relatively wide geographical spread. France. Czech Republic. and new results are added for six countries (Czech Republic. To do this. 2010). Switzerland. Slovak Republic and Slovenia). Estonia. 16. New Zealand. Ireland. as well as a varied selection of health care systems. New Zealand. Hungary. Poland. Results for 13 countries are updated to 2008-2009 or the nearest available year. 17. including the OECD Health Data database. by examining whether differences exist in preventive screening for female breast and cervical cancers. Poland. The current study covers 19 OECD countries: Austria.. Slovenia. a number of other data collections are utilised. Spain. Denmark.

The CI quantifies the degree of socioeconomic-related inequality of actual medical care use.e. adjustment for need has been made for doctor visits only. When it is positive. 21. A second principle. Interpreting values of the HI is not intuitive. a zero value indicates equality. 2008). Inequality and inequity of care are two key concepts. To statistically equalize needs for the groups or individuals to be compared. 2001. inequality in health care utilisation refers to the differences in use that can be observed between individuals or population groups. place of residence or race. It is this first principle of horizontal equity that the present study uses as a basis for international comparisons. 19.a positive index indicates pro-rich inequality (i. (2002). 22. and it is sensitive to changes in distribution of the population across socioeconomic groups (Wagstaff et al. it indicates equity.. Huber et al. and when it is negative.e. The degree of horizontal inequity is measured by the concentration index of the needsstandardised use.. DEFINING. When the HI is not significantly different from zero. means appropriately treating those people who have differing needs for care. it indicates pro-poor inequity. In this study. 20. and this is here termed the horizontal inequity index. The method used in this paper to describe and measure the degree of horizontal inequity in health care delivery is identical to that used in Van Doorslaer et al. Horizontal inequity indexes were calculated for doctor visits only. with equal need assumed for dental visits and cancer screening. that favours the poor). Studies of health care equity examine whether certain population groups systematically receive different levels of care. with this additional cover being the sole source for sizeable shares of the care package. Koolman and van Doorslaer (2004) have shown that multiplying the value of the HI by 75 gives the proportion of the health variable that would need to be (linearly) redistributed from the richer half to the poorer half of the population (assuming that health 11 . Unlike studies which only examine lowest and highest quintiles. The inequality measure adopted in this paper is the concentration index (CI). Here. 1991). 24. as is often the case for dental health care. MEASURING AND INTERPRETING HORIZONTAL INEQUITY 18. irrespective of individual characteristics such as income. All OECD countries aspire to achieve adequate access to health care for all people on the basis of need. vertical equity. Many countries explicitly endorse equality of access to health care for all people as an objective in their policy documents (Hurst and Jee-Hughes. the CI reflects the experiences of the entire population. or HI. The CI is bounded between -1 and 1. 23. whereas inequity refers to those differences remaining after adjustment for need for health care. a distribution that favours the rich). with the sign indicating the direction of inequality . and a negative index indicates pro-poor inequality (i. Most OECD countries offer options to add to the general public coverage of health costs through complementary or supplementary private coverage. this study uses the average relationship between need and treatment for the population as a whole as the vertical equity “norm” and investigates the extent of systematic deviations from this norm by income level. and Van Doorslaer and Masseria (2004). It compares the observed distribution of health care by income with the distribution of need. Horizontal equity is the principle that requires that people in equal need of health care are treated equally. it indicates pro-rich inequity. taking full advantage of the information on all individuals.DELSA/HEA/WD/HWP(2012)1 2.

increasing visits by poorer people by the same amount.05. a distribution with an index value of zero.e. 12 .DELSA/HEA/WD/HWP(2012)1 inequity favours the rich) to arrive at equity. if pro-rich inequity in doctor visits existed with a HI = 0. equalizing doctor visits across the income distribution requires redistributing 3.75% of all visits made by richer people. For example. i.

results from the most recent national health surveys gathering information on health care visits and socio-economic characteristics were used. as listed in Table 1. EHIS consists of four modules of questions on health status. Estonia.e. called hereafter the EHIS countries). Belgium. and background variables. Slovak Republic and Slovenia.1 Surveys used 25. these came from questions proposed in the European Health Interview Survey (EHIS). Poland. These modules may be implemented at a national level either as one specific survey or as elements of existing surveys (i. METHODS 3. For most European countries. Attitudes and Nutrition in Ireland (SLAN 2007) National Health Survey 2006-07 Europejskie Ankietowe Badanie Zdrowia 2009 (EHIS 2009) Európsky prieskum zdravia 2009 (EHIS 2009) Anketa o zdravju in zdravstvenem varstvu 2007 (EHIS 2007) Encuesta Europea de Salud 2009 Swiss Health Survey 2007 British Household Panel Survey 2009 Medical Expenditure Panel Survey 2008 13 . health care use. For non-EHIS countries. Eight countries have adopted the same question modules on health care use (Austria. Data on health care use were taken from national health interview surveys (see Table 1).DELSA/HEA/WD/HWP(2012)1 3. which was implemented across countries between 2006 and 2009. Hungary. Table 1. 26. Czech Republic. health determinants. national health interview surveys or other household surveys). Data sources Country Austria Belgium Canada Czech Republic Denmark Estonia Finland France Germany Hungary Ireland New Zealand Poland Slovak Republic Slovenia Spain Switzerland United Kingdom United States Survey data Österreichische Gesundheitsbefragung 2006/07 (EHIS 2006/07) Belgium Health Survey 2008 (EHIS 2008) Canadian Community Health Survey 2007/08 European Health Interview Survey in the Czech Republic 2008 (EHIS 2008) National Health Interview Survey 2005 Estonian Health Interview Survey 2006/07 (EHIS 2006/07) Welfare and services Survey (HYPA -survey) 2009 Enquête Santé Protection Sociale 2008 German Telephone Health Interview Survey (GEDA) 2009 European Health Interview Survey 2009 (EHIS 2009) Survey of Lifestyle. A detailed list of available information in each survey can be consulted in Annex 1.

2009. Germany. 31. These are the countries offering nationwide population-based free-access screening. and is of the form: “In the past 12 months have you visited a GP?”. so that the horizontal equity principle can be tested. de Looper and Lafortune. including (among the 19 studied countries) Belgium. This difference in time scales does not allow for cross-country comparison of the frequency of visits. affecting the inclusion age and frequency. Dentist consultations were measured over the past 12 months with the exception of France (past 2 years) and Denmark (past 3 months). GP and specialist visits thus need to be standardised to remove the effect of differing needs for care among persons with different income levels. 3. Slovenia and the United Kingdom (but only England. Some caution is needed in comparing the magnitude of inequalities across countries and over time because of these discrepancies in recall periods. 2011c). Based on previous findings. women aged 20-69 years who had a Pap smear in the past 3 years.. 1 It is not possible to extrapolate because of the different time scales (number of visits in the past 4 weeks versus probability of visiting in the past 12 months). There are also other countries which have national population-based free-access screening but rollout is not nationwide yet. However. Pap smear tests are available through free nationwide population-based programmes in Denmark. since the worse-off visit GPs more often. the recall period refers to the past 12 months in Denmark (for breast cancer screening) and in Ireland (for both indicators). 2008. National guidelines relating to cancer screening may also differ across countries (OECD. 32. However. they have greater needs for health care services (Mackenbach. New Zealand. Both the probability and the frequency of services were measured using sampling weights. Three types of health care services were measured. Finland. measuring inequities in visits over the past 3 months in Denmark would likely over-estimate pro-poor inequities. 2011b). 29. 2011b). For breast cancer screening. 2006. these being the utilisation of (i) physicians (and in most countries further information on separate GP and specialist utilisation are also available). Scotland and Wales) (OECD. Mackenbach et al. France. Hungary. the same age range and frequency was adopted as that used for the OECD Health Data collection. 30. and (iii) breast and cervical cancer screening services for women. They include Denmark and Switzerland. For the number of visits to doctors and dentists. in EHIS countries. Free nationwide population-based screening mammography programmes operate in many countries. Spain and the UK2. Ireland. but inequalities in the probability of a visit can be still measured. it is likely that the probability of the well-off seeing a GP in the past three months is lower than the probability that the worseoff had seen a GP. the focus was on women aged 50-69 years who reported having a mammogram in the past 2 years. individuals were asked how many visits they had made in the past 4 weeks1. a first survey question measures the probability of a visit. (ii) dentists. 2 14 . A second question measures the frequency of visits in the past 12 months. and for cervical cancer screening. 28. Typically. New Zealand. Finland.2 Estimating health care utilisation 27. the recall period was also generally the past 12 months. Since persons in lower socioeconomic groups have higher rates of morbidity. Denmark is an exception since doctor visits were recorded over the past 3 months. To perform international comparisons.DELSA/HEA/WD/HWP(2012)1 3. OECD.3 Adjusting for health care need 33. Thus. Ireland.

or Very bad”. an indirect standardisation method was applied (see Annex 2). the need for doctor visits is proxied by age. some countries differ in their response scales (see Annex 1). Further. Linear regression models were used first. To simplify comparison with the previous study. 1997). and health status variables. such as ethnicity (available for a few countries only). health authorities recommend periodic screening for all women in the target age groups. income was divided into quintiles. New Zealand. New Zealand. For some countries. The categorical variable for activity status distinguishes Employee / Self-employed / Student / Unemployed / Retired / Homemakers / Others 15 .. with equal need across all income levels assumed. only the linear regression estimates are displayed. the United Kingdom and the United States the response options differ slightly: “Excellent. an annual dental visit is recommended for all persons. if at all. Equality is determined for dentist visits and cancer screening using the concentration index CI. In practice. many countries make these services available for free. In Canada. In this study. The need-standardisation procedure also uses other explanatory variables. Health status was measured by two variables: self-assessed health and activity limitation. using the horizontal inequity index HI. and the EHIS countries. 1994). since there is high heterogeneity in individual responses. which is adjusted for need. followed by a two-stage model with logistic regressions for binary outcomes (the probability of a consultation) and 0-truncated negative binomial models for count variables (the frequency of consultations). The activity limitation indicator identifies whether individuals are “Limited. this being only available for Canada and France.. Socio-economic status was measured by household income. Poor”. 37. equivalised using the OECDmodified equivalence scale (Hagenaars et al. most data sources do not provide information on dental care needs among OECD countries. Demographic and socio-economic characteristics are included. Education level was categorised into three groups: Low / Medium / High. in order to estimate partial correlations with the need variables. Very Good..4 Estimating income 39. and for which we do not want to standardise. Fair. Van Doorslaer and Masseria’s (2004) results from linear and non-linear methods also differ little. Wagstaff and Van Doorslaer (2000) showed that HI indices are insensitive to the use of non-linear methods rather than least squares. Again. between the two methods. In most countries. consistent with the previous study. Although the appropriateness of the measure of health care needs may be questioned. the estimated rates by income level and the concentration indexes change only slightly. 3. and marital status. 3. 38. Also. In this paper. For the purpose of the analysis. the size of the household. which is not adjusted for need. Ireland. gender. Neither is needs standardisation performed for breast and cervical cancer screening. 36. Self-assessed health is most often based on a five-response variable in which individuals rate their own general health status as “Very good. equity is determined for doctor visits. 2008). and this may bias the measurement of socio-economic inequalities in health (Tubeuf et al. Good. it was not possible to calculate the equivalised income using the OECD modified equivalence scale. 35. Bad. However. but to control for. In Canada. 2005) and mortality (Idler and Benyamini. it is worth noting that self-assessed health is widely regarded as a good predictor of both health care utilisation (DeSalvo et al. Needs standardisation was not performed for dental consultations. Good. Again. and so household income categories were used for Denmark. Severely limited or Not limited” in their daily activities. only a categorical variable of equivalised income was available.5 Other explanatory variables 40. Chronic conditions as reported in national health surveys was not used. To perform the adjustment for health care needs. Fair.DELSA/HEA/WD/HWP(2012)1 34.

The region variable is based on NUTS-2 for most countries. military).DELSA/HEA/WD/HWP(2012)1 (disabled. The level of urbanisation of the living area distinguishes between three degrees: Dense / Intermediate / Thin. An indicator to identify whether people have public or private health insurance coverage is used when available. The availability of these variables is described in Annex 1. 16 .

In the case of doctor visits. 2007). and the average across the population. by income quintile. However. and so these countries are not presented in Figure 13.DELSA/HEA/WD/HWP(2012)1 4. 2011). The surveys in Austria and Ireland do not include a question on doctor visits (only on GP visits). In most countries.70 0. the gradient is not clear-cut. adjusted for need. 2009 (or latest year) Lowest income quintile France Germany Belgium Canada Czech Republic Hungary Spain Slovak Republic New Zealand United Kingdom Switzerland Slovenia Poland Estonia Finland United States Denmark* 0.40 0.1 Doctor visits 42. The average rates among the total adult population of visiting a doctor in the past 12 months vary across countries. MEASURING INEQUALITIES IN HEALTH CARE UTILISATION: RESULTS 41. the distribution is need-adjusted. Results show income-related equity for doctor visits in Korea with HI=-0. and it appears to be reversed in Denmark. similar studies on Asian countries were undertaken although methodology and data comparability differ. A second chart shows the horizontal inequity index (HI) for both the probability and frequency of visits. Detailed results on the need-standardised distribution of doctor visits (and the split by GP and specialist) across all income quintiles is available in Annex 3. Inequities in doctor visits (GP and specialists together) are presented first.009 (Lu et al. Figure 1: Needs-adjusted probability of a doctor visit in last 12 months.00 Average Highest income quintile Rates of doctor visits in the past 12 months Note: (*) in the past 3 months in Denmark. for the same level of need. and pro-rich inequity in Japan with HI=0. The first chart summarises the distribution of health care use by lowest and highest income quintiles. in the Czech Republic and the United Kingdom. 3 In addition to these selected OECD countries.80 0.50 0. Inequalities in dental visits and cancer screening follow. and then for GP and specialist visits for countries where the data permit this distinction. The distribution of visits to doctors in the last 12 months across income quintiles. Rates of doctor visits also vary by income level.60 0.. 4. high-income people are more likely to visit a doctor than low-income people..0135 (Ikegami et al. 17 . is shown in Figure 1.90 1. from 68% in the United States to 91% in France (rates in Denmark relate to the past three months).

Slovak Republic. In Figure 2 Panel A.1 in Annex 3). i. this would require a redistribution of 74 visits per 1000 population from the richest to the poorest half of the population to achieve equity. In these three OECD countries. adjusted for need. Figure 2: Inequity index for doctor visits in the past 12 months.3. The distribution of visits to GPs in the last 12 months across income quintiles. Poland.DELSA/HEA/WD/HWP(2012)1 43. Estonia. 45. equalizing doctor visits across income quintiles would require redistributing less than 3% of all visits from the richer to the poorer in Finland. According to the redistribution interpretation offered by Koolman and van Doorslaer (2004). the average number of doctor visits in the past 12 months is about 3 per person. where HI= 0. Canada.85 in Poland (see Table A3. In France. Spain and Finland – display a high level of pro-rich inequity in the frequency of doctor’s visits. Only in the United States was a higher level of inequity apparent. people with higher incomes are more likely to visit a doctor than those with lower incomes. people with a lower income were as likely to see a doctor as higher income people. is shown in Figure 3. but not at a high level (Figure 2). Denmark (for which the recall period may over-estimate inequities in favour of the worst-off) displays significant pro-poor inequities. Pro-rich inequities in the probability of a doctor visit are observed in most countries. and up to 9% in Poland. the confidence intervals of the horizontal inequity index for the United Kingdom. Since the average number of doctor visits in the past 4 weeks is about 0. since most of the inequity indices are not significantly different from zero. 5300 visits per 1000 population. (**) counts in the past 4 weeks in EHIS countries. Nonetheless. Czech Republic and Spain. 4. and Finland). the Czech Republic and Slovenia cross the zero line.e. four OECD countries – Poland. 2009 (or latest year) Note: (*) visits in the past 3 months in Denmark.e. whereas the inequity index is higher (HI= 0. 87 doctor visits per 1000 population should be redistributed from the richer to the poorer to achieve equity. the same number of doctor visits (87) should be redistributed to arrive at equity. The 18 . The share of the total adult population seeing a GP in the past 12 months varies across countries. adjusted for need.2 GP visits 47. from 58% in Finland to 86% in France. 850 visits per 1000 population. In Finland. The number of visits to be redistributed is related to the average number of visits. The pattern for the frequency of visits is less clear. and thus. The gradient appears to favour low-income people in Denmark. i. given the same need. for the same level of need for health care. the United States. meaning that the HI is not statistically significantly different from 0. Variations by income level favouring highincome groups are apparent in six countries (New Zealand.04). 44. 46.02 and the average number of doctor visits per person in the past 12 months was 5.

80 0. Switzerland.60 0. Ireland. United Kingdom. Nine countries show no significant inequities in the probability of seeing a GP (Czech Republic. and Slovenia). Finland. Hungary. Seven countries display significant pro-rich inequity (Estonia. the Slovak Republic. Belgium. the poor see a GP more often than the rich. 2009 (or latest year) Lowest income quintile France Belgium New Zealand Austria Canada Slovak Republic Spain Hungary United Kingdom Ireland Czech Republic Poland Slovenia Estonia Switzerland Finland Denmark* 0. (**) counts in the past 4 weeks in EHIS countries.00 Average Highest income quintile Rates of GP visits in the past 12 months Note: (*) visits in the past 3 months in Denmark.30 0. and Canada). by income quintile. Figure 4: Inequity index for GP visits in the past 12 months. the results are quite different. and so these countries are excluded. Figure 3: Needs-adjusted probability of a GP visit in last 12 months. For the same level of health care needs.90 1. New Zealand.70 0. Canada. and six countries display significant pro-poor inequities (Denmark. Belgium. 48. 19 . and France) but the degree is quite small.40 0. Austria. Poland. Austria.DELSA/HEA/WD/HWP(2012)1 surveys in the United States and Germany do not separately identify GP and specialist visits. Only Denmark presents significant pro-poor inequities. France. 2009 (or latest year) Note: (*) visits in the past 3 months in Denmark.50 0. New Zealand. Spain. adjusted for need. 49. The other nine countries display no significant inequity. Figure 4 shows the inequity indexes for GP consultations. With respect to the number of visits.

they consult more often. and Spain). For the frequency of visits. 2009 (or latest year) Lowest income quintile Hungary Czech Republic France Canada Slovak Republic Spain Switzerland Belgium Poland Estonia Slovenia United Kingdom Finland New Zealand Denmark* 0. 54. Concerning frequency of visits.3 Specialist visits 51. 55. high-income people display higher rates of specialist care utilisation. In Spain. 2008). France and Spain are the most inequitable in specialist visits for both probability and frequency concurring with previous findings (Or et al. Hence. the Czech Republic. Among the studied countries. Summarising these findings. Palència et al. other research has found that inequity in specialist visits are largely related to the private sector. after adjusting for need. and they vary substantially (Figure 6 Panel B).40 0. However. The average share of the total adult population seeing a specialist in the past 12 months varies from 33% in New Zealand to 60% in Hungary. Canada.DELSA/HEA/WD/HWP(2012)1 50. Figure 6 shows pro-rich inequities in both the probability and frequency of specialist visits in almost all countries. Finland.60 0. Figure 5: Needs-adjusted probability of a specialist visit in last 12 months. adjusted for need. The 20 . the Czech Republic and Slovenia. they are not significant in Slovenia. In all countries. after need-standardisation. Data from Austria.80 Average Highest income quintile Rates of specialist visits in the past 12 months Note: (*) visits in the past 3 months in Denmark. New Zealand. Slovak Republic and Belgium. they are not significant for the probability of visits in the United Kingdom. The degrees of inequity are much larger than those displayed for GP visits.00 0. is shown in Figure 5. by income quintile. 4. but once they engage with a GP. Estonia. Ireland and the United States do not permit to analyse specialist visits. with the public health system being more equitable (Regidor et al. The distribution of visits to specialists in the last 12 months across income quintiles. Switzerland. the most deprived are generally as likely to see a GP in the past 12 months as the rich. Denmark. HIs are significant for eight countries (Poland. Most countries present large variations in specialist visit rates across income quintiles. France. the well-off are generally more likely to visit a specialist than the poor (Panel A). 2011). and they also do so more often (Panel B). 2008. Hungary.20 0. Germany. 53. 52.

the average number of specialist visits in the past 12 months is about 1. Figure 6: Inequity index for specialist visits in the past 12 months. The quantity to be redistributed to achieve equality would be 66 specialist visits in Finland.7% in Spain. 2009 (or latest year) Note: (*) visits in the past 3 months in Denmark. 1760 visits per 1000 population. (**) counts in the past 4 weeks in EHIS countries. it would be necessary to redistribute 140 visits from the richest to the poorest half of the population to achieve equity.3 in Annex 3).4% in Poland to 14. and 106 in Switzerland.76 (see Table A3. 21 . For France. For every 1000 population.DELSA/HEA/WD/HWP(2012)1 percentage of specialist visits to be linearly redistributed from the richer to the poorer to achieve equity varies from 3.e. i. where this percentage is equal to 8%.

with systematically higher rates among high-income people.40 0. Switzerland. the average number of dentist visits in the past 12 months is 1. The distribution of visits to dentists in the last 12 months across income quintiles is shown in Figure 7.00 Average Highest income quintile Rates of dentist visits in the past 12 months Note: (*) visits in the past 2 years in France. Figure 8 shows large inequalities in the probability of a dentist consultation in favour of the welloff.196. the number of dentist visits to be redistributed for each 1000 population to achieve equality would be 107 in Canada and 145 in the United States. Thus. 22 .DELSA/HEA/WD/HWP(2012)1 4. by income quintile. The percentage of dentist visits to be linearly redistributed from the richer to the poorer to achieve equality is 14. 57.80 1. Figure 7: Probability of a dentist visit in last 12 months.4 Dentist visits 56.20 0. the average number of dentist visits in the past 12 months is 0. 58. In Canada.35 (see Table A3. Equalizing dentist visits across income quintiles requires redistributing fewer than 10% of all visits in the other countries. with the exceptions of the Czech Republic. where the CI equals 0. from 37% in Hungary to 71% in Czech Republic.3% in Spain. All countries display large variations in dentist visit rates across income quintiles. In the United States. large inequalities in the number of dentist visits are apparent. The average rates among the total adult population of seeing a dentist in the past 12 months vary across countries.99. (**) visits in the past 3 months in Denmark. Similarly. 2009 (or latest year) Lowest income quintile Czech Republic United Kingdom Slovak Republic Switzerland Canada Austria Finland Belgium Slovenia Ireland New Zealand Estonia Spain United States Poland Hungary France* Denmark** 0. where the CI equals 0. and Slovenia where the index is not significantly different from zero.106.00 0.4 in Annex 3).7% in the United States and 11.60 0.

Poland and Spain display the largest inequality index. In panel A: (**) visits in the past 24 months in France. 60. Spain and the United States. 2008). High-income persons were more likely to visit a dentist within the last 12 months in all countries. 2009 (or latest year) Note: (*) visits in the past 3 months in Denmark. and the majority of adult population had to use private dental care. Assessments of this policy implementation show that equity and fairness of the oral health care provision system improved (Niiranen et al. 2010). a recent study in Canada shows that access to preventive care is the most pro-rich type of dental care utilisation (Grignon et al. These results confirm findings from a recent study among Europeans aged 50 years and over which identifies significant pro-rich inequalities in access to dental treatment. Listl (2011) finds considerable income-related inequalities in dental service utilisation and attributes these to inequalities in preventive dental visits. In several countries.. however. 2011). Similarly. Before that date. followed by Austria and Belgium (Listl. either alone or in combination with operative treatment. The Finnish government introduced on 1 December 2002 a new policy offering publicly-funded dental care for the whole population. Denmark and France have different recall periods. There is. 23 . Among 14 European countries. wide variation across countries in the concentration index. 59.. The main goals were to offer publicly-funded dental care for the whole population and to equalise access to dental care by socioeconomic groups and by municipalities. In panel B: (**) counts in the past 4 weeks in EHIS countries. 62. and may also have an effect on the level of inequalities in cross-country comparison. recent policy reforms related to the extension of coverage or organisation of dental care have affected access. The distinction between preventive and curative care in dental visits is also relevant. publicly-funded dental care was limited according to age. which impacts on the average probability of dentist visits. Inequalities are larger in countries with a lower probability of a dental visit such as Hungary. Poland.DELSA/HEA/WD/HWP(2012)1 Figure 8: Inequality index for dentist visits in the past 12 months. 61. as mentioned in section 3.2.

Other studies in European countries confirm significant social inequalities in the utilisation of early detection and prevention health care services (Von Wagner et al.80 1. 2009 (or latest year) quintile. 2009 (or latest year) Lowest income quintile Spain Austria United States France New Zealand Canada Belgium Czech Republic Hungary Poland Slovak Republic Switzerland United Kingdom Slovenia Estonia Ireland* Denmark* 0. Ireland. women aged 20-69 years. Breast cancer screening participation in the past two years varies from 36% in Estonia to 85% in Spain. women who are uninsured or receiving Medicaid (health insurance coverage for the poor.00 0. Figure 11 shows that pro-rich inequalities in breast cancer screening exist in almost all countries. and are significant in all countries. Estonia. by income income quintile.. women aged 50-69 years. 2011).80 1. except Slovenia.00 Average Highest income quintile Lowest income quintile United States Austria Spain Slovenia Canada New Zealand France Poland Denmark Belgium Czech Republic Hungary Slovak Republic United Kingdom Switzerland Estonia Ireland* 0. Rates of cancer screening participation among women within the target age group vary widely across countries.5 Cancer Screening 63. 24 . The probabilities of screening for breast cancer in the last two years.00 Rates of breast cancer screening in the past 2 years Rates of cervical cancer screening in the past 3 years Note: (*) visits in the past 12 months in Denmark and Ireland. Canada. Some countries present large inequalities in screening rates across income quintiles. Note: (*) visits in the past 12 months in Ireland. or women with lower educational levels report much lower use of mammography and pap smears (NCHS. In particular. France. 65. low-income women. Caution is needed for all above analyses for the Czech Republic and Slovenia which have large confidence intervals due to small sample sizes. women from higher socioeconomic groups are more likely to have mammograms (Sirven and Or. 2010). and the United States). Poland.40 0. and cervical cancer in the last three years are shown in Figures 9 and 10. and are significant in eight countries (Belgium. and from 30% in Estonia to 85% in the United States for cervical cancer screening. In the United States.60 0. Figure 9: Probability of breast cancer screening in Figure 10: Probability of cervical cancer screening in the the last two years.00 0.20 Average Highest income quintile 0. 64. Likewise.40 0.20 0.60 0. New Zealand. income-related inequalities in cervical cancer screening favour the welloff (Figure 12).DELSA/HEA/WD/HWP(2012)1 4. disabled or impoverished elderly). by last three years. 2011).

04 -0.02 Note: (*) visits in the past 12 months in Denmark and Ireland. There are four countries (Canada. 66.DELSA/HEA/WD/HWP(2012)1 Figure 11: Inequality index for breast cancer screening in the past 2 years. such as free screening access and nationwide population-based programmes was undertaken.25 0. in Switzerland questions on health status and limitations have changed and so it is difficult to use similar information for the “need for care” standardisation procedure. marital status.14 0. good health status. a different data source is used.20 0. residential area. Switzerland. examined 22 European countries. are all recognized as important additional predictors of participation in screening (Sirven and Or. Van Doorslaer and Masseria. the data source is slightly different since the 2004 study used actual medical visits from sickness fund records whereas the current study examines visits reported by individuals.00 -0. Comparing the absolute values of indexes obtained in Van Doorslaer and Masseria’s 2004 study and in this paper should be done with caution. higher level of education.10 -0. having health insurance. The values of the inequity and inequality indices between the two studies often result from surveys which differ slightly in their methodology. 67.10 0. younger age. 2009 (or latest year) 0. many people do not make a distinction between GPs and specialists. 2004). This matters since the quantity of health information in the need standardisation procedure may lead to under.00 0.15 0. such as ethnicity.02 Figure 12: Inequality index for cervical cancer screening in the past 3 years. employment status. In Finland. Other individual characteristics. Palència et al (2010). doctor visits were assessed for the past 12 months in the 2004 study versus the past three months in the current study. the United Kingdom and the United States) where data come from the same source in both studies. and found that inequalities in use of breast and cervical cancer screening are higher in countries without a populationbased screening programme. 1993. Analysis of possible links between inequalities in cancer screening and health system features.6 Comparison with earlier results 68.10 0. 70. having a usual source of care and use of other preventative services.or over-estimation of pro-rich utilisation patterns (Van Doorslaer et al. In another study. Indeed.05 0.08 0. 4. but no clear relationship emerged.05 0. In Denmark. 2009 (or latest year) 0. which may lead to an over-estimation of pro-poor inequity. However. 2010).06 0. for many countries. In France.12 0. and so the general question in the 2000 ECHP survey on 25 . Note: (*) visits in the past 12 months in Ireland. 69.

4. and Spain). Germany.DELSA/HEA/WD/HWP(2012)1 visits to these physicians in the past 12 months might be difficult to understand. notably for dental care. Figure 13: Comparison of inequity indexes. These differences may affect comparisons. for Ireland and Spain. with few exceptions. and is particularly stable regarding doctor and GP visits. an indication is given as to the average probability of a visit in the past 12 months (or 2 years in the case of breast cancer screening. For each type of health care. Belgium. Finland. and in dentist visits because of Ireland. Hungary. France. relative to other countries. Results indicate that. Figure 13 shows that the country ranking is reasonably consistent with the 2004 study. An indication is also given as to the level of equity found in doctor visits after the adjustment for need. questions are more precise. 71. in particular for specialist visits. and 3 years for cervical cancer screening). In the 2009 Finnish HYPA survey. since they ask about visiting a doctor at health centres (public) or at occupational health clinics (used to define GP visits).7 Country overview 72. Denmark and France’s positions. and 26 . different sources of data were used. visiting a doctor at outpatient departments or in private practices (used to define specialist visits). 2000 and 2009 Note: Difference in data sources in 9 countries (Austria. Similarly. The country rank has changed slightly in specialist visits. Finland and Spain’s positions. Denmark. inequities and inequalities have remained stable over time. Ireland. due to Finland. which may cause inconsistencies in results over time. Individual country findings from the analysis are summarised in Table 2.

Equitable n. Higher probability of s creening.a. Equal (breast). these favour the well-off.a. unequal (cervical) Lower probability of screening. Lower probability of a vis it. Unequal Lower probability of a vis it. Inequitable Higher probability of a vis it. Equitable Medium probability of a vis it. Unequal Medium probability of a vis it. Equal (breas t). Inequitable Medium probability of a vis it. Inequitable Medium probability of a vis it. Equal (breast). Lowes t probability of a vis it. Unequal n. the findings for doctors refer to GPs only. Inequitable Medium probability of a (GP) vis it. Unequal Medium probability of a vis it. Inequitable Higher probability of a vis it. Unequal Medium probability of s creening. means not available. Unequal Higher probability of s creening. Equitable Higher probability of a vis it. 27 .DELSA/HEA/WD/HWP(2012)1 equality in dental visits and breast and cervical cancer screening. n. Inequitable Highes t probability of a vis it. unequal (cervical) Lower probability of s creening. unequal (cervical) Medium probability of screening. Mos t unequal n. Unequal Medium probability of a vis it. unequal (cervical) High probability of s creening. Unequal Medium probability of a vis it. Medium probability of screening. Equal (breast). Inequitable Higher probability of a vis it. unequal (cervical) Medium probability of s creening. Unequal Highes t probability of a vis it. Equitable Medium probability of a vis it.a. Country overview of probability of a doctor or dentist visit or cancer screening. Equal High probability of s creening. Table 2. and level of inequality Country Aus tria Belgium Canada Czech Republic Doctor (GP and specialist) visits Higher probability of a (GP) vis it. If inequalities exist. Most unequal Breast and cervical cancer screening High probability of s creening. Unequal Medium probability of a vis it. Unequal Lower probability of a vis it. Equal (breast). unequal (cervical) Medium probability of s creening. Inequitable Lower probability of a vis it. Unequal Higher probability of a vis it. Inequitable Medium probability of a vis it.a. Unequal Higher probability of a vis it. The words equitable/inequitable refer to doctor visits after adjustment for need and. unequal (cervical) Lowes t probability of screening. Unequal Denmark Es tonia Finland France Germany Hungary Ireland New Zealand Poland Slovak Republic Slovenia Spain Switzerland United Kingdom United States Note: For Austria and Ireland. Equal (breas t). Mos t inequitable Dentist visits Medium probability of a vis it. Equal (breast). Unequal Lower probability of a vis it. Unequal Lower probability of a vis it. Unequal Medium probability of screening.a. equal/unequal refer to dentist visits and cancer screening which are not adjusted for need. Equal (breast). Inequitable Lower probability of a vis it. Unequal Medium probability of a vis it. Equitable Lowest probability of vis it. Unequal Medium probability of screening. unequal (cervical) Lower probability of screening. Inequitable Medium probability of a vis it. Unequal Higher probability of a vis it. Inequitable Lower probability of vis it. Unequal n. Unequal Higher probability of a vis it. Unequal Higher probability of a vis it. Unequal Higher probability of s creening.

g. INEQUALITIES IN HEALTH CARE UTILISATION RELATED TO HEALTH SYSTEM FINANCING 73. characteristics related to the organisation of health systems may create barriers. 5. 4 Annex 4 presents the link between inequity and the indicator “depth of coverage” as defined in Paris et al. Health care coverage promotes access to medical goods and services. However. health insurance coverage. A first approach focuses on cultural and information barriers.. Previous studies have found a relationship between the share of public health spending in total health expenditure and lower inequity in doctor consultations (Or et al. is an imperfect indicator of accessibility.. 74. 2010) to explore macro-level relationships between inequality in health services use and the financing of health systems. since the range of services covered and the degree of cost-sharing applied to some services (e. whether GPs act as ‘gatekeepers’ to facilitate access for persons in lower socioeconomic positions. Most OECD countries have achieved universal coverage of health care costs for at least a core set of services. as well as providing financial security against unexpected or serious illness. The results for 2008-09 are complemented by information from the OECD Health Data 2011 database. or whether primary care systems explicitly focus on the needs of the most deprived are two such characteristics (Or. even if universal. in particular for vulnerable people at risk of social exclusion (Huber et al. et al. findings were not conclusive. 76. 77. the direct cost of care to individuals. Thirdly. Secondly. However. since there was little variation in depth of coverage across countries. This section provides further evidence on how the cost of care can create inequalities in health care utilisation. and the 2008 OECD Health Systems Characteristics survey (Paris et al. and are most commonly experienced among the poor and least educated. 2008).DELSA/HEA/WD/HWP(2012)1 5.. which hinder individuals obtaining knowledge of care pathways and seeking care.1 Public health settings 75. Conversely. 2004). sometimes through combinations of public and private health insurance. in the following areas: • • • Public health settings Out-of-pocket payments Private health insurance. 2010)4. 2008).. The distribution of health costs between public and private funds is an important dimension that affects health care access. and particularly whether they have secondary private health insurance.. 2008). (2010). 28 . has most often been associated with inequalities in health care use (van Doorslaer and Masseria. Explaining persisting inequalities in the utilisation of health care services in OECD countries has followed three general approaches. dental care and pharmaceutical drugs) varies substantially across countries (Paris et al. private funding is often regressive and negatively impacts on the uptake of needed services.

02 CHE 0. but still significant relationship.08 ESP 0.04 50 60 SVK POL FIN HUN CAN EST FRA NZL CZE GBR 90 SVN ESPBEL DEU 80 70 Public Health Expenditure as % of Total current health expenditure Linear (all countries) Linear (except the US) Source: OECD estimates for inequity indexes. Figure 14 highlights the relationship between high public financing and low inequity in doctor visits. Data analysis using the 2008 OECD Health Systems Characteristics survey5 lends support to these results. Figure 14: Relationship between inequity in doctor visits and the share of public health expenditure 0. stands out as having a substantially lower share of public health expenditure. publicly funded health care systems are most effective at reducing inequalities in access and reducing the effects on health of income distribution” (Gelormino et al.00 -3 0 Low NZL HUN POL CAN FIN SVK SVN CZE FRA DNK BEL CHE EST Medium 3 6 High Degree of private provision in outpatient specialists services 5 More details on the degree of public/private provision of care in Annex 4.04 R² = 0. Figure 15 shows that countries with a higher degree of private provision of care display higher levels of inequity in specialist consultations.5336 Inequity in doctor visits (HI) 0.2343 0.00 40 -0. The United States. A recent literature review focusing on health system features supports that “nationalized. Or et al.02 DNK -0. its exclusion leads to a weaker.02 GBR 0. and OECD Health Data 2011 79. (2008) highlight that National Health Services where financing and provision of care are handled centrally demonstrate less inequity in specialist visits. Figure 15: Relationship between inequity in specialist visits and the degree of private provision of services 0.06 0. 29 .2146 0.04 R² = 0.06 USA R² = 0.10 Inequity in specialist visits (HI) 0. 2011).DELSA/HEA/WD/HWP(2012)1 78. 80.. with its reliance on primary private health insurance.

2 Out-of-pocket payments 86. 83. creating barriers to access and potentially further damaging health. policies have been implemented to guarantee access to care for people with low-income or high health risks. (2008) used European data to show that social inequalities are stronger in countries where out-of-pocket payments are higher. 2010). Exemptions from out-of-pocket payments for lowincome people exist in half of OECD countries (Paris et al. These aim to relieve the burden of cost-sharing. A question in the 2008 OECD Health System Characteristics survey on “regulation of prices billed by physicians” assesses whether patients must pay extra fees that are not reimbursed when consulting a doctor. Safety nets for the elderly in Japan play a major role in ensuring equity in access.013) (Ikegami et al. Hideki Hashimoto. in people aged 65 and over. 82. underline that out-of-pocket payments increase inequalities in access to care and contribute to impoverishment. for example. 5. 2010). United Kingdom. the relationship is not strong – New Zealand has the lowest regulation of physician pricing among the countries considered. Bureaucratic requirements for reimbursement or lack of knowledge are further disincentives (Huber et al. University of Tokyo School of Public Health. have jeopardised access to a number of health goods and services. and can serve as a proxy of free access at point-of-care. However. 2011). Another financial barrier is related to access at the point of service delivery. Safety nets assist in delivering equity of access by subsidising coverage or providing services for low-income or economically disadvantaged groups. and Hungary has a medium degree of regulation but has one of the lowest levels of inequity (see Annex 4). Results from a recent study suggest pro-rich distribution of health care services6 in the population aged under 65 (HI=0. since the elderly often have a limited income based on their pension. Gelormino et al.. Findings indicate that some countries with high price regulation display lower inequities in specialist care utilisation (e. 2008). An important determinant of the degree of inequality in health care use is the size of household out-of-pocket payments. There is a lack of information on free access at the point-of-care across countries. Almost all countries have implemented policies to protect population groups from usual copayments or excessive out-of-pocket expenses. In those countries where coverage is not automatically provided to all residents through national or local health systems.0067).. Recent increases in the share of health expenditure in the most deprived households in Slovenia. (2011) in reviewing the effects of health care reforms in European countries. or direct provision of health care (Paris et al. dedicated programmes. 7 30 .g. 85. but has average inequity. either through subsidies for the purchase of insurance. and Czech Republic). Direct payments for health care penalise the worst-off. However. The majority of health care service utilisation in Japan consists of outpatient clinic visits. 84. and where the share of public health spending in GDP is smaller. Analysis carried out by Prof.. estimates suggest equity in health care services utilisation (HI=-0. therapeutic appliances. dental services and outpatient specialised health care 6 The question asks about any visit to healthcare services in the past month.DELSA/HEA/WD/HWP(2012)1 Source: OECD estimates for inequity indexes. including medicines. reflecting the reduced copayment systems in place for the elderly in Japan. Paying up-front costs that are reimbursed later creates barriers for low-income households. and OECD Health Systems Characteristics 2008 81. Or et al. including outpatient and inpatient doctor visits as well as to traditional medicine practitioners..

. Specialist visits 0. and System of Health Account 2011 89. with only basic treatment covered by public insurance. 2011).08 0. 8 Out-of-pocket payment in total basic medical and diagnostic services is the closest available measure for health spending related to doctor consultations.04 0. Analysis of out-of-pocket payments in total basic medical and diagnostic services8 in 2008-2009 and the inequity index in doctor visits found no significant relationship.12 Inequity in specialist visits (HI) Inequality in dental visits (CI) Panel B. 90. 31 . many countries have introduced safeguards against excessive out-of-pocket payments.DELSA/HEA/WD/HWP(2012)1 (WHO Europe.04 CZE 0.06 0. a positive association was observed between out-of-pocket payments in dental care and inequality in dentist’s consultations (Figure 16. adult dental care is largely absent from the benefits package. In contrast. To offset their effect. panel A). panel B).2717 HUN ESP POL 0. Similarly. Specialist medical services paid out-of-pocket—typically consultations provided at private medical facilities—is the second largest component of private expenditure. 2011). with correspondingly high out-of-pocket expenses (Garcia-Armesto et al. about twothirds of expenditure is financed privately. Dentist visits 0. In Spain. Long waiting times to access health care has led to a network of informal payments and other queue-jumping mechanisms (OECD.02 0 CAN BEL SVN AUT CZE FIN SVK NZL EST R² = 0. 87. Poland has one of the largest shares of out-of-pocket expenditure on health care among OECD countries. 88.2786 POL 0 20 40 60 80 100 OOP payments as % of total dental expenditure Source: OECD estimates for inequity indexes.1 0. For dental care. lending weight to countries’ effective use of safeguards. the analysis of out-of-pocket payments as a percentage of total expenditure on specialist care shows a significant relationship with the degree of inequity in the probability of specialist visits (Figure 16. Figure 16: Relationship between inequity and the share of out-of-pocket payments Panel A.12 0. A recent initiative proposes a number of policies and interventions to reduce social inequalities (Commission on the Reduction of Social Inequalities in Health in Spain. leading to highly unequal access. 2012).06 FRA 0.02 0 0 10 20 30 40 OOP payments as % of total expenditure on specialist care BEL FIN SVN HUN CHE ESP R² = 0. 2010).08 0.1 0.

However. a supplementary insurance which provides cover for additional health services not covered by the public scheme. Persons in countries with private health insurance markets have the possibility of not only buying more or better care. Ireland. but provides faster private-sector access to medical services where there are waiting times in public systems. or a duplicative insurance that covers services which are already included in the basic benefit package. can access public health services free of charge through a Medical Card. New Zealand. People who are publicly covered can purchase private insurance to cover cost-sharing (complementary PHI) or for additional services (supplementary PHI). PHI can also act as a secondary insurance in addition to public coverage: as a complementary insurance which covers any cost-sharing remaining after basic coverage. the United Kingdom. and the United Stated). to isolate the impact of complementary insurance. opting-out income ceiling). The complementary health insurance is privately funded. 3) bonus insurance. About 47% of the Irish population has private health insurance. About one third of the population. most notably for specialist and dental care. Germany. Multivariate logistic regression models were used to assess the relationship between having PHI and the probability of seeing a doctor. In Ireland. under a certain income threshold. The function of PHI differs across countries (OECD. Private health insurance plays both a complementary and supplementary role in Ireland. about 32% of the population choose a private insurance to supplement what they receive from the public health system with more comprehensive and timely care. with the exception of CMU-C (Couverture Médicale Universelle Complémentaire). Because of the complex role PHI serves across countries. Private health insurance also plays both a complementary and supplementary role in New Zealand.3 Private health insurance 91. more information besides the single dimension of the “share of total PHI” is needed to assess the link with inequity in utilisation. 92. the indicator refers to “having a private health insurance either as a primary coverage or as a secondary source of coverage” vs. basic primary health coverage is automatic and financed from taxes.DELSA/HEA/WD/HWP(2012)1 5. In Germany. About 3% of the population who are not entitled to a Medical Card have a GP Visit Card whereby GP visits are covered by the public system. It can be a primary insurance that represents the only available access to basic health cover because individuals do not have public health insurance. 95. but also of receiving care more quickly. “not privately insured”. 96. and to bypass waiting lists for inpatient services. the self-employed. most of the population (83%) has basic primary coverage through sickness funds. In this analysis. 97. In this analysis. the indicator refers to “having a complementary health insurance” which captures both PHI and CMU-C vs. and others are privately insured (e. some out-of-pocket charges in the primary care sector. Ordinary contracts offer the highest level of financial protection against health care spending but also have the highest level of premiums. 93. Private health insurance (PHI) plays an important role in facilitating access. civil servants. In Switzerland. no complementary insurance. basic primary health care coverage is supplied by national health services. almost all of the population is covered for basic primary health care by the social security system. This analysis was conducted on a set of countries for which national survey data provided information on PHI (France. basic coverage is mandatory and people can choose among four different options for health insurance plans: 1) ordinary basic policies. Switzerland. 94. Other forms 32 . 2004). In France. mainly to cover costs of inpatient bed use.g. In New Zealand. The CMU-C provides care free of charge to the most deprived population. 2) policies with choice of deductibles. 4) insurance with a limited choice of providers. 94% of the French population has complementary health insurance to cover cost-sharing.

Figure 17: Share of PHI by income quintile. 99. “having a public insurance or no insurance”. In Germany. Note: Figures refer to both PHI and CMU-C (public) complementary coverage in France. Despite differences in PHI settings across countries. they represent the share of population “having primary private health insurance”. dental care. which mainly provides access to acute elective care in the private sector. In the United Kingdom. since it provides cover for enhanced services such as faster access and increased consumer choice (Boyle. Supplementary insurance for additional comfort or treatment is purchased by about 30% of the population. 33 Lower income Middle-low Middle Middle-high Higher income 23 United States 42 63 81 89 . 100. the NHS provides preventive medicine. they represent the share of population “having a private health insurance either as a primary coverage or as a secondary coverage”. primary care and hospital services to all residents. Around 12% of the population is covered by a private medical insurance. PHI in the United Kingdom is mainly a supplementary insurance. higher-income people are more likely to purchase PHI (Figure 17). alternative medicines and cash benefits for sickness absence. In this analysis. 2011d). Most people (55% of the population) are covered through private insurance plans. the indicator refers to “having a private primary health insurance” vs.DELSA/HEA/WD/HWP(2012)1 of health insurance contracts offer lower premiums with either higher deductibles or restrictions in the choice of doctor or hospital (OECD. In the United States. usually tied to their employment. 81% of the population has a basic primary health insurance. In the US. 2011). Contracts typically cover one or several of the following benefits: private rooms in hospitals. 2009 (or latest year) 100% Privately insured CMU (France only) 80% Private health insurance 60% 88 94 96 97 97 73 40% 56 51 60 50 62 48 38 14 21 29 36 36 17 23 31 34 20% 40 23 15 21 0% Lower income Middle-low Middle Middle-high Higher income Lower income Middle-low Middle Middle-high Higher income Lower income Middle-low Middle Middle-high Higher income Lower income Middle-low Middle Middle-high Higher income Lower income Middle-low Middle Middle-high Higher income Lower income Middle-low Middle Middle-high Higher income 5 8 14 20 France Germany Ireland New Zealand Switzerland United Kingdom Source: National population-based surveys (see Table 1). selected OECD countries. while Medicare (for those aged over 65) and Medicaid (for the poor) provide public coverage. 98.

For specialist visits. similar results are found9 (Panel C). ethnicity. 34 41% 25% 42% 38% 41% 81% 60% 88% . it refers to “having a private primary health insurance” vs. and Switzerland). education level. health status. Ireland. Regarding GP visits. and the US) people with PHI are more likely to consult a doctor. 102. size of the household. income level. the indicator refers to “having a complementary insurance” (both CMU-C and PHI) vs. Doctor visits 100% Probability of visisting a doctor 90% 80% 70% 94% 88% 91% 90% 90% 85% 88% 87% 60% 50% 40% 30% France Germany New Switzerland United Zealand Kingdom United States 87% 84% Privately insured Not privately insured Panel B. marital status. although with some exceptions. occupation. “no complementary insurance”. Regarding dentist visits. Dentist visits 100% Privately insured Not privately insured Probability of visisting a specialist Probability of visisting a dentist 80% 80% 60% 60% 78% 67% 61% 44% 57% 46% 58% 56% 47% 51% 70% 65% 40% 40% 76% 69% 29% 20% 20% 0% France New Zealand Switzerland United Kingdom 0% France Ireland New Switzerland United Zealand Kingdom United States Source: National population-based surveys (see Table 1). in selected OECD countries Panel A. results show in all 9 Irish data does not provide information on specialist visits. In four out of six countries (France. Figure 18 shows that in most countries people with PHI are significantly more likely to consult doctors and dentists. The effect of PHI is not significant in Germany and the United Kingdom (Panel A).DELSA/HEA/WD/HWP(2012)1 101. Switzerland. the UK displaying no significant difference (Panel B). GP visits 100% 90% 80% 70% Privately insured Not privately insured Probability of visisting a GP 89% 85% 83% 81% 83% 74% 80% 72% 58% 50% 40% 30% France Ireland New Zealand Switzerland 70% United Kingdom Panel C. In France. Specialist visits 100% Privately insured Not privately insured Panel D. Note: Figure 18 shows the adjusted probabilities of medical visits. In Germany. all other things being equal. people with PHI are more likely to consult in four out of five countries (France. region. New Zealand. PHI enhances access to medical goods and services. In the US. The analysis controlled for a range of covariates: age. in particular to secondary and dental care. “having a public primary insurance or no insurance”. it refers to “having a private health insurance either as a primary basic coverage or as a secondary coverage”. gender. Figure 18: Adjusted probabilities of medical visits by PHI status. New Zealand.

and report fewer unmet care needs. specialists and dentists more often. In France. 103. the authors found that people with private health insurance or CMU-C compared to people without complementary insurance consult GPs.. although the CMU-C-insured do not reach the level of utilisation of privately insured people. 2011) As expected. a recent study analysed the complementary insurance and its relationship with access to care (Perronin et al. 35 .DELSA/HEA/WD/HWP(2012)1 studied countries that people having PHI are significantly more likely to consult than those without PHI (Panel D).

Income-related inequalities in breast cancer screening appear in around half of all countries. these being mostly methodological in nature. A greater share of out-of-pocket payments is associated with inequity in specialist and dental care. There is no strong evidence of diminishing inequities although some discrepancies are found in country ranks for specialist and dentist visits. Findings highlight the important effects of certain health system features on equity. The higher the share of public health expenditure. CONCLUSION 104. 2004) shows that the relative position of countries between 2000 and 2009 has remained stable for inequities in doctor and GP visits. the scenario is different. 36 . the worse-off are as likely as the better-off to contact a GP. Secondary private health insurance facilitates the use of care. A comparison with previous results (Van Doorslaer and Masseria. the better-off are more likely to visit doctors especially specialists .than those with lower incomes. and they visit more often.DELSA/HEA/WD/HWP(2012)1 6. Horizontal inequities and inequalities in health care utilisation persist across the 19 OECD countries studied. the lower the inequity in doctor visits. Pro-rich inequalities in dental visits and in cervical cancer screening are present in almost all countries. With GP contacts. After adjustment for needs for health care. In most countries. 105. 106. with a higher rate among the better-off. greater inequity in specialist visits accompanies a higher degree of private provision. with the privately insured more likely to visit doctors and dentists. Broader health insurance coverage improves access. Similarly. for the same level of need for health care.

Vol. International Journal of Health Services. Zaidi (1994). Hurley. 54. Bambra.124. 41(2). Vol. J.. T. De Looper M. Office for Official Publications of the European Communities. Poverty Statistics in the Late 1980s: Research Based on Micro-data. Social Policy and Equality. Vol. United Kingdom (England): Health System Review. and Lafortune G. 1-295. M. K. L. Rockville. Health Policy. “Subjective Unmet Need and Utilization of Health Care Services in Canada: What are the Equity Implications?”. (2005). Le Grand J. Madrid. S. 98. 28. IRDES Question d’économie de la santé No. 465-472. MD. “Social Inequalities in the Use of Health Care Services After 8 years of Health Care Reforms – A Comparative Study of the Baltic Countries”. 209-230. A Proposal of Policies and Interventions to Reduce Social Inequalities in Health in Spain. Spadea. Vol. Kiivet. S. Naiditch and C. M. Vol. C. A. García-Armesto. pp. Hagenaars. Allin (2010). pp. 250–259. Allin. K. 40. Commission on the Reduction of Social Inequalities in Health in Spain (2011).. pp. Bellini. M. “The Effects of Health Care Reforms on Health Inequalities: A Review and Analysis of the European Evidence Base. van Doorslaer. 43. (2010). 37 . DeSalvo. Guillaume. 12(4).. E. 280–289. Int J Public Health.A. 13(1). et al.DELSA/HEA/WD/HWP(2012)1 REFERENCES AHRQ (Agency for Healthcare Research and Quality) (2012). Habicht and A. Health Systems in Transition. 1234–1246. S. (2009). AHRQ. Vol. Jones. Wang and S. E.. pp. Grignon. A. (2010). Kunst (2009). pp. OECD Health Working Paper No.. (2011). Bago d’Uva T. et al. Vol. Directorate General for Public Health and Foreign Health Ministry of Health. Health Services Research. Habicht J. Journal of Health Economics. pp. Luxembourg. Dourgnon P. E. “Predicting Mortality and Healthcare Utilization With a Single Question”. Ordonneau (2007). Costa (2011). 70(3). National Healthcare Disparities Report 2011. “Introducing Gate Keeping in France: First Assessment of the Preferred Doctor Scheme Reform”. Soc Sci Med. Vol.. Grignon. de Vos and M. S. R. Gelormino. “Measuring Disparities in Health Status and in Access and Use of Health Care in OECD Countries”. Moving Forward Equity. “Measurement of Horizontal Inequity in Health Care Utilisation Using European Panel Data”.. Health Systems in Transition. (2009). T. “Inequity in a Market-Based Health System: Evidence From Canada’s Dental Sector”. S. and G. M. Spain: Health System Review. pp. 81-90. Boyle.

DELSA/HEA/WD/HWP(2012)1 Huber, M., A. Stanicole, J. Bremner and K. Wahlbeck (2008), Quality in and Equality of Access to Healthcare Services, European Commission Directorate-General for Employment, Social Affairs and Equal Opportunities, Luxembourg. Hurst, J. and M. Jee-Hughes (2001), “Performance Measurement and Performance Management in OECD Health Systems”, Labour Market and Social Policy Series, Occasional Paper No. 47, OECD, Paris. Idler, E. L. & Y. Benyamini (1997), “Self-rated Health and Mortality: A Review of Twenty-seven Community Studies”, Journal of Health and Social Behaviour, Vol. 38, pp. 21-37. Ikegami N., B.-K. Yoo, H. Hashimoto et al. (2011), “Japanese Universal Health Coverage: Evolution, Achievements, and Challenges”, The Lancet, Vol. 378(9796), pp. 1106-1115. Kelley, E. and J. Hurst (2006), “Health Care Quality Indicators Project: Conceptual Framework Paper”, OECD Health Working Papers, No. 23, OECD Publishing, Paris. Koolman, X. and E. van Doorslaer (2004), “On the Interpretation of a Concentration Index of Inequality”, Health Economics, Vol. 13, pp. 649-656. Listl, S. (2011), “Income-Related Inequalities in Dental Service Utilization by Europeans Aged 50+”, Journal of Dental Research, Vol. 90(6), pp. 717-723. Lu J. R., Leung G. M., Kwon S., et al. (2007), “Horizontal Equity in Health Care Utilization Evidence From Three High-Income Asian Economies”, Social Science & Medicine, 64(1), 199-212. Mackenbach, J. P. (2006), Health Inequalities: Europe in Profile, Department of Health, London. Mackenbach, J. P. et al. (2008), “Socioeconomic Inequalities in Health in 22 European Countries”, New England Journal of Medicine, Vol. 358, pp. 2468-2481. Niiranen, T., E. Widström and T. Niskanen (2008), “Oral Health Care Reform in Finland – Aiming to Reduce Inequity in Care Provision”, BMC Oral Health, Vol. 8:3, doi:10.1186/1472-6831-8-3. National Center for Health Statistics (NCHS) (2011), Health, United States, 2010: With Special Feature on Death and Dying, NCHS, Hyattsville, MD. O'Donnell, O., E. Van Doorslaer, A. Wagstaff and M. Lingelow (2008), Analyzing Health Equity Using Household Survey Data, The World Bank, Washington D.C. OECD (2004), Private Health Insurance in OECD countries, OECD Publishing, Paris. OECD (2011a), Health at a Glance 2011: OECD Indicators, OECD Publishing, Paris. OECD (2011b), “What explains different cancer survival rates?” OECD Background Paper DELSA/HEA/ (2011)6. OECD (2011c), How’s Life? Measuring Well-Being, OECD Publishing, Paris. OECD (2011d), OECD Reviews of Health Systems – Switzerland, OECD Publishing, Paris. OECD (2012), OECD Economic Surveys: Poland 2012, OECD Publishing, Paris. 38

DELSA/HEA/WD/HWP(2012)1 Or, Z., F. Jusot and E. Yilmaz (2008), “Impact of Health Care System on Socioeconomic Inequalities in Doctor Use”, IRDES Working Paper No. 17, IRDES, Paris. Palència L., S. Espelt, M. Rodriguez-Sanz, et al. (2010). “Socio-economic inequalities in breast and cervical cancer screening practices in Europe: influence of the type of screening program”, International Journal of Epidemiology, 39: 757-765.Palència, L., S. Espelt, M. Rodriguez-Sanz, et al. (2011), “Trends in Social Class Inequalities in the Use of Health Care Services Within the Spanish National Health System, 1993-2006”, Eur J Health Econ., Published online: 10 November 2011. Paris, V., M. Devaux and L. Wei (2010), “Health Systems Institutional Characteristics: A Survey of 29 OECD Countries”, OECD Health Working Paper No. 50, OECD Publishing, Paris. Perronin, M., A. Pierre and T. Rochereau (2011), “Complementary Health Insurance in France: WideScale Diffusion but Inequalities of Access Persist”, IRDES Questions d’économie de la santé No.161. Regidor, E., D. Martínez, M. E. Calle, P. Astasio, P. Ortega and V. Domínguez (2008), “Socioeconomic Patterns in the Use of Public and Private Health Services and Equity in Health Care”, BMC Health Services Research, Vol. 8:183. Sirven, N. and Z. Or (2010), “Disparities in Regular Health Care Utilisation in Europe”, IRDES Working Paper No. 37, IRDES, Paris. Tubeuf, S., F. Jusot, M. Devaux and C. Sermet (2008), “Social Heterogeneity in Self-Reported Health Status and Measurement of Inequalities in Health”, IRDES Working Paper No. 12, IRDES, Paris. Van Doorslaer, E., A. Wagstaff and F. Rutten (eds) (1993), Equity in the Finance and Delivery of Health Care: an International Perspective, Oxford University Press, Oxford. Van Doorslaer, E., X. Koolman and F. Puffer (2002), “Equity in the Use of Physician Visits in OECD Countries: Has Equal Treatment for Equal Need Been Achieved?”, Measuring Up: Improving Health Systems Performance in OECD Countries, OECD, Paris, pp. 225-248. Van Doorslaer, E. and C. Masseria (2004), “Income-related Inequality in the Use of Medical Care in 21 OECD Countries”, OECD Health Working Paper No. 14, OECD, Paris. Von Wagner, C. et al. (2011), “Inequalities in Participation in an Organized National Colorectal Cancer Screening Programme: Results from the First 2.6 Million Invitations in England”, International Journal of Epidemiology, Vol. 40, pp. 712-718. Wagstaff, A., P. Paci and E. Van Doorslaer (1991), “On the Measurement of Inequalities in Health”, Soc Sci Med, Vol. 33(5), pp. 545-557. Wagstaff, A. and E. Van Doorslaer (2000), “Equity in Health Care Finance and Delivery”, In: A. J. Culver and J. P. Newhouse (eds.), Handbook of Health Economics, Volume 1, Elsevier. WHO (2012), World Conference on Social Determinants of Health, www.who.int/sdhconference/en/. WHO Europe (2011), Health Inequalities in Slovenia, National Institute of Public Health, Ljubljana.

39

DELSA/HEA/WD/HWP(2012)1

ANNEXES

ANNEX 1. SET OF AVAILABLE INFORMATION IN EACH SURVEY

40

1-5years 5.DELSA/HEA/WD/HWP(2012)1 Table A1. have you been screened or tested for brest cancer . have you visited a dentist at health centers months or in private practices? in past 24 months. have you seen. disorder. orthopaedist. orthopaedist.4-12months 4. emotional or mental health)? Dentists visits: frequency In the past 12 months.any other medical doctor or specialist such as a surgeon. or talked to any of the follow ing health professionals about your physical.>=5 years Cervical cancer screening (pap sm ear) Have you ever had a pap smear test? Yes/ No When w as the last time? 1. 3-<5 years 5. have you visited a specialist? Have you consulted other health Number of times care providers during the past 3 months? Yes. dental hygienist or orthodontist (about your physical. When w as the last time you consulted a GP or family doctor on your ow n behalf? [1. 1-<3 years 4. my general practitioner Number of times Finland 2009 Number of visits in past 12 in past 12 months have you visited a doctor at health centers months (public) or at occupational health clinics ? in past 12 months have you visited a GP? Number of visits in past 12 months During the past 3 months. or talked on the telephone. or talked to an eye specialist. 6months-<1year 3.<6months ago 2. emotional or mental health w ith a dentist or orthodontist? Breast cancer screening (m am m ography) Have you ever had a mammogram. about your physical. 6months-<1year 3. Dependent variables Country GP visits: probability In the past 12 months.<6months ago 2. have you visited a doctor at outpatient departments or in private practices? in past 12 months. a practising specialist physician in past 12 months. Denm ark 2005 illness or injury? Yes. Betw een 1 and 12 mths ago / 3. allergist. or talked to a dentist. such as an ophthalmologist or optometrist (about your physical. Never] missing missing missing missing missing Ireland 2007 missing When w as the last time you visited a dentist.any other medical doctor or specialist such as a surgeon. emotional or mental health w ith a family doctor (pediatrician) or general practitioner? Specialists visits: probability Specialists visits: frequency In the past 12 months. 1-<2 years 4. 2-<5 years 5. 1-2 years ago/ 4. More than 2 years ago/ 5. Betw een 1 and 12 mths ago/ 3. how many times have you seen.>=5 years Canada 2007-08 During the past 3 months. how many times have you seen. emotional or mental health: a family doctor (pediatrician) or general practitioner? GP visits: frequency In the past 12 months. allergist. about your physical. [gynaecologist/urologist] or psychiatrist (about your physical. Never] In the last 12 months. In the last 4 w eeks/ 2.1.past 4 w eeks Germ any 2009 2. a dentist Mammography in the past 12 months Pap smear in the past 3 years Number of visits in past 12 months Number of visits in past 12 in past 12 months. emotional or mental health)? . that is. or talked on the telephone. dental hygienist or orthodontist on your ow n behalf? [1. emotional or mental health w ith an eye specialist (such as an ophthalmologist or optometrist)? . [gynaecologist/urologist] or psychiatrist (about your physical. have you seen. disorder. a breast xray? Yes / No When w as the last time? 1. how many times have you seen. or talked on the telephone.1-3month 3. More than 2 years ago / 5. have you consulted a physician because of disease. about your physical.>5years missing a. have you consulted a physician because of disease. have you visited a dentist? missing missing missing Number of visits in past 12 months France 2008 Last mammography: w ithin past 2 years / w ithin past 2-3 years / more than 3 years / never missing Last pap smear: w ithin past 3 years / w ithin past 3-5 years / more than 5 years / never missing missing When w as the last time you saw Number of visits w ith private practitionners in past 12 months a doctor (except dentist)? 1. have you seen.mammogram? In the last 12 months. emotional or mental health)? Number of times Dentists visits: probability In the past 12 months. have you been screened or tested for cervical cancer? 41 . illness or injury? Yes. 1-2 years ago / 4. emotional or mental health)? In the past 12 months. In the last 4 w eeks / 2.

DELSA/HEA/WD/HWP(2012)1 Country New Zealand 2006-07 GP visits: probability In the last 12 months. When w as the last time you b. 3-5 4. 6-10 5. 6-10 5. 1-2 3. When w as the last time you b. >36months 5.(if a=1) During the past 4 w eeks. 6-10 5. not w hile only accompanying a child. have you seen a GP about your ow n health? GP visits: frequency Number of times Specialists visits: probability Specialists visits: frequency In the last 12 months. never] a.(if a=1) During the past 4 consulted a medical or surgical w eeks. 3-5 4. 4 more than 12 months / 5.12-24months 3. 12 months ago/ 3.less than 12 months ago / 2. >10 Number of office-based physician visits in 2008 Number of visits in past 12 months: 1. 1-2 3. 3-5 4.less than 4 w eeks ago/ 2. 6-10 5. and Slovenia. 12 months ago/ 3. >10 missing Number of visits in past 12 months: 1.never] a. Belgium. etc. 42 . 1-2 3. 12 months ago/ 3. have you Number of visits in past 12 seen a specialist? months Date of the last visit Date of the last visit UK 2008 Number of visits in past 12 months: 1. 1-2 3. 12months ago or more/ 3. 12months ago or more/ 4..never] Number of visits in past 12 months Number of visits in past 12 months Breast cancer screening (m am mography) Mammography in the past 2 years Cervical cancer screening (pap smear) Pap smear in the past 3 years Spain 2009 a. more than three years ago if answ er 1 => w hich date? (month and year) Last pap smear: 1. how many times? consulted a GP (general practitioner) or family doctor on your ow n behalf? [1.)? [1. past 12 months 2. 12months ago or more / 3..less than 4 w eeks ago/ 2.(if a=1) During the past 4 w eeks. 2006) Number of dental care visits in 2008 Number of dental care visits in 2008 USA 2008 Last mammography: Within past year Within past 2 years Within past 3 years Within past 5 years more than 5 years Never Last mammography: 1. spouse. more than three years ago Sw itzerland 2007 In the past 12 months.24-36 months 4. >36months 5. 12months ago or more/ 4. 2007) Health check-up in past 12 months: cervical smear (Data collected in 2008.less than 4 w eeks ago/ 2. never had mammography Last pap smear: Within past year Within past 2 years Within past 3 years Within past 5 years more than 5 years Never Last pap smear: 1. none 2.never] In the past 12 months.less than 12 months ago/ 2. none 2. Poland. have you seen any medical specialists about your ow n health? Number of times Dentists visits: probability Dentists visits: frequency When w as the last time you missing visited? [1.24-36 months 4.never] b. none 2. When w as the last time you b. how many times? specialist on your ow n behalf? [1. 12 months ago or more/ 3. not w hile only accompanying a child. When w as the last time you b. how many times? Note : EHIS countries refer to Austria. how many times? consulted a GP (general practitioner) or family doctor on your ow n behalf? [1. Estonia. w ithin three years 2. have you Number of visits in past 12 seen a general practitioner? months a. etc. past 12 months 2. never had mammography EHIS countries a. w ithin three years 2. When w as the last time you b. >10 Number of office-based physician visits in 2008 Number of of visits in past 12 months: 1.(if a=1) During the past 4 w eeks. 2007. none 2.)? [1. Slovak Republic. how many times? specialist on your ow n behalf? [1.less than 12 months ago/ 2. >10 missing Health check in past 12 months: missing dental Health check-up in past 12 months: breast screen (Data collected in 2008.never] a.(if a=1) During the past 4 consulted a medical or surgical w eeks. 12months ago or more/ 4. Czech Republic. Hungary. When w as the last time you visited a dentist or orthodontist on your ow n behalf (that is.12-24months 3. spouse. how many times? on your ow n behalf (that is.never] Last mammography: 1. 3-5 4.less than 12 months ago/ 2.(if a=1) During the past 4 visited a dentist or orthodontist w eeks.

uninsured 1. military. Black / Medical card: 0. employee / 2. insured / 1. publicly insured 1 / 2 / 3 / 4 / 5+ single / married / other Ireland 2007 Excellent / very good / good / fair / poor 0. employee / 2. 1. selfemployed / 3.Working in paid w ork / 2.Asian 43 .limited Bad /Very bad severely equivalised income 1. student / 4.no / 1.intermediate /3.Maori / 1 / 2 / 3 / 4 / 5+ married / 3.Pacific / unknow n 4. retired / 8 regions high 6. uninsured missing 1 / 2 / 3 / 4 / 5+ single / married / other Finland 2009 0. full card or GP card 1 / 2 / 3 / 4 / 5+ single / married / other New Zealand 2006-07 Excellent / very good / good / fair / poor 0.limited 5 groups of income 1.limited 5 groups of income 1.no / 1. homemakers / 7.densely /2. privately insured / 1. homemakers / 6.intermediate /3. no 3. selfemployed / 3. unemployed / 4.limited severely equivalised income 1. w ork or school. unemployed / 4.disabled.limited 5 groups of income 1.thinly missing missing 1 / 2 / 3 / 4 / 5+ single / married / other France 2008 Very good / 0. Explanatory variables: Need variables and socio-economic characteristics Need standardisation variables Country Selfassessed health Activity lim itation Incom e Education Activity status Socio-econom ic variables Region Degree of urbanization Insurance coverage Ethnicity Size of household Marital status Canada 2007-08 Excellent / very good / good / fair / poor Impact of health problems on three main domains: home.yes / 1. retired 4 regions high / 5. low / middle / unemployed / 5. other 1. White / not insured 2.yes / 1.densely /2.disabled.no / 1.limited equivalised income 1.intermediate /3.Not in missing high paid w ork and not looking for a job 1. low / middle / unemployed / 5. employee / 2. homemakers / 7.DELSA/HEA/WD/HWP(2012)1 Table A1. retired 7 regions high / 5. homemakers / 7.densely /2. selfemployed / 3. other 1.Not in paid w ork and low / middle / looking for a job / 3. Asian card 1. other missing supplementary insurance 0. and other activities (sometimes / often / never) 5 quintiles based on the adjusted ratio of total household income to low / middle / w orking status : yes / no the low income cut-off high corresponding to the household and community size 11 provinces missing missing missing 1 / 2 / 3 / 4 / 5+ single / married / other Denm ark 2005 Very good / Good / Fair / Bad / Very bad Very good / Good / Fair / Bad / Very bad 0. homemakers / 6.limited but not Good / Fair / severely / 2.limited but not severely / 2.no / 1. privately insured / missing 1. low / middle / unemployed / 5.no. other 1.European Other / single / 2.thinly 0.no. student / 4.thinly 0. retired / 22 regions high 6. employee / 2.disabled. military.densely /2. retired / 5 regions high 6.2.thinly Private Health Insurance :0. student / 4. student / low / middle / 3.disabled.intermediate /3.no / 1.disabled.thinly complementary insurance 0.densely /2. student / low / middle / 3.intermediate /3. employee / 2. uninsured missing 1 / 2 / 3 / 4 / 5+ single / married / other Germ any 2009 Very good / Good / Fair / Bad / Very bad 0.no / 1. other 1.

intermediate /3. Supplementary insurance (indicator based on type of missing services received w hen hospitalised).urban /2. student / 4.thinly Insurance coverage Ethnicity Size of household Marital status Spain 2009 Equivalised income 1.disabled /7.rural type of insurance. military. other 1.no / 1. employee / 2. other 1 Employed / 2 Selfemployed / 3 student 4 unemployed / 5 retired 6 homemakers / 7 disabled.no / 1. low / middle / unemployed / 5.limited but not severely / 2.disabled. retired / 7 regions high 6. selfemployed / 3. military.yes / 1.no / 1. others missing private medical insurance 0. other missing missing 1 / 2 / 3 / 4 / 5+ single / married / other Sw itzerland 2007 Very good / Good / Fair / Bad / Very bad 0. unemployed / 4. low / middle / unemployed / 5. uninsured 1. White / 2. military. homemakers / 7.intermediate /3.Black / 3.disabled. student / low / middle / 3. 44 . homemakers / 7. student / 4. employee / 2.Other 1 / 2 / 3 / 4 / 5+ single / married / other EHIS countries Very good / Good / Fair / Bad / Very bad 1. Estonia. retired 19 regions high / 5. Hungary.limited severely equivalised income 1. other 1.limited severely Incom e Education Activity status Socio-econom ic variables Region Degree of urbanization 1. homemakers / 6.no.densely /2. student / 4.DELSA/HEA/WD/HWP(2012)1 Need standardisation variables Country Selfassessed health Very good / Good / Fair / Bad / Very bad Activity lim itation 0. retired / 12 regions high 6. Other 1 / 2 / 3 / 4 / 5+ single / married / other USA 2008 Excellent / very good / good / fair / poor no / yes eqvincome based on the percentage of family income relative to years of poverty line (adjusted education for the composition and size of family) missing missing uninsured: 0. Belgium. Slovak Republic. recoded into 5 NUTS-2 severely / limited severely high 6.yes equivalised income 1. homemakers / quintiles 7.Hispatic / 2. retired / income.disabled. Black / 3.limited but not severely / 2. employee / 2. selfemployed / 3. employee / 2.Asian / 4.densely /2. Subsidies for insurance payments 1 / 2 / 3 / 4 / 5+ single / married / other UK 2008 Excellent / very good / good / fair / poor 0. Asian / 4. selfemployed / 3.thinly missing missing 1 / 2 / 3 / 4 / 5+ single / married / other Note : EHIS countries refer to Austria. 10 deciles of household no / limited but not low / middle / unemployed / 5.no / 1. Czech Republic. uninsured 1. and Slovenia. Poland.yes.

X a set of need indicator variables including demographic and morbidity variables. on average : [2] Y where The indirect standardisation method estimates a health regression10 such as the following: α βX δZ the sample mean values. the alternative of using intrinsically non-linear regression models does not change the final results. 45 . β and δ are parameters vectors. or X-expected. [1] where Y denotes the dependent variable (e. and Z a set of non-need variables (variables for which we do not want to standardise. where µ is the weighted sample mean of Y. Equation 1 can be used to generate need-predicted. values of Y. are then obtained as the difference 110. The distribution of across income can be interpreted as the distribution of health care utilisation to be expected if need were equally distributed across income. but to control for. covw denotes the weighted covariance and Ri is the (representatively positioned) relative fractional rank of the ith individual in the income distribution. .g. α. 112. This annex presents a overview of the need-standardisation procedure. 108. as discussed in Van Doorslaer and Masseria (2004). (2008). We used the indirect standardisation method as described in the appendix of Van Doorslaer and Masseria (2004). More information on Stata programs can be found in O’Donnel et al. 10 Linear OLS regression is preferred since. and ε an error term. Estimates of the indirectly need-standardised utilisation. 109. YX represents the amount of medical care an individual would have received if she/he had been treated as others with the same need characteristics. The concentration index of a variable Y can be computed using a simple “convenient covariance” formula: [4] . doctor visits of individual in a given period). between actual and x-expected utilisation. plus the sample mean : [3] 111. in order to estimate partial correlations with the need variables). METHODOLOGICAL ANNEX 107. The concentration index of the actual health care utilisation measures the degree of inequality and the concentration index of the need-standardised utilisation (which is the horizontal inequity index HI) measures the degree of horizontal inequity.DELSA/HEA/WD/HWP(2012)1 ANNEX 2.

and the estimated standard error of β1 provides the estimated standard error of C.DELSA/HEA/WD/HWP(2012)1 113. where σr2 is the variance of Ri and β1 is equal to C. Another option to obtain robust estimates for C and its standard error consists in running the following “convenient regression” of the transformed Y on relative fractional rank: [5] . 46 .

81 0.41 101127 probability of visit in the past 12 months 0. The surveys in Austria and Ireland do not provide information on doctor visits.86 4.76 3.90 Hungary total visits in past 4 weeks 0.78 0.75 0.024 0.92 0.79 0.107 -0.72 0.124 -0.74 0.82 Denmark total visits in past 3 months 0.03 1.83 0.28 5.013 0.011 -0.83 0.01 0.74 0.26 4.013 * * * * * * * -0.009 0.010 -0.004 -0.69 5.86 0.029 0.005 0.64 0.88 0.96 0.010 0.24 5.99 0.99 0.94 Germany total visits in past 12 months 5.060 * Note: Significant HI and CI indices highlighted with a * (p<0.088 -0.044 * 0.88 0.07 3.30 0.02 4113 probability of visit in the past 12 months 0.25 5.88 0.85 0.11 4.005 -0.79 0.004 0.79 0.80 Spain total visits in past 4 weeks 0.84 0.155 -0.005 * * * * * * * * * * * * * * -0.90 0.82 0.86 0.102 -0.74 0. 47 .022 0.68 10174 probability of visit in the past 12 months 0.027 0.92 0.116 0.42 0.78 0.83 0.78 Finland total visits in past 12 months 2.14 4.83 0.58 19765 probability of visit in the past 12 months 0.81 0.028 -0.01 1.026 0.78 1452 probability of visit in the past 12 months 0.006 -0.19 3916 probability of visit in the past 12 months 0.52 4.89 0.12 4.69 4392 probability of visit in the past 12 months 0.70 0. QUINTILE DISTRIBUTION OF HEALTH CARE UTILISATION Table A3.90 5.42 0.003 0.62 0.19 22611 probability of visit in the past 12 months 0.88 0.27 0.114 -0.056 0.70 0.05 0.94 0.86 0.78 0.80 2.74 0.84 0.018 * * * * * * * * * * * * 3.86 0.35 5.79 3.82 United Kingdom total visits in past 12 months (categorical) 11949 probability of visit in the past 12 months 0.029 0.68 -0.89 0.012 0.72 0.01 1.67 0.76 0.71 1528 probability of visit in the past 12 months 0.1: Quintile distribution of the probability and number of doctor visits after need-standardisation.014 -0.63 0.66 3.80 United States total visits in past 12 months 2.027 -0.70 0.85 0.017 -0.45 0.019 0.90 0.84 0.10 3.011 0.85 0.66 0.17 4.78 0.79 0.74 0.10 3.77 0.92 1.70 0.06 1.88 0.012 0.65 0.74 2.59 0.05).06 3.84 0.014 0.016 0.084 * 0.007 -0.79 0.86 Poland total visits in past 4 weeks 0.91 5.023 0.14 5.66 0.99 4508 probability of visit in the past 12 months 0.75 0.013 0.44 0.052 0.70 0.79 0.85 1.77 1.014 0.11 0.47 0.032 0.97 0.79 0.32 0.22 4.94 1.71 3.62 0.93 0.28 4.84 -0.62 0.84 0.83 0.025 -0.90 23181 probability of visit in the past 12 months 0.048 0.88 0.13 1.022 0.41 Estonia total visits in past 4 weeks 0.82 Slovak Republic total visits in past 4 weeks 0.81 0.65 0.85 0.008 -0.009 -0.61 5833 probability of visit in the past 12 months 0.85 Switzerland total visits in past 12 months 4.012 0.73 0.56 0.77 0.84 0.022 0.109 -0.82 0.80 0.DELSA/HEA/WD/HWP(2012)1 ANNEX 3.76 Total CI HI 0.06 4.43 0.88 12040 probability of visit in the past 3 months 0.001 -0.22 0.019 -0.83 0.83 0.010 -0.82 0.97 0.80 0.021 * 0.91 0.90 Czech Republic total visits in past 4 weeks 0.86 0.10 3.88 Canada total visits in past 12 months 4.65 14491 probability of visit in the past 12 months 0.82 4.82 0.78 0.83 10629 probability of visit in the past 12 months 0.83 4.89 0. inequality index (CI before need-standardisation) and inequity index (HI after need-standardisation) Quintile Quintile Quintile Country Richest Poorest 2 3 4 Sample size Austria total visits in past 4 weeks probability of visit in the past 12 months Belgium total visits in past 4 weeks 0.86 0.61 0.80 0.55 3.020 * 0.138 -0.79 0.88 0.89 Ireland total visits in past 12 months probability of visit in the past 12 months New Zealand total visits in past 12 months 4.010 0.21 17253 probability of visit in the past 12 months 0.039 0.77 1.66 0.002 0.85 Slovenia total visits in past 4 weeks 0.81 0.72 0.73 France total visits in past 12 months 5.76 0.20 5.005 * * * * * * * * * -0.074 -0.79 3.85 0.95 0.60 0.

001 -0.73 2.77 2.028 -0.77 0.29 0.71 0.75 0.58 3.71 0.83 0.68 0.79 0.40 0.75 0.007 0.35 0.82 0.45 0.72 0.58 0.74 2.2: Quintile distribution of the probability and number of GP visits after need-standardisation.50 0.076 0.36 0.80 2.76 0.56 0.80 0.66 0.75 0.016 0.44 0.006 -0.140 -0.74 2.81 0.027 0.77 0.76 0.84 Quintile Quintile Quintile Richest Total 2 3 4 0.018 -0.76 0.017 0.79 0.005 -0.020 * Note: Significant HI and CI indices highlighted with a * (p<0.014 -0.011 0.76 2.54 0.71 0.61 0.78 0.DELSA/HEA/WD/HWP(2012)1 Table A3.85 0.69 0.49 0.018 -0.37 0.38 0.67 0.174 -0.55 0.030 0.81 0.41 0.002 0.78 0.78 2.59 0.91 1.016 0.78 0.83 0.84 0.72 0.095 -0.80 0.77 0.75 0.73 0.002 * * * * * * * * * * * * -0.46 0.068 -0.34 0.66 0.76 0.008 -0.78 0.020 -0.71 2.96 1.76 3.52 0.003 0.029 -0.57 0.60 0.90 0.119 -0.84 2.017 0.80 0.50 0.139 * -0.021 -0.68 0.019 -0.59 1.33 0.79 2.82 2.31 0.68 0.076 0.41 0.57 0.49 0.86 0.40 0.79 0.45 0.106 -0.77 0.53 0.82 0.022 * -0.51 0.009 -0.05).76 0.76 0.024 0.46 0.67 0.027 -0.014 0.78 0.59 0.78 0.71 0.46 0.74 0. The surveys in Germany and the United States do not separately identify GP and specialist visits.85 0.54 3.80 0.82 0.77 2.76 -0.51 0.003 -0.77 2. 48 .78 0.67 1.007 0.70 0.034 -0.77 0.62 0.56 0.72 0.48 0.008 * * * * * * * * * * * * 0.76 0.77 0.34 0.71 0.65 0.79 0.007 -0.58 3.51 0.79 0.77 0.83 0.96 0.79 0.77 0.013 -0.54 3.89 0.46 3.141 -0.39 0.005 -0.005 -0.84 0.59 0.021 -0.65 0.71 0.022 0.57 0.006 * * * * -0.40 0.72 0.65 0.59 0.083 -0.018 -0.78 0.65 3.77 0.038 -0.67 0.52 0. inequality index (CI before need-standardisation) and inequity index (HI after need-standardisation) Country Sample size Austria 14951 Belgium 4392 Canada 101127 Czech Republic 1452 Denmark 12040 Estonia 5833 Finland 3916 France 10174 Germany Hungary 4508 Ireland 8569 New Zealand 10629 Poland 23181 Slovak Republic 4113 Slovenia 1528 Spain 17253 Switzerland 14491 United Kingdom 11949 United States Poorest total visits in past 4 weeks probability of visit in the past 12 months total visits in past 4 weeks probability of visit in the past 12 months total visits in past 12 months probability of visit in the past 12 months total visits in past 4 weeks probability of visit in the past 12 months total visits in past 3 months probability of visit in the past 3 months total visits in past 4 weeks probability of visit in the past 12 months total visits in past 12 months probability of visit in the past 12 months total visits in past 12 months probability of visit in the past 12 months total visits in past 12 months probability of visit in the past 12 months total visits in past 4 weeks probability of visit in the past 12 months total visits in past 12 months probability of visit in the past 12 months total visits in past 12 months probability of visit in the past 12 months total visits in past 4 weeks probability of visit in the past 12 months total visits in past 4 weeks probability of visit in the past 12 months total visits in past 4 weeks probability of visit in the past 12 months total visits in past 4 weeks probability of visit in the past 12 months total visits in past 12 months probability of visit in the past 12 months total visits in past 12 months (categorical) probability of visit in the past 12 months total visits in past 12 months probability of visit in the past 12 months 0.55 0.67 0.61 1.57 0.70 0.168 -0.52 0.157 -0.71 0.51 0.54 0.74 0.73 0.36 0.86 CI -0.73 2.051 -0.45 0.87 0.42 0.79 2.120 -0.67 0.37 0.77 0.29 0.001 * * * * * * * * * * * * HI -0.76 2.87 0.78 0.006 -0.012 -0.78 1.73 0.48 0.84 0.80 0.019 0.081 -0.60 0.023 0.58 0.80 0.56 0.026 -0.75 2.33 0.55 0.54 0.57 0.37 0.

028 0.39 0.49 0.41 1.09 0.29 0.08 0.30 0.56 Czech Republic total visits in past 4 weeks 0.57 0.22 0.34 0.043 0.061 * * * * * * 0.54 0.46 Spain total visits in past 4 weeks 0.106 0.42 0.57 0.054 0.24 4392 probability of visit in the past 12 months 0.16 0.96 1.15 0.55 Slovenia total visits in past 4 weeks 0.37 0.40 France total visits in past 12 months 1.46 0. Ireland.08 0.59 0.005 0.39 0.42 0.44 0.057 0.46 0.77 2.024 -0.71 1.27 0.44 4113 probability of visit in the past 12 months 0.76 9432 probability of visit in the past 12 months 0.42 1.50 0.53 0.50 0.29 0.46 0.052 -0.048 0.23 0.26 0.58 0.32 0.010 0.019 0.33 Poland total visits in past 4 weeks 0.007 0.40 0.52 0.30 17253 probability of visit in the past 12 months 0.196 0.72 1.17 1.055 * Note: Significant HI and CI indices highlighted with a * (p<0.36 0.32 0.40 0.051 0.08 0.17 0.26 0.29 1.55 0.061 -0.56 0.36 1.44 0.21 1.017 0.025 0.022 * * * * * * * * * * 0.43 0.50 0.46 0.49 0.43 1452 probability of visit in the past 12 months 0.53 Switzerland total visits in past 12 months 1.56 0.13 1.030 -0.41 0.51 0.68 0.075 0.51 1.29 0.048 0.13 0.43 0.007 -0.004 0.57 Germany total visits in past 12 months probability of visit in the past 12 months Hungary total visits in past 4 weeks 0.060 * 0.122 0.135 -0.38 0.06 0.19 0.063 -0.05).14 0.52 0.37 0.42 0.01 1.049 0.59 0.54 0.17 0.040 0.61 0.22 0.43 1.015 0.58 Denmark total visits in past 3 months 0.3: Quintile distribution of the probability and number of specialist visits after need-standardisation.029 * -0.52 0.022 -0.24 0.47 0.35 0.19 0.24 5833 probability of visit in the past 12 months 0.60 0.15 1.63 0.030 0.42 0.027 0.95 14491 probability of visit in the past 12 months 0.050 -0.23 0.071 0.19 2.15 12040 probability of visit in the past 3 months 0.10 0.52 United Kingdom total visits in past 12 months (categorical) 11949 probability of visit in the past 12 months 0.30 0.48 0.29 0. inequality index (CI before need-standardisation) and inequity index (HI after need-standardisation) Quintile Quintile Quintile Country Poorest Richest Total 2 3 4 Sample size Austria total visits in past 4 weeks probability of visit in the past 12 months Belgium total visits in past 4 weeks 0.079 * * * * * * * * * * -0.54 0. Germany.50 0.008 * * * * * * * * -0.05 1.33 0.58 0.08 Estonia total visits in past 4 weeks 0.29 0.059 0.46 0.33 23181 probability of visit in the past 12 months 0.35 1.62 0.31 0.46 0.15 3916 probability of visit in the past 12 months 0.59 0. The surveys in Austria.040 0.46 0.60 0.46 0.32 1.119 -0.41 0.14 1.067 -0.011 -0.47 Finland total visits in past 12 months 1.46 0.50 0.072 0.021 0.27 0.60 Ireland total visits in past 12 months probability of visit in the past 12 months New Zealand total visits in past 12 months 1.076 0.21 0.31 0.44 0.023 0.37 0.40 0.45 0.44 0.39 101127 probability of visit in the past 12 months 0.001 -0.44 0.37 4508 probability of visit in the past 12 months 0.54 0.42 0.038 0.77 2.55 0.020 0. 49 .45 0.36 0.44 0.005 0.DELSA/HEA/WD/HWP(2012)1 Table A3.46 0.24 0.045 0.60 0.48 Slovak Republic total visits in past 4 weeks 0.74 1.58 0.060 * -0.022 0.47 0.36 0.53 0.45 0.25 0.26 1528 probability of visit in the past 12 months 0.42 United States total visits in past 12 months probability of visit in the past 12 months CI HI -0. and the United States do not provide information on specialist visits.046 -0.51 Canada total visits in past 12 months 1.08 10629 probability of visit in the past 12 months 0.10 1.076 0.63 0.66 0.60 1.

012 0.40 0.10 0.39 1.07 0.37 0.64 0.23 0.16 0.15 0.062 0.42 0.09 0.42 1.66 0.032 0.106 0.58 1.79 0.38 1.35 0.30 0.58 1.28 0.38 0.71 0.092 * 0.58 0.61 0.151 0.51 0.104 0.15 0.38 0.51 0.55 0.45 1.20 0.56 0.39 0.49 1.57 0.75 CI 0.45 0.69 0.61 0.52 0.49 0.64 Quintile Quintile Quintile Richest Total 2 3 4 0.67 1.34 0.75 1.60 0.61 0.70 0.52 0.65 0.DELSA/HEA/WD/HWP(2012)1 Table A3.65 0.04 0.79 0.65 0.27 0.092 * 0.094 0.35 0.088 * 0.46 0.20 0.4: Dentist visits in the past 12 months.35 0.069 * 0.40 0.48 0.69 0.049 * * * * * * * * * 0.10 0.95 0.78 0.26 0.82 0.52 0.69 0.49 0.34 0.20 0.26 0. and inequality index (CI) Country Sample size Austria 14951 Belgium 4392 Canada 101127 Czech Republic 1452 Denmark 12040 Estonia 5833 Finland 3916 France 9904 Germany Hungary 4508 Ireland 8515 New Zealand 9730 Poland 23181 Slovak Republic 4113 Slovenia 1528 Spain 17253 Switzerland 11366 United Kingdom 11949 United States 22611 Poorest total visits in past 4 weeks probability of visit in the past 12 months total visits in past 4 weeks probability of visit in the past 12 months total visits in past 12 months probability of visit in the past 12 months total visits in past 4 weeks probability of visit in the past 12 months total visits in past 3 months probability of visit in the past 3 months total visits in past 4 weeks probability of visit in the past 12 months total visits in past 12 months probability of visit in the past 12 months total visits in past 24 months probability of visit in the past 24 months total visits in past 12 months probability of visit in the past 12 months total visits in past 4 weeks probability of visit in the past 12 months total visits in past 12 months probability of visit in the past 12 months total visits in past 12 months probability of visit in the past 12 months total visits in past 4 weeks probability of visit in the past 12 months total visits in past 4 weeks probability of visit in the past 12 months total visits in past 4 weeks probability of visit in the past 12 months total visits in past 4 weeks probability of visit in the past 12 months total visits in past 12 months probability of visit in the past 12 months total visits in past 12 months (categorical) probability of visit in the past 12 months total visits in past 12 months probability of visit in the past 12 months 0.094 0.19 0.44 1.061 0.48 1.17 0.79 0.059 * 0.70 0.30 0.28 0.104 0.49 0.19 1.19 0.38 0.35 0.53 0.59 0.28 0.50 0.75 1.47 0.27 0.53 0.57 0.20 0.180 * Note: Significant CI indices highlighted with a * (p<0.65 0.27 0.55 0.058 0.14 0.52 0.99 0.10 0.24 0.15 0.44 0.05).77 0.011 0.23 0.69 0.12 0.42 0.72 0. by income quintile.048 * 0.76 0.71 0.24 0.19 0.45 0.45 0.55 0.74 0.22 0.09 0.37 0.060 0.24 0.33 0.25 0.64 0.26 0.22 0.050 * 0.60 0.40 0.63 0.084 0. The survey in Germany does not provide information on dentist’s visit.60 0.16 0.43 0.56 1.36 0.49 0.48 1.61 0.31 1.26 0.120 0.20 0.55 0.42 0.42 0.47 0.65 0.58 0.67 0.75 0.50 0.55 1.45 1.45 0.25 0.36 0. 50 .23 0.48 0.69 0.60 0.20 0.21 0.17 0.32 0.76 0.67 0.48 0.28 0.57 0.24 0.36 0.40 0.59 0.70 0.13 0.19 0.31 0.196 * 0.21 0.025 * * * * * * * * * 0.17 0.34 1.15 0.121 * 0.66 0.37 0.028 0.131 0.44 0.080 0.063 0.30 0.

63 0. The surveys in Finland and Germany do not provide information on breast cancer screening.36 0.5: Breast cancer screening participation in the past 2 years in women aged 50-69.55 0.53 0.49 0.66 0.59 0.54 0.21 0.64 0. 51 .85 0.64 0.36 0.033 -0.51 0.67 0.001 0.65 0.75 0.168 * * * * 0.77 0.60 0.81 0.73 0.13 0.49 0.77 0.61 0.68 0.57 0.83 0.53 0.80 0.038 0.59 0.013 0.51 0.78 0.61 0.73 0.64 0.15 0.066 * * * * * * Note: Breast cancer screening in the past 12 months in Denmark and Ireland.58 0.78 0.20 0.05).029 0.33 0.72 0.47 0.52 0.DELSA/HEA/WD/HWP(2012)1 Table A3.85 0.31 0.59 0.82 0.87 0.003 0.71 0.69 0.15 0.43 0.87 0.55 0.65 0.78 0.44 0.56 0.70 0.71 0.74 0.83 0.77 0.79 0.50 0. and inequality index (CI) Country (Sample size) Austria (2465) Belgium (691) Canada (9596) Czech Republic (262) Denmark* (2038) Estonia (947) Finland France (1571) Germany Hungary (856) Ireland* (1289) New Zealand (1587) Poland (4584) Slovak Republic (692) Slovenia (253) Spain (2545) Switzerland (2526) United Kingdom (1975) United States (3152) Poorest Quintile 2 Quintile 3 Quintile 4 Richest Total 0.50 0.81 0.15 0.72 0.59 0.87 0.66 0.50 0.011 0.012 0.84 0. by income quintile.52 0.56 0.76 0.52 0.022 -0.89 0.74 0.85 0.039 0.36 0.43 0.60 0.074 0.053 * 0.51 0.72 0.60 0.62 0.54 0.023 0.12 0.77 0.54 0.33 0.81 0.29 0.78 CI 0.024 0.57 0.17 0.125 0.75 0. Significant CI indices highlighted with a * (p<0.79 0.69 0.

60 0.72 0.62 0.039 * 0.58 0. 52 .56 0.56 0.75 0.089 * 0.30 0.25 0.80 0.054 * 0.75 0.50 0.34 0.87 0.064 * 0.023 0.71 0.85 0.63 0.89 0.35 0.79 0.67 0.77 0.59 0.029 * 0.79 0.74 0.76 0.87 0.DELSA/HEA/WD/HWP(2012)1 Table A3.29 0.67 0. Significant CI indices highlighted with a * (p<0. by income quintile.87 0.52 0.71 0.81 0.079 * 0.79 0.46 0.79 0.028 * Note: Cervical cancer screening in the past 12 months in Ireland.66 0.79 0.81 0.37 0.55 0.80 0.28 0.69 0.31 0.45 0.19 0.18 0.68 0.76 0.64 0.062 * 0.79 0.43 0.66 0.82 0.80 0.13 0.69 0.051 * 0.064 * 0.85 CI 0.72 0.61 0.83 0. and inequality index (CI) Country (Sample size) Austria (6339) Belgium (1771) Canada (18556) Czech Republic (565) Denmark (5040) Estonia (2156) Finland France (4092) Germany Hungary (1941) Ireland* (4030) New Zealand (5053) Poland (10040) Slovak Republic (1822) Slovenia (651) Spain (5875) Switzerland (6183) United Kingdom (5168) United States (9363) Poorest Quintile 2 Quintile 3 Quintile 4 Richest Total 0.73 0.50 0.71 0.62 0.69 0.090 * 0.82 0.78 0.70 0.80 0. The surveys in Finland and Germany do not provide information on cervical cancer screening.53 0.071 * 0.56 0.81 0.29 0.78 0.20 0.73 0.79 0.14 0.066 * 0.80 0.81 0.68 0.67 0.6: Cervical cancer screening participation in the past 3 years in women aged 20-69.69 0.84 0.82 0.74 0.05).78 0.038 * 0.74 0.049 * 0.13 0.80 0.75 0.35 0.034 * 0.59 0.79 0.16 0.61 0.39 0.64 0.52 0.61 0.

or what people are in principle entitled to.DELSA/HEA/WD/HWP(2012)1 ANNEX 4. a score was attributed to each function of care. Some indicators were constructed in order to summarize information such as the three following indicators: (a) depth of coverage. First. Depth of coverage 116. 118. 119. and (c) regulation of price billed by provider 115. 2010). These three aggregated indicators are described here. Then. DETAILS ON HEALTH SYSTEMS CHARACTERISTICS 114. To be able to perform the analysis on the largest number of countries. This sub-indicator reflects the “theoretical” level of coverage of the population. (b) degree of private provision of care. missing data were estimated by the Secretariat or imputed by the OECD mean. Scope and depth of basic coverage defined by regulation 117. we added 0. a global score was computed as the weighted average of function-related scores. Scoring rules for scope and depth of basic coverage defined by regulation: 53 . It is based on country replies to questions 13 and 16 of the survey questionnaire. according to the share of typical costs covered by basic primary coverage (see table below).5 if exemption mechanisms exist for socially vulnerable people (people whose income is under a designated threshold or beneficiaries of social benefits). To account for the exemption from copayments (question 16 in the HSC questionnaire). Most of the information on health system features was collected through the OECD Health Data 2011 database or through the 2008 Health Systems Characteristics survey (see the questionnaire in Paris et al. • • • The indicator “Depth of coverage” is based on 3 sub-indicators: “scope and depth of basic coverage defined by regulation” “actual level of coverage by health insurance’ “out-of-pocket payments for essential care”.5 to the total score if exemption mechanisms exist for people with high medical needs (people with certain medical conditions or disabilities or people who have reached an upper limit for out-of-pocket payments) and 0.

second. the sub-score “out-of-pocket payments for essential care” has a negative sign. 123. For instance. no copayment Covered. 11 These weights were estimated from System of Health Account 2009 series. the share of individual spending covered by any type of insurance (basic or secondary. financial access to the first point of entry in medical care system guarantees that people are more likely consult in case of problems. weighted by the OECD average share of each function in total health spending (0. This indicator was chosen as a proxy to measure “real financial accessibility to health care”. which varies from 48. it is based on two principles. In contrast to the other sub-indicators. 124. “out-of-pocket payments for essential care”. social insurance and private insurance) This share. notably because many people remain uninsured. the Slovak Republic has a high score for the scope and depth of coverage defined by regulation. Data on the share of out-of-pocket payments for inpatient curative care and for basic medical and diagnostic services were drawn from the System of Health Account 2009 or estimated by the Secretariat when not available. 121. copayments 51-75% Covered. for some countries. the share of population covered by any type of insurance. For different reasons.5 3 1.1: Scoring of depth of coverage Level of coverage Covered.72 for inpatient curative care and 0. but in reality. In the aggregated score of “Depth of coverage”. In this case. First. and is confirmed by a medical prescription. a high score reflects a lower level of coverage. The actual level of coverage results from several parameters: first. is based on the share of outof-pocket payments for inpatient curative care and for basic medical and diagnostic services. acute inpatient care is generally costly. The third sub-indicator. which increases their opportunity to receive the diagnosis and treatments they need.9% to 92. Instead. in Mexico.DELSA/HEA/WD/HWP(2012)1 Table A4. and even though it corresponds to a real need. scores are very different from those obtained for “scope and depth of coverage”. Country scores are based on the average of both series. but one of the lowest score for the actual level of coverage. This sub-indicator is based on the share of total health spending financed by any type of health insurance (public sector. was rescaled on a 0 to 6 range to obtain a normalised score. copayments 1-50% Not covered Score 6 4. out-of-pocket payments are high. copayment 75-99% Covered. but the actual level of coverage of the whole population is lower.6% in OECD countries. and third. the scope and depth of coverage is high for people covered by health insurance. Second. Selecting only two functions of care rather than total spending does not mean the other functions of care are less important.28 for basic medical and diagnostic services11). high copayments could be a financial hurdle for many households. 54 . public or private) for covered goods and services.5 0 Actual level of coverage by health insurance 120. Out-of-pocket payments for essential care 122. the size of the whole population insured and entitlements are wide. the share of self-consumption of not covered services in health spending. Therefore.

representing more than 20% of physicians’ contacts.0019x + 0. representing more than 20% of services) for both primary care services and specialist services . when the score indicates that the provision of physicians’ services is exclusively public (or private).0553 R² = 0. this does not mean that no other mode of provision exists.DELSA/HEA/WD/HWP(2012)1 125. There is little variation in the depth of coverage across countries (score from 0 to 6) and results are inconclusive. but only that it represents a small proportion of physicians’ services.06 0. 55 . 127. as well as the second mode of delivery when relevant (i.00 4 5 FRA DNK ESP y = -0.08 0. The degree of private provision of physicians’ services aggregates information on primary care and out-patient specialists’ services. The degree of private provision of physicians’ services was measured using the predominant mode of health care delivery.1: Relationship between inequity and depth of coverage 0. Figure A4. Countries were asked to provide information on the predominant mode of delivery and to indicate whether another significant mode exists. Separate sub-scores were first calculated for primary care services and specialist services (see table below) and further aggregated with a simple average. Figure A4. i.e.e.10 Inequity in specialist visits (HI) 0.1 shows the relationship between inequity in specialist services use and the depth of coverage.04 0.0005 BEL CHE FIN SVK POL EST HUN SVN CZE GBR 6 CAN NZL Depth of coverage (score 0-6) Degree of private provision in physicians’ services 126. Consequently.02 0.

Japan.2: Scoring the degree of private provision of physicians’ services Provision of primary care services Public health care centres. Regulation of prices billed by providers 130. the relationship. Norway and Switzerland. public hospitals private clinics. private group/solo practice.2 shows the relationship between inequity in specialist services use and regulation of prices billed by specialists suggesting that higher regulation is associated with lower inequity in some countries (e.DELSA/HEA/WD/HWP(2012)1 Table A4. the Netherlands. public hospitals private clinics. only Q28 (predominant mode) public centres. if any. 131. Portugal. 129. However. exam. private group/solo practice. private group/solo practice private clinics. (2010) on the determination of prices/fees for specific services (consultation. private group/solo practice Q28b (second mode) 6 private clinics. Denmark. The indicator is based on the information collected in Table 25 in Paris et al. private hospitals private clinics. private group/solo practice.g. 56 . private group/solo practice public centres Q27b (second mode) Score 0 3 Score Mix of public and private provision 3 Private provision only private clinics. The data on “prices billed by providers” approximates a measure of “free care at the point of care delivery”. is not strong – New Zealand has the lowest regulation of physician pricing among the countries considered. At the other end of the spectrum. Spain. Luxembourg. private hospitals 6 128. private group/solo practice Private provision only private clinics. private hospitals public centres. United Kingdom. private group/solo practice Provision of out-patient specialist services Public health care centres. Physicians’ services are predominantly provided by the public sector in five countries: Italy. Sweden and Turkey. procedure). France. public hospitals 0 public centres. It summarises how prices are set. the supply is exclusively private in nine OECD countries: Belgium. Figure A4. Germany. public hospitals public centres. private hospitals private clinics. private group/solo practice. and Hungary has a medium degree of regulation but has one of the lowest levels of inequity. The score for the private provision of physicians’ services is calculated as the simple average of the two previous sub-scores: private provision of primary care services and private provision of out-patient specialists’ services. Korea. private hospitals public centres. and Czech Republic). public hospitals private clinics. but has average inequity. private group/solo practice. for both prices paid by third party payers and prices billed by providers. only Mix of public and private provision Q27 (predominant mode) public centres public centres private clinics.

DELSA/HEA/WD/HWP(2012)1 Figure A4.Prices can 2 sometimes exceed prices/fees paid by third-party payers and statutory copayments 3 Low regulation-Prices can always exceed prices/fees paid by third-party payers and statutory copayments Degree of regulation of prices billed by specialists 57 .10 HI in specialist visits 0.Prices must be equal to prices/fees paid by third-party payers + statutory copayments POL CAN CHE SVK CZE GBR FRA DNK BEL FIN HUN NZL OECD mean Medium .08 ESP 0.00 0 1 High regulation.04 0.06 0.2: Relationship between inequity and regulation of prices billed by physicians 0.02 0.

57 No.DELSA/HEA/WD/HWP(2012)1 OECD HEALTH WORKING PAPERS A full list of the papers in this series can be found on the OECD website: www. Lihan Wei. Marion Devaux and Lihan Wei THE CHALLENGE OF FINANCING HEALTH CARE IN THE CURRENT CRISIS (2010) Peter Scherer. 51 No. 50 No. 44 58 . Gay. Marion Devaux. Jeremy Lauer and Dan Chisholm HEALTH CARE QUALITY INDICATORS PROJECT: PATIENT SAFETY INDICATORS REPORT 2009 (2009) Saskia Drösler. 47 No. and ANNEX Saskia Drösler EDUCATION AND OBESITY IN FOUR OECD COUNTRIES (2009) Franco Sassi. Jody Church. Valerie Paris. Marion Devaux. 45 No. Michele Cecchini and Elena Rusticelli THE LONG-TERM CARE WORKFORCE: OVERVIEW AND STRATEGIES TO ADAPT SUPPLY TO A GROWING DEMAND (2009) Rie Fujisawa and Francesca Colombo No. 46 No.org/els/health/workingpapers No. TACKLING OBESITY: THE HEALTH AND ECONOMIC IMPACT OF PREVENTION STRATEGIES (2009) Franco Sassi. Marion Devaux IMPROVING LIFESTYLES. Michael de Looper NURSES IN ADVANCED ROLES: A DESCRIPTION AND EVALUATION OF EXPERIENCES IN 12 DEVELOPED COUNTRIES (2010) Marie-Laure Delamaire and Gaetan Lafortune COMPARING PRICE LEVELS OF HOSPITAL SERVICE ACROSS COUNTRIES: RESULTS OF A PILOT STUDY (2010) Luca Lorenzoni GUIDELINES FOR IMPROVING THE COMPARABILITY AND AVAILABILITY OF PRIVATE HEALTH EXPENDITURES UNDER THE SYSTEM OF HEALTH ACCOUNTS FRAMEWORK (2010) Ravi P. 53 No.oecd. Marion Devaux. 52 No. 56 THE IMPACT OF PAY INCREASES ON NURSES’ LABOUR MARKET: A REVIEW OF EVIDENCE FROM FOUR OECD COUNTRIES (2011) James Buchan and Steven Black DESCRIPTION OF ALTERNATIVE APPROACHES TO MEASURE AND PLACE A VALUE ON HOSPITAL PRODUCTS IN SEVEN OECD COUNTRIES (2011) Luca Lorenzoni and Mark Pearson MORTALITY AMENABLE TO HEALTH CARE IN 31 OECD COUNTRIES:ESTIMATES AND METHODOLOGICAL ISSUES (2011) Juan G. 49 No. Michele Cecchini. 48 No. Michele Cecchini and Francesca Borgonovi THE OBESITY EPIDEMIC: ANALYSIS OF PAST AND PROJECTED FUTURE TRENDS IN SELECTED OECD COUNTRIES (2009) Franco Sassi. 55 No. 54 No. Patrick Romano. Rannan-Eliya and Luca Lorenzoni EFFECTIVE WAYS TO REALISE POLICY REFORMS IN HEALTH SYSTEMS (2010) Jeremy Hurst HEALTH SYSTEMS INSTITUTIONAL CHARACTERISTICS A SURVEY OF 29 OECD COUNTRIES (2010) Valerie Paris.

37 No. 40 No. Maria Luisa Gil Lapetra. Lihan Wei. 36 No. 29 HEALTH CARE QUALITY INDICATORS PROJECT 2006 DATA COLLECTION UPDATE REPORT (2007) Sandra Garcia-Armesto. 31 No. 28 59 . 43 No. 35 No. 39 No. 42 No. 30 No. Cooper MIGRATION POLICIES OF HEALTH PROFESSIONALS IN FRANCE (2008) Roland Cash and Philippe Ulmann NURSE WORKFORCE CHALLENGES IN THE UNITED STATES: IMPLICATIONS FOR POLICY (2008) Linda H. 33 No. Hofmarcher. Edward Kelley and the Members of the HCQI Expert Group PHARMACEUTICAL PRICING AND REIMBURSEMENT POLICIES IN SWEDEN (2007) Pierre Moïse and Elizabeth Docteur No. Howard Oxley. 34 No.DELSA/HEA/WD/HWP(2012)1 No. and Elena Rusticelli CARE No. Susanna Baldwin and Miranda Munro THE US PHYSICIAN WORKFORCE: WHERE DO WE STAND? (2008) Richard A. Jody Church and Christine Le Thi PHARMACEUTICAL PRICING & REIMBURSEMENT POLICIES IN GERMANY (2008) Valérie Paris and Elizabeth Docteur MIGRATION OF HEALTH WORKERS: THE UK PERSPECTIVE TO 2006 (2008) James Buchan. Aiken and Robyn Cheung MISMATCHES IN THE FORMAL SECTOR. EXPANSION OF THE INFORMAL SECTOR: IMMIGRATION OF HEALTH PROFESSIONALS TO ITALY (2008) Jonathan Chaloff HEALTH WORKFORCE AND INTERNATIONAL MIGRATION: CAN NEW ZEALAND COMPETE? (2008) Pascal Zurn and Jean-Christophe Dumont THE PREVENTION OF LIFESTYLE-RELATED CHRONIC DISEASES: AN ECONOMIC FRAMEWORK (2008) Franco Sassi and Jeremy Hurst PHARMACEUTICAL PRICING AND REIMBURSEMENT POLICIES IN SLOVAKIA (2008) Zoltán Kaló. 38 No. Pascal Zurn. Elizabeth Docteur and Pierre Moïse IMPROVED HEALTH SYSTEM PERFORMANCE THROUGH BETTER COORDINATION (2007) Maria M. 41 MEASURING DISPARITIES IN HEALTH STATUS AND IN ACCESS AND USE OF HEALTH CARE IN OECD COUNTRIES (2009) Michael de Looper and Gaetan Lafortune POLICIES FOR HEALTHY AGEING: AN OVERVIEW (2009) Howard Oxley THE REMUNERATION OF GENERAL PRACTITIONERS AND SPECIALISTS IN 14 OECD COUNTRIES: WHAT ARE THE FACTORS EXPLAINING VARIATIONS ACROSS COUNTRIES? (2008) Rie Fujisawa and Gaetan Lafortune INTERNATIONAL MOBILITY OF HEALTH PROFESSIONALS AND HEALTH WORKFORCE MANAGEMENT IN CANADA: MYTHS AND REALITIES (2008) Jean-Christophe Dumont. 32 No.

org/health/healthdata) OECD REVIEWS OF HEALTH SYSTEMS . or write for a free written catalogue to the following address: OECD Publications Service 2.TURKEY (2009) For a full list.oecd.org.oecd. OECD HEALTH POLICY STUDIES OECD REVIEWS OF HEALTH SYSTEMS .DELSA/HEA/WD/HWP(2012)1 RECENT RELATED OECD PUBLICATIONS OECD HEALTH DATA 2012 (Database available from http://www. rue André-Pascal. 75775 PARIS CEDEX 16 or to the OECD Distributor in your country 60 .oecd.org/health/healthataglance for more information HEALTH REFORM: MEETING THE CHALLENGE OF AGEING AND MULTIPLE MORBIDITIES (2011) A SYSTEM OF HEALTH ACCOUNTS 2011 EDITION HELP WANTED? PROVIDING AND PAYING FOR LONG-TERM CARE (2011) IMPROVING VALUE IN HEALTH CARE: MEASURING QUALITY (2010) IMPROVING HEALTH SECTOR EFFICIENCY COMMUNICATION TECHNOLOGIES (2010) – THE ROLE OF INFORMATION AND MAKING REFORM HAPPEN – LESSONS FROM OECD COUNTRIES (2010) HEALTH AT A GLANCE: ASIA/PACIFIC (2010) HEALTH AT A GLANCE: EUROPE (2010) VALUE FOR MONEY IN HEALTH SPENDING (2010) OBESITY AND THE ECONOMICS OF PREVENTION: FIT NOT FAT (2010) ACHIEVING BETTER VALUE FOR MONEY IN HEALTH CARE (2009).RUSSIAN FEDERATION (2012) OECD REVIEWS OF HEALTH CARE QUALITY – KOREA (2012) SICK ON THE JOB? MYTHS AND REALITIES ABOUT MENTAL HEALTH AND WORK (2012) OECD REVIEWS OF HEALTH SYSTEMS . consult the OECD On-Line Bookstore at www.SWITZERLAND (2011) HEALTH AT A GLANCE 2011: OECD INDICATORS (2011) See http://www.

Sign up to vote on this title
UsefulNot useful