8931.pdf
Media
- extracted text
-
Bucking the Trend
How Sri Lanka has achieved good health at low cost:
challenges and policy lessons for the 21st century
7 Save the Children
Bucking the Trend
How Sri Lanka has achieved good health
at low cost - challenges and policy
lessons for the 21st century
Kirsty McNay, Regina Keith and Angela Penrose
Save the Children
Published by
Save the Children
1 St.John’s Lane
London
EC1M4AR
UK
Tel: +44 (0)20 7012 6400
Fax: +44 (0)20 7012 6963
© Save the Children 2004
All rights reserved. This publication is copyright, but may be reproduced by
any method without fee or prior permission for teaching purposes, though not
for resale. For copying in other circumstances, prior written permission
must be obtained from the publisher and a fee mav be pavable.
Registered Charity No. 213890
•A
r
N 011 V W M O J NI
ONV
z.avyaii
>
V' (S -1 oo
U8831
Contents
Acknowledgements............................................................................................. iv
Abbreviations and acronyms.............................................................................. v
Executive summary.............................................................................................. 1
1. Introduction..................................................................................................... 4
2. Sri Lanka’s health achievements................................................................7
3. Development of the public health system in Sri Lanka................... 10
3.1. Other interventionist policies and factors facilitating Sri Lanka’s health
achievements........................................................................................................................... 12
3.2. Evidence on the causal links between Sri Lanka’s health achievements and the
public healthcare system........................................................................................................ 14
4. What makes Sri Lanka’s public healthcare system successful?.... 15
4.1. The public healthcare system as an equitable system...............................................15
4.2. The public healthcare system as an efficient system............................................... 17
4.3. Tire public healthcare system as an insurance mechanism against catastrophic
illness.........................................................................................................................................18
4.4. Complementarity between die public and private healdicare systems................. 19
5. Current and future challenges facing Sri Lanka’s public healthcare
system.................................................................................................................... 21
6. Conclusion..................................................................................................... 25
6.1.
6.2.
Implications for policy.................................................................................................. 25
Recommendations ......................................................................................................... 26
Bibliography........................................................................................................ 28
Notes...................................................................................................................... 32
List of figures
Figure 1. Infant mortality rate and GDP per capita, various countries, 2000
5
Figure 2. Sri Lanka’s demographic transition, 1900-2000
7
Figure 3. Infant mortality rate and maternal mortality ratio, Sri Lanka 1900-2000
8
Figure 4. Population pyramids, Sri Lanka 1960 and 2000
21
List of tables
Table 1. Trend in life expectancy, Sri Lanka, 1946-2001
9
Table 2. Trend in total fertility rate, Sri Lanka, 1963-2000
9
Table 3. Growth of public health facilities and personnel, Sri Lanka, 1926-2001
11
Table 4. Trends in public and private provision of health services, Sri Lanka, 1990s
12
Table 5. Under-five child mortality rates by educational level of mother, Sri Lanka, 2000
13
Table 6. Distribution of public health spending by the poorest and richest income
quintiles, various countries, 1980s and 1990s
16
Table 7. Public health expenditure as a percentage of GDP, various countries, late 1990s
17
Table 8. Percentage of children aged less than five years who are underweight, various
countries, 1995-2000
24
iii
Acknowledgements
The authors wish to thank RP Rannan-Eliya from the Institute of Policy Studies for
sharing his work and knowledge of the Sri Lankan health policy arena and Stephen Sabey
from die Department for International Development (DFID).
Thanks also to Greg Duly (Save the Children UK in Sri Lanka) for sharing his research
on new trends in education in Sri Lanka, and to Arany Nerminathan, David Walker and
Arun Muttreja. Special thanks to Michael Bailey, Bill Bell and Anna Taylor for their
support in the editing of the paper.
iv
Abbreviations and acronyms
DFID
Department for International Development
GDP
gross domestic produce
GNP
gross national product
LTTE
Liberation Tigers of Tamil Eelam
MDGs
Millennium Development Goals
NGO
non-governmental organisation
UNICEF
United Nations Children’s Fund
WHO
World Health Organisation
V
Executive summary
There is growing concern that the Millennium Development Goals (MDGs) for child
and maternal health will not be met by 2015, recharging the debate on what is required to
achieve them. While a ‘one size fits all’ approach is widely rejected, the prescriptions of
international organisations such as the World Bank and the World Health Organisation
(WHO) continue to shape the model on offer as they have over the past two decades.
The model may be characterised as focusing public expenditure on out-patient care, cost
recovery, and growth of the private sector.
In general, poor countries have far higher levels of infant mortality- than rich ones, but Sri
Lanka’s experience, as one of the few countries to have achieved remarkable progress in
health and social development despite its low-income status, demonstrates that factors
odier than income can significandy affect health outcomes, including child mortality. The
state has financed and provided a high level of healthcare for its citizens at a relatively^
low cost (around two per cent of GDP), which has contributed to low infant and
maternal mortality and high life expectancy.
Considerable attention has been paid to the Sri Lankan example and this paper
investigates the factors diat have contributed to its achievements in order to see what can
be learnt for die benefit of children elsewhere and to consider how these achievements
can be sustained when threatened by new challenges. In certain key areas it can be seen
that die policies adopted and implemented by die Sri Lankan government have been and
in some cases still are at variance widi diose proposed and widely- followed in the general
pursuit of the MDGs.
The factors that have contributed to the Sri Lankan success in improving the health
status of its citizens may- be categorised into three groups. First, diere are cultural, social
and historical reasons. These include high levels of women's autonomy- and relative
gender equality-, a democratic system based on universal franchise and a significant level
of consensus on national priorities relating to the government’s provision of social
services.
Second, die government managed a series of complementary- policies outside the healdi
sector diat have had direct impact on healdi outcomes. Whilst it may be difficult to
quantify die exact impact of relations betw-een interventions, public programmes in other
sectors since before independence have undoubtedly- created positive synergies with the
healdi sector. Free education since 1947 and high levels of female literacy have
engendered attitudinal changes and created a knowledge base that has weathered periods
of economic decline. The nutritional status of poor families has been improved through
subsidised distribution of rice, and w-ater and sanitation systems have been developed in
parallel with die healdi system.
Third, diere have been important healdi policy- decisions, based on principles of equitv
and efficiency, from which some lessons may- be drawn, while bearing in mind die wider
context in which they are located:
a)
an emphasis on public financing of in-patient care, rather than out-patient care
b)
creation of a motivated and trained body- of healdi personnel
c)
rejection of cost recovery as a general financing policy.
1
In-patient care
The focus on freely available in-patient care with equitable access has had a number of
effects:
a)
It has created a dynamic for a strong referral system widiin a wide network of
health facilities available across the country and close to where people live. This
in turn has created a strong demand for health services and universal
understanding of how to use diem, as well as of die main health issues. Health
provision is an important political issue, and politicians need to ensure diat it is
present in dieir constituencies.
b)
Most women elect to give birth in public hospitals as public resources are
directed at in-patient deliver}' care.
c)
A complementary private sector has evolved. Recognising diat in-patient
provision is wide spread and of adequate quality, a growing private sector has
focused on out-patient care and die provision of higher cost care to diose who
can pay. The division of services between the sectors contributes to a pro-poor1
redistributive effect: public facilities are available to evervone, but wealthier
patients self-select out of die public system for out-patient care, aldiough they
continue to use and support it for in-patient services.
Trained health personnel
In addition to Sri Lanka’s strong public service edios, the healdi sector emphasis on
financing in-patient care has encouraged healdi workers to remain in the pubke healdi
system, dirough strong non-financial incentives such as professional development which
in turn increases dieir status. At die same time, they are permitted to work privately,
aldiough, until relatively recendy, the level of public sector remuneration has been
adequate for the purpose of retaining staff, who can add to their income with private
work. The focus on in-patient care has provided the infrastructure necessary to maintain
a committed health sendee staff.
Rejection of cost recovery
Sri Lanka has developed its public healthcare system without mobilising resources
beyond the level of most odier developing countries, and at significandy less cost than
the World Bank’s ‘minimum cost-effective package’ of basic preventive and curative
health services. Total annual public spending on health has generally averaged less than
two per cent of GDP. Maintaining extensive access to and use of services has been
achieved through lowering unit costs by high staff productivity and the intensive use of
infrastructure. Such improvements in productivity have been sustained over time
apparendy through careful use of non-financial incentives for staff and a strong
institutional service culture among health professionals.
Cost recover}’ can have perverse effects in that it may reduce die pressure on die public
sector to be efficient by allowing it to collect additional income outside the tax system.
By denying itself additional non-tax income the government is increasingly pressured to
ensure that the system is efficient. As care is free, die poor use the system extensively and
the population expects it to deliver dieir requirements.
2
Implications of these achievements for the future
Sri Lanka’s achievements continue to be sustained but the country now faces some
complex health challenges. The Sri Lankan Government’s commitment to investing in
public health measures was pioneering, both in terms of its early start and the low level
of national income at which it began. The decline in mortality began in the 1920s, and a
remarkable ‘breakthrough’ period (1946-53) occurred during which infant mortality' fell
by 50 per cent, maternal mortality by' 68 per cent and life expectancy increased by' over a
decade. This success in achieving both an epidemiological and demographic transition
now means that it has a rapidly ageing population and by 2010 Sri Lanka will be the third
oldest country' in Asia, after Japan and Singapore. Although data from government
hospitals shows that morbidity' remains dominated by infectious conditions and injuries,
it also shows a clear increasing trend in chronic disease cases associated with older
populations, such as heart disease, cerebrovascular conditions, diabetes and cancers. At
the same time, continued fertility' decline will eventually' mean that the ratio of middleaged women (the primary providers of informal home-based care of the elderly in Sri
Lanka) to tire elderly will decrease.
These processes raise issues about both the pattern of healthcare expenditure and the
level of spending. Sri Lanka’s system developed with an emphasis on the prevention and
treatment of communicable diseases, but the epidemiological transition means that
patterns of healthcare expenditure will have to become more diverse and complex, with
pressure for personnel and technology' to become increasingly specialised. Additional
funding is likely to be needed to meet both these new requirements and the increased
demand for formal, government-provided long-term care that accompanies the rise in
chronic diseases.
Although there is awareness of a need for reform to meet these challenges, the political
consensus, which shaped government priorities in die past, has yet to coalesce. Resource
constraints are mounting, and within the health sector there is pressure to increase
spending on salaries to pre-empt an exodus of staff to die private sector and to
modernise public provision in line with private provision. There is also a lack of
managerial capacity' at bodi central and provincial levels to plan and manage the
necessary' reform.
1. Introduction
Save die Children UK has over 80 years of experience in maternal and child healdicare.
In this time our role as an international non-governmental organisation (NGO) has
changed dramatically from only providing relief services following conflicts to supporting
nationally owned and led solutions. Throughout this time our focus on supporting the
realisation of all children’s rights has not changed? In die last decade Save die Children
UK has continually called for increased recurrent resources for health systems in
developing countries. These resources remain scarce and child health in die poorest
countries has suffered. In 2001 Save the Children UK released a report on the negative
impact of economic policies on healdi expenditure and healdi indicators (Save die
Children UK and Medact, 2001). In die same year WHO also released a report from their
Commission on Macroeconomics and Health, linking healdi to poverty, and including
estimates on how much health system strengthening would cost (WHO, 2001). Three
years later and despite lots of new disease-specific globally-led initiatives for HIV, TB
and malaria, dicrc is still no solid action plan for ensuring that fewer children in the
poorest countries die from preventable diseases by 2015.
In 2003 an NGO report was launched by die UK Secretary of State for International
Development, entided ‘SO Million Lives: Meeting the Millennium Development Goals in child and
maternal survival’. In it, 20 UK-based NGOs called for urgent policy change, and a return
to a social model of healdi, with more recurrent resources for healdi and increased
accountability’ from all stakeholders in order to achieve die health-related MDGs. (Grow
Up Free from Poverty, 2003). That report highlighted how child-focused and pro-poor
healdi and economic policy changes were urgendy needed to move die international
health debate away from a focus on cost effectiveness and other ncoliberal principles,
towards a rights-based pro-poor model to achieve die healdi-related MDGs by 2015.
Save the Children UK’s experiences and research point to a need for rapidly increased
resources for healdi, accompanied by dramatic policy change, to enable 76 million
children’s lives to be saved by 2015.
Sri Lanka has long been put forward as an example of a low-income country that has
achieved remarkable progress in health and social development. This paper explores why
Sri Lanka, despite widespread poverty, weak economic growth and high levels of
malnutrition, still manages to maintain low child and maternal mortality rates at
expenditure levels of less than two per cent GDP annually.
The World Bank has presented Sri Lanka as a having a pro-poor health policy when
compared to other Asian nations like Bangladesh, which is seen as pro-rich despite
allocating 67 per cent of its health sector budget to ‘initiatives which target the poor’. In a
six-nation Asian study, presented at die World Bank’s 2004 Washington conference —
‘Reaching the Poor’ — the countries classified as pro-poor had used universal access to
health as a policy priority. Those ranked as pro-rich had employed user fees combined
with targeting measures to reach the poor (World Bank, 2004b). Following the
publication of the World Bank’s World Development Report 2004 (^Making Services II ot-k
for Poor People', World Bank, 2004a), and its recognition that progress towards the healthrelated MDGs will be inadequate unless critical services reach die poorest households
with the greatest burden of ill-health, an understanding of how Sri Lanka has achieved
pro-poor health success is even more paramount.
4
Figure 1. Infant mortality rate and GDP per capita, various countries, 2000
Note: The two data points to the right ofSri Fanka represent Jamaica and Ukraine.
Source: United Nations 2002a.
Figure 1 shows GDP per capita and the infant mortality rate for a cross-section of
countries for which both indicators are available. The overall relationship clearly indicates
that wealth is an important factor in achieving better health status; in general poor
countries have far higher levels of infant mortality- than do rich ones. However, there are
also notable deviations, both positive and negative, from the overall relationship. Sri
Lanka’s position indicates a positive deviation from the overall relationship; the deviation
indicating that given its income level, mortality- achievement is higher than expected. Sri
Lanka’s experience therefore indicates drat factors other than income can also be
important in realising progress. It can therefore be placed at tire centre of the debate
about whether income growth is necessarily the driving force behind health and social
development, or whether other routes to high achievement are also possible.
Sri Lanka’s route is generally characterised as early and sustained government
intervention. Dreze and Sen (1989) describe Sri Lanka’s path as one of‘support-led
security’ in which public provisioning of health and social sector services has been used
as the crucial means of promoting progress without waiting for economic growth to do
so. Other examples of countries following tire ‘support-led security’ path include Chile,
China, Costa Rica, Cuba, Jamaica, and the Indian state of Kerala.’ However, according to
Dreze and Sen, the Sri Lankan experience is particularly interesting because of both its
pioneering timing and tire fact drat it was undertaken at a lower income level than the
other examples of‘support-led security’. Dreze and Sen assert that ‘Sri Lanka’s strategic
experience as a pioneer in overcoming the major penalties of low income remains one of
great significance for understanding tire prospects for support-led security in poor
countries.’
With tills in mind, the paper outlines Sri Lanka’s health achievements and places its
current performance within its historical context. Given that this paper has been written
to promote the achievement of the United Nation’s MDGs 4 and 5 (relating to child and
maternal mortality respectively), we emphasise Sri Lanka’s progress in lowering these two
aspects of mortality. Throughout the paper we comment on some of the implications for
5
children of Sri Lanka’s public intervention in the health sector.1 We then discuss tire
development of the public health system. We also recognise that government
intervention in Sri Lanka has been multisectoral in nature, and we note some of the
positive synergistic effects of such an approach. We mention some of die cultural and
political factors present in Sri Lanka diat have facilitated its achievements in healdi status.
We dien briefly review die literature that causally links public intervention to those
achievements. We go on to investigate several specific features of die public healdi
system in more detail, explaining how they have contributed to its success. Finally we
comment on some of die healdi issues and challenges Sri Lanka now faces, and will face
in die near future, and what die implications are for the healdi system. We pay particular
attention to die implications for die healthcare system of epidemiological transition, and
the growing presence of die private healdi sector, and we argue that appropriate health
system reform could enhance die already existing complementarity between die public
and private systems.
At the outset, it should be noted diat diis paper does not discuss die healdi implications
of Sri Lanka’s internal conflict between die Liberation Tigers of Tamil Eelam (LTl'E)
and the Sri Lankan Government, although we acknowledge that significant health effects
exist. Regional data shows diat die conflict-affected Nordiern and Eastern districts lag
behind the rest of Sri Lanka in terms of various demographic and health indicators. For
example, the average maternal mortality7 ratio per 1,000 live birdis for 1995-96 in die five
Northern districts of Jaffna, Kilinochchi, Mannar, Mullaitivu, and Vavuniya is 0.7
compared with a ratio of 0.2 for Sri Lanka as a whole (Department of Census and
Statistics, 2002b).’ In terms of healdi services, the Northern and Eastern districts suffer
from poor health infrastructure and a lack of healdi professionals and medical supplies
(UNICEF, 2003). Healthcare in the affected regions is characterised by lower access,
availability7 and quality7 compared widi the rest of die country (Reilly et al, 2002). Water
and sanitation, and other social sectors such as education, are also adversely7 affected.
Children are especially7 vulnerable in these conditions. For example, chronic malnutrition
is particularly7 high among children displaced by7 conflict (Save the Children, 2000). The
school drop-out rate in the Nordi and East is 16 per cent compared with die national
average of 3.9 per cent, and underage combatant recruitment by the LITE means that
the release and reintegration of child combatants will be a priority in a post-conflict
environment (UNICEF, 2003).
Such negative disparities in health and social indicators between the Nordiem and
Eastern provinces and die rest of die country7 almost wholly7 date from after die start of
continuous insurgency in 1983. Prior to 1983, most of the Nordicrn and Eastern districts
reported above average indicators for healdi, education and general living standards. This
implies that if it were not for the conflict, the good health indicators in die rest of die
country7 should have been replicated throughout the island. In diis paper we present data
at the national level only. We note if districts affected by the conflict are omitted from
the national figures.
6
2. Sri Lanka’s health achievements
Figure 2 illustrates Sri Lanka’s demographic transition since 1900.6 Figure 3 focuses on
trends in infant and maternal mortality.7 Table 1 shows die increase in life expectancy at
birth diat has accompanied the trends, and Table 2 shows the fall in the total fertility
rate.8 As can be seen from die figures, mortality decline began as early as the 1920s, vith
rapid falls occurring during the 1940s and 1950s, and the secular downward trend being
sustained diereafter. Caldwell (1986) refers to die years 1946-53 as Sri Lanka’s
‘breakthrough’ period, a phase of only seven years during which infant mortality fell by
50 per cent, maternal mortality by 68 per cent, and over a decade was added to life
expectancy'.9 During the 1960s to the mid-1970s the infant mortality rate levelled, but
dien resumed its decline from die late 1970s. Sri Lanka’s infant mortality rate is currendy
14 per 1,000 live births (Department of Census and Statistics, 2002a).10 This rate
compares with an average of 76 for low-income countries, 33 for lower middle-income
countries, 28 for upper middle-income countries and six for high-income countries
(World Bank, 2002).11
Figure 2. Sri Lanka’s demographic transition, 1900-2000
Sources: Anand and Kanbur 1995; Department of Census and Statistics website accessed at
www.statistics, gov. ik
The child mortality rate for those under five years of age has also continued its decline in
recent years; the Demographic and Health Survey 2000 indicates diat during the 1990s it
fell from 27 per 1,000 live births to 15 (Department of Census and Statistics, 2002a). In
die mid 1990s, the maternal mortality ratio was 0.2 per 1,000 live births (Department of
Census and Statistics, 2002b). Life expectancy in Sri Lanka is now 71 years for males and
75 years for females (Department of Healdi, 2001) and total fertility is below
replacement level, at 1.9 (Department of Census and Statistics, 2002a). Many
commentators point out that these overall achievements do not mask large disparities in
progress between rural and urban location or between different population groups
(FIsiao, 2000; Rannan-Eliya et al, 2000). One exception is die significandv higher
7
mortality experienced by plantation estate workers producing products such as tea,
rubber and coconut, aldiough recently this has also been declining.12
Figure 3. Infant mortality rate and maternal mortality ratio, Sri Lanka 1900-2000
Note: Between 1921 and 1945, the maternal mortality ratio trend is based on linear interpolation
between values at 1921, 1925, 1930, 1935, 1940, and 1945. For 1945 onwards, annual values are
plotted.
Sources: Anand and Kanbnr 1995; Department of Census and Statistics website accessed at
www.statistics.gov. Ik
8
Table 1. Trend in life expectancy, Sri Lanka, 1946-2001
1946
1953
1963
1967
1971
1981
1991-1996
1996-2001
Life
expectancy
male
43.9
58.8
61.9
64.8
64.2
67.8
69.5
70.7
Life
expectancy,
female
41.6
57.5
61.4
66.9
67.1
71.7
74.2
75.4
Source: Department ofHealth 2001
Table 2. Trend in total fertility rate, Sri Lanka, 1963-2000
1963
1974
1981
1982-1987
1988-1993
1995-2000
Total fertility rate
5.0
3.4
3.7
2.8
2.3
1.9
Source: Department ofHealth 2001.
9
3. Development of the public health system in
Sri Lanka
Most writers agree that government inputs into the health sector have been decisive in
Sri Lanka’s health achievements both prior to and during die ‘breaktlirough’ period, and
subsequendy. The early expansion of public health interventions gave increasing
emphasis to preventive measures. Cholera was contained through quarantine measures
and sanitary7 improvements, major infectious diseases like smallpox and plague were
brought under control drrough vaccination programmes, and health education campaigns
helped reduce hookworm (Caldwell, 1986; Gunatilleke, 2000). In 1926, preventive and
curative services were amalgamated under die Ministry of Public Healdr and die first
healdi unit staffed with a physician, nurses, midwives, and a sanitary inspector was
established. The devastating malaria outbreak in 1934-35 stimulated plans to expand
healdi units across all rural areas and, following die end of the Second World War, a
dense but well-dispersed network of health facilities was developed.13 These have been
organised as an effective health referral system, ranging from maternity homes and
dispensaries at die lower levels to sophisticated teaching and national hospitals at die top
(Rannan-Eliya et al, 2000). Table 3 shows the growth of public sector healdi
infrastructure since die early twentiedi century7. The rapid extension in public healdi
facilities meant that, by 1950, Sri Lanka was already enjoying a higher bed-to-population
ratio and was providing more hospital admissions per capita, than any other country- in
South or Soudi-East Asia in 2000, except Singapore and Malaysia (Rannan-Eliya, 2001a).
Newly available health technologies were also embraced. For example, Sri Lanka’s
campaign against malaria dirough DDT spraying (and later through the use of maladiion)
is well known, and the debate about its contribution to total mortality7 decline during the
‘breakdirough’ period has been particulady vigorous.14 The attention given to maternal
and infant healthcare early in the twentieth century was further enhanced mid-century7.
Household visits by7 health workers to identify7 pregnant women and encourage their
referral for antenatal services and institutional delivery assisted by7 trained midwives, and
to provide care and immunisation of their newborns, expanded. Gunatilleke (2000) states
that by 1951, 50 per cent of births already7 occurred in medical institutions or with the
care of trained personnel. A World Bank mission to Sri Lanka in 1952 noted the growing
demand for institutional childbirth, saying Tractically everywhere one goes, be it a
provincial, district, or rural hospital or maternity home, there are a large number of
waiting maternity7 patients and like other hospital cases they are fed and treated free of
charge’ (IBRD, 1952, p. 397, cited in Pathmanathan et al, 2003). In 1953 public
healthcare was made freely available to the entire population. In 1968, Sri Lanka’s (noncoercive) family planning programme was operationalised with tire establishment of tire
Family7 Health Bureau.
Authors argue that the dynamism of public health intervention slackened somewhat in
the 1970s and 1980s, as from 1977 tire government progressively implemented reform
measures aimed at liberalising the economy7 and confronted fiscal pressures (UNICEF,
Colombo, 1988; Dreze and Sen, 1989; World Bank, 1998). In 1987, following the passing
of the 13'1’ Amendment to the Sri Lankan constitution, responsibility for lower-level
health services was devolved from the central Ministry of Health to eight Provincial
Councils, each of which established its own Provincial Health Ministry7 (Hsiao, 2000).
From tire 1960s, Sri Lanka also witnessed tire growth of a private health sector, especially
10
Table 3. Growth ofpublic health facilities andpersonnel, Sri Lanka, 1926-2001
No. of
beds *
No. of
hospitals
No. of
maternity
homes
No. of
central
dispensaries
No. of
doctors
No. of
assistant
MPs
No. of
nurses
Population
per bed
Population
per health
worker
1926
1930
1935
1940
1945
1950
1955
1960
1965
1970
1975
1980
1984
1988
1994
1998
2000
2001
8089
9477
11893
11992
15650
19959
24821
29816
35167
37735
39518
42275
43877
44454
49577
53507
57027
57197
98
112
112
126
153
263
274
289
296
326
345
380
408
413
415
469
558
498
-
-
-
12
38
107
107
108
72
82
103
104
93
35
75
-
-
-
543
595
632
632
260
2340
283
283
292
332
355
339
336
350
370
395
404
-
285
341
339
404
514
674
952
1173
1494
1932
2118
2055
2239
2316
4047
6427
7963
8384
380
409
416
469
548
676
990
1107
1242
1205
1068
1018
984
1100
1357
1324
1349
1343
437
605
618
744
1165
1387
2210
3232
3642
5542
5685
6834
7400
8317
13060
14621
14716
15797
609
554
472
534
425
385
359
332
317
332
342
349
355
371
360
351
339
327
4472
3876
4084
3693
2986
2805
2100
1795
1750
1442
1522
1489
1509
1406
968
836
806
734
Notes: ^excludes maternity homes.
Sources:for data to 1994: Giinatilleke 2000;for 1998 onwards: Department of Census and Statistics 2002b.
11
for out-patient care.15 Private practitioners consist largely of public sector staff working in
a private capacity outside their official working hours, as well as full-time private
practitioners. Out-of-pocket fees fund private services, and the role played bv health
insurance is small.16 Rannan-Eliya ct al (2001) estimate that between 50 and 70 per cent
of the private sector case load is dealt with by government doctors acting in their private
capacity. A recent survey of full-time private doctors undertaken by die Institute of
Policy Studies in 2000 (Rannan-Eliya et al, 2003) indicates that there are now between
600 to 650 private practices employing about 1,200 physicians who provide about 15 per
cent of annual out-patient consultations. Table 4 shows the trends in public and private
sector sendees during the 1990s. Bv the late 1990s, it can be seen that while the two
sectors divide out-patient care almost equally, in-patient provision continues to be
overwhelmingly dominated by die public sector, die latter providing 94 per cent of such
care.17
Table 4. Trends in public and private provision ofhealth sendees, Sri Lanka,
1990s
1990
Public sector
In-patient admissions (thousands) 2,533*
Out-patient visits (thousands)
28,401*
1992
1994
1996*
1997
3,023
36,827
3,204
35,276
3,339
35,348
3,454
39,503
153
176
204
32-36,000
153
986
176
1,502
204
1,617
Private sector (total)
In-patient admissions (thousands)
117
135
Out-patient visits (thousands)
30-34,000 32-35,000
Private sector (hospital)
In-patient admissions (thousands)
Out-patient visits (thousands)
117
568
135
810
No/e: ^excludes some dis!rids in Northern mid Eastern provinces.
Source: Hsiao 2000.
3.1. Other interventionist policies and factors facilitating Sri
Lanka’s health achievements
Before we review the evidence on die causal links between Sri Lanka’s public healdi
system and its health achievements and examine specific characteristics of die public
healthcare system in more detail, it is worth noting diat public intervention in Sri Lanka
has not been confined only to the health sector but has been multisectoral in nature. Tills
point is important because the broad scope of government support in Sri Lanka is argued
to have significandy enhanced social progress via positive synergies between different
sectors of intervention (Gunatilleke, 1985; Alailama and Sanderatne, 1998). Government
intervention in education has been active since before independence. A major expansion
of schools in the 1930s and 1940s, and die enactment of free education in 1947, meant
that literacy rates for the population aged five and above increased from 58 per cent in
1946 to 65 per cent in 1963 (Gunatilleke, 2000). Female education has increased rapidly,
and 77 per cent of (ever married) women in Sri Lanka now have above primary schooling
(Department of Census and Statistics, 2002a). Although gradually eliminated during the
1980s, the Sri Lankan government was also noted for its free or heavily subsidised
12
distribution of rice, which provided a minimum consumption level for the poor (Dreze
and Sen, 1989).
Interactions between these interventions and those in the health sector have been
beneficial. For example, in many developing country contexts, female education has been
shown repeatedly to have a strong depressing effect on child mortality and fertility.18
These relationships are present in the Sri Lankan context, indicating that in addition to
the direct benefits of schooling that children gain from a positive government
educational policy, they also benefit from their mothers’ education. Table 5 shows that
the under-five mortality' rate is over 30 per 1,000 live births among mothers with no
education or only primary schooling, and less than 20 among mothers with above
primary schooling (Department of Census and Statistics, 2002a). Evidence suggests that
this relationship exists because educated women are better able to deal with their
children’s illness episodes; they are more likely to take them to modern health facilities
for treatment, follow healdi providers’ instructions carefully, and take their children back
if medication docs not seem to be working. For Sri Lanka, writers also emphasise links
between female education and women’s use of healthcare for themselves.” Gunatilleke
(2000) comments that the associations between health and education have contributed to
die country’s intrinsic capacity' for progress, stating that ‘gains in health and education
keep compounding even during periods of economic decline because they have created
institutions, a knowledge base, and attitudinal changes drat are all conducive to furdier
progress’.
Table 5. Under-five child mortality rates by educational level ofmother, Sri
Lanka, 2000
Under five child
mortality rate(per
1,000 live births)
No schooling
(31-8)
Primary
33.3
Secondary
18.8
14.5
G.C.E (O/L)
G.C.E. (A/L) and higher
13.8
All educational levels
20.8
Noles: the child mortality rates arefor the ten-yearperiod preceding the survey. Figures in parentheses art
based onfewer than 500 cases.
Source: Department of Census and Statistics, 2002a.
In addition to the government’s crucial role in expanding the supply of health and
educational facilities, factors within Sri Lankan society have fostered a high level of
demand for such services and have therefore enhanced receptivity' to them (Caldwell,
1986; Caldw'ell et al, 1989). Historically high levels of women’s autonomy, and interest in
education for girls as w'cll as boys, have been important drivers in the use of governmentled supplies. Female children in particular are likely to have benefited from Sri Lanka’s
relative gender equality'; the country' is noted for having significandv lower son preference
than its South Asian neighbours. Unlike in drese countries little girls in Sri Lanka are less
likely to die in infancy and early childhood than boys and, similarly, anthropometric
indicators of health do not indicate gender bias against females in tire allocation of
resources such as food and healthcare (Abeykoon, 1995).
13
The high level of demand for and receptivity to publiclv provided social services have
also facilitated a broad social consensus as to Sri Lanka’s priorities, both among
successive governments and among die general population (Alailama and Sanderatne,
1998). Moreover, Sri Lanka’s competitive democratic political system and universal adult
franchise has enabled these priorities to be translated into actual government policy.20 In
fact, in 1928 the Donoughmore Commissioners specifically advocated the franchise of
women as well as men on the grounds diat political pressure from women was required
to make healtii services, particularly diosc implicated in lowering infant mortality, a
priority (Rannan-Eliya. 2001a). Rannan-Eliya (ibid) further asserts that ‘die introduction
of universal franchise is the single most important determinant of subsequent social
policy in Sri Lanka, and its healdi achievements. It led rapidly to the establishment of
comprehensive healdi services throughout die island in response to electoral demand, as
well as die introduction of universal free primary and secondary education’. Rannan-Eliya
goes on to explain that electoral pressures reflecting die public preference for equity and
universalism of access have influenced certain characteristics of the Sri Lankan public
healdi system. Such characteristics help account for die system’s success, and are
examined in more detail below.
3.2. Evidence on the causal links between Sri Lanka’s health
achievements and the public healthcare system
While the temporal connection between Sri Lanka’s health achievements and
government intervention in die health sector seems clear, establishing a direct causal
relationship between die two is more complex, and during the 1980s in particular, was
die subject of a great deal of analysis. There now exists much empirical work supporting
such a causal relationship. Some such studies employ a ‘cross-country’ approach that
attempts to identify whedier Sri Lanka is an outlier in terms of its social achievements
and to explain its relative position in terms of public intervention (see for example the
debate between Bhalla and Glewwe (1986) and Bhalla (1988), Iscnman (1980; 1987) and
Sen (1981; 1988)). Recent contributions to this literature include Hertz et al (1994) and
Shiffman (2000).21 In die former analysis, Sri Lanka remains a positive outlier in die
sample in terms of its rates of infant and maternal mortality and life expectancy, even
when variables relating to the availability of medical personnel and hospital beds,
sanitation, education, and wealth are included in the regression. Hertz et al attribute this
result to the government’s multisectoral interventionist approach and argue that the
resulting interactions between different interventions are impossible to capture
statistically. The absence of variables that adequately capture Sri Lanka’s facilitating
factors — such as women’s autonomy and political processes — may also be relevant.
Anodier approach is to use econometric analysis of Sri Lankan time-series data. For
example, Anand and Kanbur (1995) find that for the period 1952-81 income growth
alone would not have achieved Sri Lanka’s positive health outcomes; public intervention
has had a significant effect. Rannan-Eliya (2001b) incorporates Sri Lankan time-series
data for the period 1947-98 with data from four other countries with similar healthcare
systems to form a panel data set.22 Having controlled for income growth, he finds a
government health spending impact that is larger than that found in previous studies for
Sri Lanka and for other countries, and argues that this result may indicate the particular
value of die specific healdi strategies pursued in diese five countries.21 In all diese studies,
public intervention in the health sector is shown to have a positive impact on the level of
infant mortality, consistendy indicating diat children have benefited from the
government’s pro-active policy.
14
4. What makes Sri Lanka’s public healthcare system
successful?
Having outlined the development of Sri Lanka’s public health system, as well as the
literature attributing a causal relationship between it and the country’s health
achievements, we now consider some of the characteristics of the system that have made
it unusually successful. Why has Sri Lanka’s public health system made such a difference?
We discuss a number of characteristics in turn. Where appropriate, we focus on these
aspects of die health system as they relate particularly to child and maternal health.
4.1. The public healthcare system as an equitable system
Equity applies to different aspects of die healdicare system. Writers note that Sri Lanka’s
system is equitable in several respects and that its equitable nature contributes to its
success. Equity of access and service delivery' is one of the system’s strengths and has
been achieved via several mechanisms (Rannan-Eliva, 2001a). The expansion of a dense
but well-dispersed network of healdi facilities has minimised travel costs to diem. Hsiao
(2000) states tiiat most rural people live within 5-10km of a peripheral health facility. The
proximity of health facilities has also meant that information and knowledge about their
services and willingness to use them has spread quickly.
Odier barriers to access have been avoided by' providing services free of charge. Even in
the face of resource constraints, user fees have remained an undesirable policy option
pardy' because Sri Lanka’s democratic system makes diem politically cosdy (Hsiao,
2000).24 The egalitarian values to which the pohtical system responds has meant that
equity of access has remained a priority, and given the government’s resource constraints,
it has been maintained by' reducing unit costs (see more on this below). The rapid growth
of a private hospital sector - which could have creamed off die best medical staff and
left the poor widi access to inferior services — has been discouraged by maintaining die
technical quality of government facilities at a level similar to that offered by private
practitioners.25
All diese mechanisms are associated widi high levels of use of public health services.
Moreover, the well-developed network of facilities means that poor rural populations
benefit disproportionately, and access is more or less equal across income groups,
gender, and location. In terms of reproductive healthcare for example, Sri Lanka’s
Demographic and Healdi Survey' 2000 (Department of Census and Statistics, 2002a),
indicates diat 92 per cent of the births occurring in the five years prior to the survey took
place in government hospitals or maternity homes. Compare this high proportion widi
the corresponding statistic for Sri Lanka’s neighbour, India: data from India’s 1998/99
National Family’ Healdi Survey' shows that only 16 per cent of birdis occurring in the
three years preceding die Survey' took place in government medical institutions (UPS and
ORC Macro, 2000).26 Sri Lanka’s Demographic and Healdi Survey 2000 (Department of
Census and Statistics, 2002a) also shows that only 0.3 per cent of Last births in the five
years preceding die survey occurred among motiiers who had received no prenatal care.
Only 1.1 per cent of last births that occurred among modicrs in the estate sector were
without such care. The 2000 Survey’ also indicates diat 97 per cent of children under five
years of age (excluding infants), with a Child Health Development Record issued by the
health autiiorities, have complete immunisation coverage (Department of Census and
Statistics, 2002a).27
15
Commentators agree that tire high level use of publicly provided reproductive health
services reflects the fact that the government successfully provides mothers with
integrated and convenient maternal and child health, family planning, and infant care
services of acceptable quality (Somanathan and ]ayawardhane, 2001). These services
include visits to pregnant women at home and provision of postnatal follow-up for all
mothers in the community. Rannan-Eliya et al (2000) argue that the public sector has
inherent comparative advantages over tire private sector in providing such services
associated with economics of scale in tire logistics involved. The provision of these
services and their widespread use have clear positive implications for both maternal and
child health.
Sri Lanka’s public healthcare system is equitably funded. It is financed largely by indirect
taxation, which has been found to be mildly progressive (Rannan-Eliya, 2001a).
Redistribution is further enhanced because it is die poorer income groups that benefit
most from public healdrcare services: Hsiao (2000) show’s drat die poorest income
quintiles receive a larger share of the benefits of taxation-funded health expenditure than
do die richest quintiles.28 Out-of-pocket expenses to private healdrcare providers increase
monotonically with income (Rannan-Eliya, 2001a). Sri Lanka has not achieved this
redistributive effect by attempting to explicidy target or means-test access to public
healdi services (ibid). Such services are available to die entire population, and exist
alongside a private health sector of similar technical quality. We noted above diat die
private sector mainly provides out-patient care, while there continues to be a near public
sector monopoly in the provision of in-patient services. This healdrcare structure enables
wealthier patients bodi to self-select out of die public sector for out-patient care in
response to the private sector’s better quality in terms of aspects such as waiting times
and overcrowding, but to continue to benefit from and support the public sector for in
patient care. Pro-poor targeting is dierefore implicit, but is much more effective dian diat
achieved in most other developing countries. This point is clearly shown in Table 6 on
die relative distribution of public health spending by die poorest and richest income
quintiles for a cross-section of countries.29
Table 6. Distribution ofpublic health spending by the poorest and richest income
quintiles, various countries, 1980s and 1990s
Sri Lanka
Jamaica
Malaysia
Hong Kong
Ireland
Bangladesh
Brazil
Egypt
Kenya
Indonesia
Vietnam
Ghana
1991
1989
1989
1989
1987
1996
1985
1995
1993
1989
1992
1992
Poorest quintile
30
30
29
27
27
17
17
16
14
12
12
11
Richest quintile
9
9
11
16
15
25
42
24
24
29
29
34
Note: Note the similarities between Sri Kanka, Jamaica, Malaysia, Hong Kong and Ireland - see
endnote 29.
Source: }kannaibE.liya 2001a.
16
These characteristics indicate that equity of access and pro-poor redistribution are
important elements in Sri Lanka’s healthcare system success. However, Rannan-Eliya
(2001a) points out that both are rejected by the WHO’s health system assessment
framework (WHO, 2000), suggesting that international health policy advisers give
insufficient recognition to diem.50 The Sri Lankan experience suggests that ensuring
universal access to public health services makes for their high utilisation among all
population groups, and together with a complementary private healthcare system,
effectively redistributes health spending and health services towards the poor without the
need for explicit targeting (Rannan-Eliya, 2001a).
4.2. The public healthcare system as an efficient system
Sri Lanka’s public healdicarc system has developed without mobilising resources beyond
what most other developing countries spend, and at significantly less cost than the World
Bank’s ‘minimum cost-effective package’ of basic preventive and curative health services
(World Bank, 1994). Total annual public spending on health has generally averaged less
than two per cent of GDP, although Gunatillcke (2000) shows that in the second half of
the 1950s and the 1960s, expenditure was slightly more than two per cent. Government
intervention in other areas, including education and food subsidy provision, has tended
to be higher dian that in health. Table 7 compares Sri Lanka’s recent health expenditure
as a proportion of GDP with that of a cross-section of odrer countries of comparable life
expectancy. Sri Lanka has not resorted to otiier methods of financing public health
services, such as user fees, or attempted to restrict access to healthcare.
Table 7. Public health expenditure as a percentage of GDP, various countries, late
1990s
Croatia
Slovakia
Columbia
Estonia
Poland
Lithuania
Bulgaria
Oman
Mexico
Venezuela
Public health
Life
expenditure
expectancy
(% of GDP)
2000
1998
Argentina
73.8
9.5*
Tunisia
73.3
5.7
71.2
China
5.2
Albania
70.6
5.1*
Thailand
73.3
4.7
72.1
Mauritius
4.7*
70.8
Sri Lanka
3.9*
71.0
2.9
Paraguay
72.6
Ecuador
2.6
Malaysia
72.9
2.6
Public health
Life
expenditure
expectancy
(%ofGDP)
2000
1998
73.4
2.4*
70.2
2.2
70.5
2.1*
73.2
2.0
70.2
1.9
71.3
1.8
72.1
1.7*
70.1
1.7
70.0
1.7
72.5
1.4
Note: *pnblic health expenditure data refer to 1999.
Source: United Nations 2002a
Instead, the Sri Lankan government has managed to achieve and maintain extensive
access to and use of healthcare facilities by continuously attaining efficiency gains in the
operation of facilities. Using personnel and infrastructure more intensively has lowered
unit costs, and the country' is noted for its efficient use of facilities. For example, a
recent comparison of total national per capita expenditure on reproductive health
services in Sri Lanka and Egypt illustrates that despite lower total expenditure levels in
17
Sri Lanka than Egypt, utilisation of services is higher in the former country (RannanEliya ct al, 2000).31 Once standardised in 1995 US dollars, Sri Lanka is estimated to have
spent §4.94 per capita in 1997 and Egypt is estimated to have spent §5.29 per capita in
1994/95. The public sector component of tlais expenditure is only slightly higher in Sri
Lanka than in Egypt. Yet in Sri Lanka for example, 92 per cent of women give birth in
government medical institutions (Department of Census and Statistics, 2002a). This
proportion compares with 56 per cent in Egypt.32 Rannan-Eliya (2001a) comments that
‘Sri Lanka’s unit costs arc considerably below those achieved in virtually all developing
countries. These low unit costs are achieved through high staff productivity, high
throughput rates, and correspondingly low lengths of stay’. He notes that this high
productivity is not a static characteristic, but is a product of an institutional system that is
able to achieve sustained improvements in performance over time. Specific explanations
of such ability are poorly understood, but possibilities include a strong and shared
institutional culture of serving the people as well as significant non-financial incentives
among staff that favour good performance.
This aspect of Sri Lanka’s public healthcare system is important because it illustrates that
centrally administered government systems can be inherently capable of continuous
productivity improvements, and that this characteristic enables such systems to maintain
universal access to healthcare (Rannan-Eliya, 2001a). Further investigation of the
institutional and organisational factors facilitating such improvements in tire Sri Lankan
system would provide valuable information about possible wavs to improve the
perfonnance of public health providers in other developing countries.
4.3. The public healthcare system as an insurance mechanism
against catastrophic33 illness
Sri Lanka’s public health system devotes an unusually high proportion of funds to
providing freely available in-patient hospital care. Using national health accounts
estimates for 1996, Hsiao (2000) shows that 69 per cent of government healdi spending
is on government hospitals. Table 4 shows that in-patient care is overwhelmingly
provided by such hospitals. Based on arguments about cost-effectiveness of resource
allocation, this strategy has been persistently criticised by international bodies, including
the World Bank. Their advice has been to reallocate money to non-hospital, out-patient
care because such treatments are typically more cost-effective (ibid). However, writers
argue that cost-effectiveness of resource allocation has never been the underlying
objective behind the Sri Lankan government’s use of funds. Instead, the government’s
orientation towards hospital care reflects a political concern to meet public demand for
such services in a situation where demand has risen faster than the capacity to pay for
private services and the supply of private providers, and where there is major market
failure in health insurance for in-patient care.34 The facilitating role of Sri Lanka’s political
system is important here, and Hsiao (2000) comments that ‘tire state responded to
aggregate social preferences as expressed through the ballot box, when it emphasized
hospital care and not to some technical view of what was optimal or rational’.
In tire Sri Lankan context therefore, the government recognises that, irrespective of its
cost-effectiveness or indeed its impact on health, there are strong insurance arguments in
favour of committing substantial public funds to provide in-patient care. Such a
commitment maximises marginal social welfare and is pro-poor because die less well-off
are unlikely to be able to afford eidrer private sector care or insurance for hospital
treatment, and in the absence of publicly provided services, would struggle to cope with
18
illness costs.3S In die area of reproductive healthcare, Rannan-Eliya et al (2000) point out
that a large proportion of government resources is devoted to in-patient delivery care and
that diis is reflected in die fact that most mothers give birth in an institution. As we have
seen from the Demographic and Health Survey 2000 (Department of Census and
Statistics, 2002a), most choose to do so in a public institution rather than incur the
expense of delivering in a private hospital. To the extent that fertility is higher among
poorer women, die public provision of free delivery services benefits them
disproportionately, protecting them from either going without care or having to pay for
it. In the light of die Sri Lankan experience, Rannan-Eliya (2001a) notes that the WHO’s
healdi system assessment framework (WHO, 2000), that recognises only health goals as
being health system goals, is likely to result in a seriously deficient view of health policy
that may discourage interventions that arc pro-poor in a much more general sense. One
policy implication that emerges from this characteristic of Sri Lanka’s public healthcare
system is to ensure delivery of public sector healthcare, particularly those aspects of care
whose provision represents a form of social insurance for the poor.
4,4. Complementarity between the public and private
healthcare systems
We have already commented on some aspects of the complementarity between the
public and private healthcare systems in Sri Lanka. We have mentioned that the public
system focuses on in-patient care, while die private emphasises out-patient sendees, and
that diis division, together with the maintenance of similar levels of technical quality but
differentials in consumer quality- across die two sectors, makes for a pro-poor
redistributive effect. The latter is achieved as public facilities arc available to everyone,
but wealdiier patients self-select out of die public system for out-patient care, aldiough
continue to use and support it for in-patient services.
We have already noted above that in terms of reproductive health practices, most
mothers in Sri Lanka choose to give birdi in government institutions, so it is important
that die public sector can deliver such services. Wc have also noted that infant and child
healdi services are also predominandy provided by the government sector, indicating that
children’s health benefits from a well functioning public hcaldicarc system. On the other
hand, general gynaecological care is largely privately provided and most women choose
to consult a private practitioner for such care; it is less expensive than assistance at
childbirth, and is more suited to out-patient services as it is tends to be client-initiated
and individually-spccific (Rannan-Eliya et al, 2000). In addition, in such cases patients
may value quality’ aspects such as privacy and confidentiality that diey may feel are more
forthcoming from a private practitioner.
The Sri Lankan experience dierefore indicates that reproductive healthcare is not solelv
dependent on public healdi services, and health policy’ in general must take account of
die varied choices women make when accessing care. That they may prefer to use the
private sector for certain aspects of care, even when the public system is working well,
means that attention should be given to its performance in addition to diat of die public
sector. Specifically, greater awareness should be given to die private sector’s role in
providing out-patient services. Rannan-Eliya et al (2003) note that diis has not been the
case in die World Bank’s 2001 Poverty Reduction Strategy’ Paper for Sri Lanka, in which
expansion of die private hospital sector is given more consideration as part of a povertyreducing health strategy’. The Sri Lankan experience calls fordi die need for more data to
19
be collected on the private out-patient sector so that we are better able to understand its
role.
Other aspects of complementarity are also significant, and contribute to the success of
Sri Lanka’s healdicare system as a whole. Rannan-Eliya (2001a) argues that the public
healthcare system has played a crucial role in creating and maintaining quality in Sri
Lanka’s private healdtcare system. This has occurred in several ways. Firsdy, die early and
widespread expansion of public healdi facilities has encouraged Sri Lankans to leam
about and use modem health facilities, better enabling diem to select appropriate
providers, w-hcther public or private. This has meant that although use of the private
sector has grown, it has not involved die expansion of unqualified, informal providers
diat characterise diis sector in many other developing countries. Hsiao (2000) provides
survey data showing diat only two per cent and 7.6 per cent of patients who reported
themselves sick in 1996/97 consulted Ayurcdic practitioners in the government and
private sector respectively.
Secondly, as we noted above, die majoritv of private sector healdi providers are
government staff working in dicir private capacity outside official working hours. The
government provides die opportunity of private practice to public sector doctors because
it recognises their need to supplement their income as well as the role of the private
sector in relieving pressure on die government system.36 This structure means diat most
private sector practitioners are trained in die public sector and that the public sector is
also able to retain dieir services.37 Studies of die characteristics of full-time private
general practitioners (GPs) indicate diat diey also provide quality services. Among his
sample of private GPs and Ministry of Healdi providers taken during die mid-1980s,
Varnam (1987) reports diat die private providers were older and more experienced than
die public providers and had all completed some post-graduate training in addition to die
compulsory government service. He also comments that diey spent longer with patients
and were more likely to keep patient records, perform immunisations, and conduct tests.
A recent survey undertaken in 2000 (Rannan-Eliya et al, 2003) confirms the existence of
relatively high quality care among full-time private practitioners. It is important to note
that this quality exists without regulation of the private health sector and in a situation
where lack of capacity may make effective regulation difficult (ibid).
These aspects of complementarity between the public and private health systems in Sri
Lanka indicate that a successful public system can indirecdy bolster a high quality private
system. The latter does not necessarily preclude die former, and should not only be
advocated as an alternative to a failing public system (Rannan-Eliya, 2001a). A successful
private health sector can then in turn contribute to die achievement of positive healdi
outcomes. When a public health system works w'ell, the two systems can be mutually
reinforcing.
20
5. Current and future challenges facing Sri Lanka’s
public healthcare system
Sri Lanka’s exceptional record in health achievements continues to hold true today.
Nevertheless, the country now confronts various health and health policy challenges that
will become increasingly testing in the near future. In this section, we briefly examine
some of diese challenges, and discuss their policy implications.
Sri Lanka’s demographic success means that it now has a rapidly ageing population. The
shift in the population age structure that reflects this ageing process is quite dramatic,
and is illustrated in Figure 4.
Figure 4. Population pyramids, Sri Lanka 1960 and 2000
80+
75-79
70-74
65-69
60-64
55-59
50-54
45-49
40-44
35-39
30-34
25-29
20-24
15-19
10-14
5-9
0-4
1,000
800
600
males, thousands
400
200
0
200
400
600
800
1,000
females, thousands
80+
75-79
70-74
65-69
60-64
55-59
50-54
45-49
40-44
35-39
30-34
25-29
20-24
15-19
10-14
5-9
0-4
1.000
800
800
600
males, thousands
1,000
females, thousands
Source: United Nations 2002b accessed at http:/1esa.iiii.org.iinpp
03 9 31
According to United Nations data (United Nations, 2002b), in 1960 die population aged
60 and older already represented 5.7 per cent of Sri Lanka’s total population. By 2000,
this proportion had increased to ten per cent. In particular, die number of very elderly
people, aged 80 years and older who arc significandy more likely dian the younger elderly
to be chronically disabled or ill, is increasing rapidly in Sri Lanka. Hsiao (2000) comments
that diis ageing process will mean diat, by 2010, Sri Lanka will be die diird oldest country
in Asia, after Japan and Singapore.
Epidemiological change is accompanying ageing, and chronic diseases associated with
advanced demographic regimes, such as ischemic heart disease, cerebrovascular
conditions, diabetes, and cancers are becoming more prevalent. For example, aldiough
data from individuals seeking treatment as in-patients in government hospitals shows
that morbidity remains dominated by infectious conditions and injuries, it also shows a
clear increasing trend in chronic disease cases (Department of Healdi, 2001).58 At the
same time, continued fertility decline will eventually mean that the ratio of die elderly to
middle-aged women (the primary providers of informal, home-based care of die elderly
in Sri Lanka) will decrease. Rannan-Eliya (1999) comments diat Sri Lanka ‘can expect to
face a significant care issue from 2010 onwards (if not earlier) due to the decline in die
relative numbers of women in die middle age groups’.
These processes have important repercussions for die public healdicare system. They
raise issues about both the pattern of healdicare expenditure and the level of spending.
Sri Lanka’s system developed with an emphasis on the prevention and treatment of
communicable diseases, but the epidemiological transition means that patterns of
healdicare expenditure will have to become correspondingly more diverse and complex;
for example, personnel and technology- will have to be increasingly specialised and
sophisticated. Such requirements are relatively expensive, and coupled with die increased
demand for long-term care that accompanies the rise in chronic diseases, suggest the
need for additional funding. The contraction of relative numbers of informal care
providers will also probably result in an increasing need for government action to supply
formal support services to help families care for their elderly relatives and to look after
those elderly people without family (ibid). However, it is probably important not to
exaggerate die impact that population ageing will have on increased healthcare costs
(ibid). Limited available evidence to date for Organisation for Economic Co-operation
and Development (OECD) countries (for example, OECD, 1998), suggests diat die
impacts of ageing on total national health expenditures are more limited than usually
imagined. Nevertheless, Gunatilleke (2000) states that ‘the low’-cost health care system
that was effective in dealing widi the disease burden associated with communicable
diseases will have to undergo a major reorientation and acquire new capacity- to deal witii
these challenges’.
However, achieving such a reorientation is not easy. There is litde political consensus on
health sector reform and the shape it should take, and as yet litde public pressure for
reform. In addition, resource constraints persist. These constraints include financial ones.
Relatively neglected sectors, particularly economic infrastructure, increasingly compete
with the health sector for scarce public funds. Within the health sector itself, spending is
increasingly required to maintain salaries at a high enough level to limit die mounting
exodus of staff to the private sector, and to keep up with die sophistication of the
growing private hospital sector (Hsiao, 2000). These issues raise questions about die
capacity- of the government to also provide sufficient funds to respond to
epidemiological change. New- financing options may have to be considered.
22
Constraints also relate to human capital, and Sri Lanka is increasingly confronting a
shortage of specialist, nursing, and administrative and managerial staff at the same time
as it is facing an over-supply of physicians (ibid). Tire latter is leading to concerns diat
increasing numbers of newly qualified, inexperienced doctors will be tempted to enter
the private sector immediately, with dubious implications for quality of service. The
mismatch in supply of different categories of staff also implies the need for a more
coherent policy on the output of medical graduates, and the shortage of managerial
personnel raises doubts as to the capacity of the system to plan, manage, and monitor
necessary reform.3’ Lack of managerial capacity exists at both the central and provincial
levels and this, together with the absence of coherent dialogue between the two levels
about their respective responsibilities and commitments, makes for an environment that
does not seem to be conducive to reform (ibid).
In addition to tackling the rise in chronic diseases associated with ageing and the
epidemiological transition, Sri Lanka continues to face high levels of malnutrition. These
high levels suggest diat die country’s health achievements have occurred among a
population that remains frail. For example, despite Sri Lanka’s impressive achievements
in infant and child survival rates, and die benefits children experience from the provision
of child-related public health services, it seems diat children’s health remains vulnerable
in several respects. Sri Lanka’s Demographic and Health Survey 2000 indicates that 17
per cent of children born in die five years preceding the Survey had low birth weight
(Department of Census and Statistics, 2002a). This compares with 19 per cent in the
1993 Demographic and Healdi Survey (Department of Census and Statistics, 1995). In
the 2000 Survey, 29 per cent of children aged 3-59 mondis were classified as
underweight, compared widi 38 per cent in die 1993 Survey (Department of Census and
Statistics, 1995; 2002). Despite diis improvement, Table 8 shows diat Sri Lanka’s
percentage of underweight children is still much higher dian diat for any other country of
comparable life expectancy for which data is available. Diarrhoeal diseases are still very
common among children and have been recognised as one of the main factors adversely
affecting growth and development in the early childhood years (Department of Census
and Statistics, 2002a).
These indicators reflect continuing poverty in Sri Lanka, and point to the need for the
government to foster participator}' economic growth in addition to its long-standing
emphasis on supplying pro-poor social services (Asian Development Bank, 1997).40 They
may also contribute to the rise in adult chronic and non-communicable diseases
mentioned earlier: die Barker hypothesis, which has mounting evidence in its favour,
argues diat foetal undernutrition later predisposes healthy populations to higher rates of
diabetes, heart disease, and many other chronic diseases. Sn Lanka may be particularly
prone to this link between poor prenatal and childhood nutrition and adult chronic illhealth.
Finally, we note some of die issues that arise from die growdi and increasing complexity
of die private healdicarc system in Sri Lanka. Given these trends, many writers
recommend diat the government moves away from its current levels of engagement with
die private sector, to develop a coherent policy on die appropriate public-private mix of
healdicare financing and provision that builds on the existing complementarity between
the two sectors (Hsiao, 2000). Although, as we mentioned above, evidence suggests that
die private sector generally provides reasonable quality services, and that die public
system contributes to this feature, better regulation is also recommended because of its
growing importance. However, as with odier aspects of reform, writers question the
capacity of die government to undertake such initiatives (Russell and Attanayakae, 1997;
Rannan-Eliya ct al, 2003). Lack of routine data collection on, and regular consultation or
co-ordination widi, the private sector mean diat policy-making is difficult. Tine legal basis
for regulating die private sector is outdated and, even if renewed, die systems required
for die effective registering of private practitioners as well as monitoring and enforcing
dieir standards arc not in place (Hsiao, 2000).41 Devolution from the centre to the
provinces is again relevant, as it leads to confusion about which level is responsible for
private sector regulation. Resource constraints also come into play, and the lack of
financial and human capital necessary for effective policy implementation weakens the
capacity for reform.
Table 8. Percentage ofchildren aged less than five years who are underweight,
various countries, 1995-2000
Underweight
children as a % of all
Life expectancy
children aged less
than 5 years
2000
1995-2000*
7
Columbia
71.2
24
71.0
Oman
8
Mexico
72.6
5
Venezuela
72.9
4
Tunisia
70.2
10
China
70.5
19
Thailand
70.2
16
71.3
Mauritius
Sri Lanka
33
72.1
5
Paraguay
70.1
15
Ecuador
70.0
18
Malaysia
72.5
Note: *data refers to the most recentyear available during 1995-2000. Note also that although the
figure on undenveight children given in textfor Sri Tankafrom the 2000 Demographic and Health
Suniey is lower than the one in this table, it is still higher than any other country in the table.
Source: United Nations 2002a.
Many of the health activists in Sri Lanka are fearful of die increasing privatisation of
services and the reductions in social spending following the recent poverty reduction
strategy' discussions (Peiris, 2003). Recent research on the impact of reductions in
education spending has shown a drop in literacy rates and attendance levels of girls
especially in the nordi and east of the country' (Save die Children UK, 2003). These
changes may well be reflected in the near future in health outcomes if social investment
continues to fall, as may' occur widi the drafting of die new poverty reduction strategies.
Sri Lanka needs to be supported to implement its multisectoral approach to health so
that it maintains good healdi outcomes and continues to lead die world in providing
good health at a low cost. Interest will no doubt remain as to the progress and success of
die reform process.
24
6. Conclusion
Sri Lanka is well known for its exceptional health achievements, and the role of
government intervention in attaining diese has been, and continues to be, a topic of great
interest, both for Sri Lanka itself and for the policy lessons it offers to other countries,
particularly in their efforts to lower infant and maternal mortality. All studies concur that
the public healdicare system plays a crucial role in Sri Lanka’s positive health outcomes,
including those of children. However, die importance of bodi government intervention
in other sectors and die beneficial synergies that result from such a multisectoral
approach to service provision must also be acknowledged. Cultural and political factors
that enhance die demand for government-provided services have proved essential,
demonstrating die importance of an informed civil society in creating the demand for
accessible and affordable services. The proportion of GDP spent on health provision
and Sri Lanka’s low income status demonstrate that a comprehensive health system can
be achieved with comparatively modest resources.
We have argued that several characteristics of Sri Lanka’s public healthcare system
contribute to its success. Particularly notable has been its ability to provide universal
access to health services, as well as its equitable and efficient funding. These features
create a pro-poor redistributive effect in die provision of government health facilities,
and help maintain high utilisation of them. Equity of access has remained a political
priority despite resource constraints and periodic international pressure to adopt
different models. It has been maintained over time by the lowering of unit costs. The
large proportion of government healdi expenditure on in-patient hospital facilities works
to furdicr protect die poor by providing them widi an insurance mechanism against the
costs of illness.
The government healdi sector exists in tandem with a growing private healthcare sector,
predominandy providing out-patient care. It seems diat this complementarity of service
provision, coupled with die maintenance of similar levels of technical quality across both
sectors, means diat even relatively wealthy patients still largely use the public sector for
in-patient care, and therefore continue to support public provision of such care. The
public health sector has also probably indirectly contributed to the quality of services
offered by die private sector, which in turn increasingly play a role in Sri Lanka’s health
achievements. The Sri Lankan experience dicreforc demonstrates that public healthcare
systems can work well, and that when diis is die case, diere can be a mutually reinforcing
relationship between this sector and private health providers.
However, Sri Lanka also indicates that as demograpliic progress occurs, and the diversity
of healdicare providers increases, healdi system reform becomes increasingly necessary.
It is now perhaps within this context diat the Sri Lankan case will continue to provide
invaluable insights into healdi systems and how diey can be made to work well.
6.1. Implications for policy
Sri Lanka’s success in healdi provision must be seen in the context of the island’s
changing demographic structure, itself partly a result of die success of the healdicare
system. However, notwithstanding the challenges posed by demographic change and
other processes specific to Sri Lanka, die undoubted success of die Sri Lankan ‘social
model’ begs die question of whether there arc there lessons to be learned that can be
25
applied to other countries, and diat can be seen as alternatives to die standard Bretton
Woods and WHO approaches.
The first point to note is diat die system developed over a long period, and must be seen
as endogenously developed — die government has not been so aid-dependent as to be
obliged to follow the changing prescriptions of creditors and donors. In relation to this,
die government and die public sector have been subject to popular and political
pressures to provide sendees, and have responded. Many of these conditions do not
apply to countries diat have not yet achieved reasonable healdi indicators.
Neverdieless, die question arises as to whether the ‘social model’ may not in any case be
a necessary7 condition for developing a working health service. The Sri Lankan model has
developed at far lower costs (as a percentage of national output and in terms of costs per
capita) than that proposed by7 the World Bank in its minimum health package, because it
has emphasised efficiency7 and retention of good staff widiin a credible and responsive
institutional structure — people of all incomes want to use public hospital facilities, while
diose who can pay, can use out-patient facilities provided by the private sector.
The impact of failed cost recovery policies are well documented in other countries.
Failure in terms of reduced facility utilisation, inappropriate prescription practices and
financial inefficiency, pardy as a result of lack of facilities and reduced demand. Similarly,
die divisions created by the combination of private sector, high quality provision for the
rich and low quality7 provisions for die poor are apparent in many countries.
The present move in Sri Lanka towards reduced social spending (from diree per cent
GNP for education in 1990 to two per cent in 2001) is already having negative
consequences in the education sector, including falling literacy rates and reduced
attendance levels at school. If this trend continues, health outcomes will also be
compromised.
The Sri Lankan case at die very7 least shows diat countries would be well advised to
consider the standardised policies proposed by the World Bank and die WHO with some
scepticism before adopting diem.
6.2. Recommendations
26
•
The Government of Sri Lanka should continue to support public sector spending
through the interim Poverty7 Reduction Strategy Paper and explore equitable
measures to improve economic growth and reduce levels of malnutrition.
•
The Government of Sri Lanka should begin to increase engagement with the
private sector and increase government capacity7 for enhanced regulation and
policy coherency.
•
The World Bank should support further qualitative and quantitative research into
the reasons why7 user fees combined with mechanisms to target die poor are not
reaching die poor and why universal access results in pro-poor health outcomes.
•
International donors should support (bodi financially and in terms of policy) the
abolition of user fees while exploring pro-poor (equitable) alternative
mechanisms to ensure universal healthcare provision.
The World Bank should evaluate the impact of the investment policy outlined in
‘Investing in Health’ on access to healthcare for the poorest, and consider
supporting die implementation of pro-poor policies as expressed in the OECD
poverty and healdi guidelines.
DFID and WHO should carry out a child-focused analysis of the impact of
World Bank policies on health resource allocation for children, disabled and
elderly people and use the findings to influence World Bank/International
Monetary Fund policies and conditionalities.
WHO and die World Bank should analyse the institutional and organisational
factors which have enabled the centrally administered system of Sri Lanka to be
capable of continuous productivity improvements, thus maintaining universal
access to healthcare, in order to inform public health providers in other
developing countries.
WHO should apply lessons learned in Sri Lanka to die development of the
WHO primary healthcare lens.42 International donors should explore the
possibility of fully supporting five resource-poor governments to pilot Sri Lanka’s
approach for at least ten years, specifically looking at what the critical threshold
for recurrent investment is, regarding die amounts of support, the time taken for
gains to be made on statistical trends and the cost per person of these gains.
27
Bibliography
Abcykoon, ATPL, 1995, Sex preference in South Asia: Sri Lanka an outlier’”, Asia-Pacific
Population journal, 10/3: 5-16
Alailama, P and Sanderatne, N, 1998, ‘Social policies in a slowly growing economy: Sri
Lanka’, in S Mehrotra and R Jolly (eds) Development with a Human Face. Experiences in Social
.Achievement and Economic Growth, Oxford: Clarendon, 235-263
Anand, S and Kanbur, SM Ravi, 1995, ‘Public policy and basic needs provision:
Intervention and achievements in Sri Lanka’, in J Dreze, A Sen, and A Hussain (eds) The
Political Economy ofHunger, Oxford: Clarendon, 298-331
Asian Development Bank, 1997, Sri Eanlza. Responding to New Soria! Challenges, Asian
Development Bank
Basu, AM, 2002, ‘Why does education lead to lower fertility? A critical review of some of
the possibilities’, World Development, 30/10: 1779-1790
Bhalla, S, 1988, ‘Is Sri Lanka an exception? A comparative study of living standards’, in
TN Srinivasan and PK Bardhan. Rural Poverty in South Asia, Oxford University7 Press, 89117
Bhalla, S and Glewwe, P, 1986, ‘Growth and equity in developing countries: A
reinterpretation of Sri Lankan experience’, IPor/<7 Rank Economic Review, 1/1: 35-63
Caldwell, JC, 1986, ‘Routes to low mortality’ in poor countries’, Population and Development
Review, 12/2: 171-220
Caldwell, J, Gajanayake, I, Caldwell, P, and Pieris, I, 1989, ‘Sensitization to illness and the
risk of death: An explanation for Sri Lanka’s approach to good health for all’, Social
Science and Medicine, 38/4: 365-379
Department of Census and Statistics, Sri Lanka, 1995, Demographic and Health Survey 1995,
Department of Census and Statistics
Department of Census and Statistics, Sri Lanka, 2002a, Demographic and Health Survey
2000, Department of Census and Statistics
Department of Census and Statistics, Sri Lanka, 2002b, StatisticalAbstract 2002,
Department of Census and Statistics, accessed at www.statistics.gov.lk
Department of Health, Sri Lanka, '2SXA, Annual Health "bulletin, Department of Health
Dreze, J and Sen, A, 1989, Hunger and Public Action, Oxford: Clarendon
Grow Up Free From Poverty, 2003, 80 Million Fives: Meeting Millennium Development Goals
for Child and Maternal Survival, Save the Children UK and CAFOD
Gunatilleke, G, 1985, ‘Health and development in Sri Lanka - An overview’, in SB
Halstead, JA Walsh, KS Warren, Good Health at Eow Cost. Proceedings of a Conference held at
the Rellagio Conference Center, Rellagio, Italy, April 29 — May 3, New York: Rockefeller
Foundation, 111-124
28
Gunatilleke, G, 2000, ‘Sri Lanka’s social achievements and challenges’, in D Ghai, Social
Development and Public Policy. A Study ofSome Successful Experiences, Macmillan, 139-189
Hertz, E, Herbert, JR, and Landon, J, 1994, ‘Social and environmental factors and life
expectancy, infant mortality, and maternal mortality rates: Results of a cross-nadonal
comparison’, Social Science and Medicine, 39/1: 105-114
Hobcraft, J, 1993, “Women’s education, child welfare and child survival: A review of the
evidence’, Health Transition Review, 3/2: 159-174
Hsiao, WC, 2000, ‘A preliminary assessment of Sri Lanka’s health sector and steps
forward’, Cambridge MA: Harvard University'
Hurst, J and Jee-Hughcs, M, 2001, ‘Performance measurement and performance
management in OECD healdi systems’, Labour Market and Social Policy Occasional Paper No.
47. Paris: OECD
IBRD (International Bank for Reconstruction and Development), 1952, ‘Economic
development of Ceylon’, Part 2 ofSelected Fields ofDevelopment: Report of a Mission Organised
by IBRD at the Request of the Government of Ceylon, Colombo: Ceylon Government Press
UPS and ORC Macro, 2000, India National Family Health Survey 1998-99, Mumbai:
International Institute of Population Sciences
Iseruman, P, 1980, ‘Basic needs: The case of Sri Lanka’, IPonW Development, 8/3: 237-258
Isenman, P, 1987, ‘Growth and equity' in developing countries: A reinterpretation of die
Sri Lankan experience: A comment’, World Bank Economic Review, \/y. 521-531
Jayawardena, L, 1974, ‘Sri Lanka’, in H Chenery, M Ahulwalia, CLG Bell, JH Duloy, and
R J oily. Redistribution with Growth: Policies to Improve Income Distribution in Developing Countries
in the Context ofEconomic Growth: A joint Study, Oxford University Press
Langford, CM, 1996, ‘Reasons for die decline in mortality in Sri Lanka immediately after
the Second World War: A re-examination of die evidence’, Health Transition Review,
6/1: 1-24
Leach, ER, 1971, PnlEliya, A Village in Ceylon: A Study of Land Tenure and Kinship,
Cambridge: Cambridge University Press
Organisation for Economic Co-operation and Development (OECD), 1998, Maintaining
Prosperity in an Ageing Society, Paris: OECD
Padimanathan, I, Liljestrand, J, Martins, JM, Rajapaksa, LC, Lissner, C, de Silva, A,
Sclvaraju, S, and Singh, PJ, 2003, Investing in Maternal Health. Learningfrom Malaysia and Sri
Lanka, Washington DC: The World Bank
Peiris Sirimal, 2003, ‘Fast deteriorating health services in Sri Lanka’, People’s Health
Forum Sri Lanka
Rannan-Eliya, RP, 1999, ‘Economic impacts of demographic ageing: With special
emphasis on Sri Lanka and old-age income security'", IPS HPP Occasional Paper No. / /,
Institute of Policy' Studies of Sri Lanka, Health Policy Programme
Rannan-Eliya, RP, 2001a, Strategiesfor Improving the Health ofthe Poor — The Sri Lankan
Experience, Institute of Policy' Studies, Health Policy' Programme
29
Rannan-Eliya, RP, 2001b, ‘Contribution of income and public health spending to healdi
status in five developing countries during 1943-98’, unpublished paper, Cambridge MA:
Harvard University
Rannan-Eliya, RP, Berman, P, Eltigani, EE, Somanathan, A, and Sumadiiratnc, V, 2000,
‘Expenditure for reproductive health sendees in Egypt and Sri Lanka’, IPS HPP Occasional
Paper No. 13, Institute of Policy Studies of Sri Lanka, Healdi Policy Programme
Rannan-Eliya, RP, Pande, B, Killingworth, J, and Somanathan, A, 2001, ‘Equity in
Financing and Delivery of Health Sendees in Bangladesh, Nepal and Sri Lanka’, Data
International Ltd, Nepal Health Economics Association, and Institute of Policy- Studies
Rannan-Eliya, RP, Jayawardhane, P and Karunaratnc, L, 2003, ‘Private primary care
practitioners in Sri Lanka’, in AS Yazbeck and DH Peters (eds) Health Polity Research in
Asia: Building Capacityfor Reform, Washington DC: The World Bank
Reilly, B, Ford, N, and DuBois, M, 2002, ‘Sri Lanka’s healdi sendees is a casualty of 20
years of war’, British Medicaljournal, 324/361
Russell, S and Attanayake, N, 1997, ‘Sri Lanka — Reforming the healdi sector: Does
government have the capacity?’, The Bole of Government in Adjusting Economies Paper 14, The
University of Birmingham, Development Administration Group, School of Public Policy
Save the Children, 2000, War Brought Us Here: Protecting Children Displaced Within Their Own
Countries By Conflict, Save the Children
Save the Children UK and Medact, 2001, The Bitterest Pill ofAU: The Collapse ofAfrica’s
Health Systems, Save the Children UK
Save the Children UK, 2003, Rights based approach to education: implicationsfor Sri Eanka,
Save the Children UK Sri Lanka Office
Sen, AK, 1981, “Public action and the quality of life in developing countries’, Orford
Bulletin of Economics and Statistics, 43/4: 287-319
Sen, AK, 1988, ‘Sri Lanka’s achievements: How and when’, in TN Srinivasan and PK
Bardhan, Rairal Poverty in South Asia, Oxford University Press, 549-56
Shiftman, J, 2000, ‘Can poor countries surmount maternal mortality?’, Studies in Family
Planning, 31/4: 274-289
Simetka, O, Reilley, B, Joseph, M, Collie, M and Leidingcr,J, 2002, ‘Obstetrics during
civil war: Six months on a maternity ward in Mallavi, Northern Sri Lanka’, Medicine,
Conflict and Survival, 18/3: 258-270
Somanathan, A and Jayawardhane, P, 2001, ‘Maternal and child healdi services in Sri
Lanka: A survey of primary level government providers’, Institute of Policy Studies
UNICEF, Colombo, 1988, ‘Sri Lanka: The social impact of economic policies during die
last decade’, in GA Cornia, R Jolly and F Stewart, Adjustment with a Human Face, Volume
II: Country Case Studies, Oxford: Clarendon Press, 238-271
UNICEF, 2003, UNICEF Humanitarian Action Report 2003, UNICEF
United Nations, 2002a, Human Development Report 2002. Deepening Democrag' in a Fragmented
World, Oxford University Press
30 V '
United Nations, 2002b, World Population Prospects: Tbe 2002 Revision Population Database,
http://csa.un.org/unpp/
Varnam, M, 1987, “Primary medical care in Sri Lanka: Hospital or general practice?’
journal of tbe Royal College of Genera! Practitioners, 37: 31-33
World Bank, 1994, World Development Report: Infrastructurefor Development. Oxford
University Press
World Bank, 1998, ‘Sri Lanka social services: A review of recent trends and issues’,
Poverty Redaction and Economic Management South Asia Region Report, No. 17748-CE, World
Bank
World Bank, 2002, IPor/<7 Development Indicators 2002, World Bank
World Bank, 2004a, World Development Report 2004. Making Services Work for Poor People,
World Bank
World Bank, 2004b, ‘Reaching tire Poor’ meeting in Washington DC, February 2004
World Health Organisation, 2000, World Health Report 2000, Geneva: WHO
World Health Organisation, 2001, Tbe Report on the Commission on Macroeconomics and Health,
Geneva Switzerland: WHO
Notes
1 1 his means that the poorest 20 per cent of the population benefit more from public services than the
wealthiest 20 per cent, who may choose to use private services but who still have to pav taxes.
2 Although the United Nations Convention on the Rights of the Child (UNCRC) was not ratified until
1989. Save the Children UK’s founder (Eglantine Jebb) established a set of tights for children which was
eventually incorporated into the UNCRC.
3 Dreze and Sen (1989) also distinguish ‘growth-mediated securin'’ as a development option in which the
state plays a major role in promoting social securin' by using the fruits of economic growth, either to
directly provide services or to ensure participatory growth. They mention South Korea, Hong Kong,
Singapore, Kuwait, and United Arab Emirates as following this route. They contrast both ‘support-led
securin ' and ‘growth-mediated securin'’ with ‘unaimed opulence’, where wealtii increases but is not used to
secure widespread social progress. Brazil and Oman are examples of the latter.
4 Millennium Development Goal 4 is to reduce child mortalin' by two-thirds between 1990 and 2015.
Millennium Development Goal 5 is to reduce maternal mortalin' by three quarters between 1990 and 2015.
See www.developmentgoals.org
5
Calculated from data in 3.6 in Department of Census and Statistics, 2002b.
6 The demographic transition is the process of sustained decline in mortalin' and subscquentlv fertility, such
that high and approximately equal death and birth rates eventually give way to low and approximately equal
rates.
7 Sri Lanka is well known for having relatively reliable official vital statistics data, although Hsiao (2000)
notes that diere is likely to be some underreporting of the infant mortalin- rate. Maternal mortality statistics
arc probably more accurate than those in many other developing countries because of the close
involvement of health providers in pregnancy, delivery, and the postpartum period.
s Life expectancy is the average years of life expected by a hypothetical cohort of individuals who would be
subject all their lives to the mortality conditions of a given period. It is expressed in years. The total fertility
rate ts the average number of children a hypothetical cohort of women would have at die end of their
reproductive period if they were subject during their whole lives to the fertility rates of a given period and
they were not subject to mortalin'. It is expressed as children per woman. A total fertility rate of 2.1 is
known as the replacement level of fertility, required for a population to replace itself.
9
These calculations are based on official data available at www.statistics.gov.lk
10
This statistic is for the five years preceding the 2000 Demographic and Health Survey.
11
For details on the country classification groups, see World Bank (2002).
12 Rannan-Eliya (2001a) notes that the unfavourable mortality conditions of the estate population followed
their disenfranchisement in 1947-48 and the related fact that they did not benefit from the expanding
provision of government health services. The subsequent narrowing of the gap between die mortality
conditions of the estate workers and the rest of the population followed die estate workers’ increasing
inclusion and re-enfranchisement in the political system in the mid-1980s. Following this, estate sector
health facilities, historically provided separately by British-owned plantation firms, were brought under the
supervision of the Ministry of Healtii in the 1990s. Rannan-Eliya comments that ‘since increasing
involvement of the government health system in estate areas has led to tiiis improvement in health
indicators, this should be interpreted as a positive indicator of the ability of the health system to reduce
inequity in health outcomes. It also underlines the importance of a political system which gives voice to the
socially marginalized’. We say more about both these points more generally later.
13 According to Rannan-Eliya (2001a), it was during the 1934-35 malaria outbreak that officials began to
recognise that severe illness frequently bankrupted poor families, and that this eventuality could be avoided
by providing free hospital care. We say more about this public health sector role later.
14 See the references to this debate in Caldwell (1986), Dreze and Sen (1989), and Gunatilleke (2000). See
also Langford (1996). For the 1990s, Hsiao (2000) reports that more than half the case load of malaria
mortality is in the North-Eastern province where the Ministry of Health faces difficulties in operating
effectively. For evidence on the adverse impact of the conflict on maternal health indicators, see Simetka et
al (2002). Although these studies indicate the unfavourable health impact of Sri Lanka s internal conflict,
Hsiao (2000) also points out that such adverse effects are partly mitigated by the fact that the Ministry of
32
Health is able to enforce junior medical graduate postings there (as anywhere else) for a minimum time
period.
15 As in most developing countries, there is a dearth of reliable data on the private healthcare system. The
Private Clinic Survey carried out by the Institute of Policy Studies in 2000 helps fill this gap (Rannan-Eliya
et al, 2003).
16 The exception to this latter statement is the financing of urban private hospitals (Rannan-Eliva et al,
2003).
17 Hsiao (2000) notes that the continued presence of the public sector in the provision of out-patient
services contrasts with the much lower levels in other South and South-East Asian countries such as India,
Bangladesh, Malaysia, Thailand and Indonesia.
18
Useful reviews of these relationships arc Hobcraft (1993) and Basu (2002).
19 In this context, Caldwell (1986) cites the work of the anthropologist Edmund Leach (1971). Leach
reports that despite the traditional role washerwomen had as midwives and authorities on pregnanev in the
Sri Lankan village he studied, pregnant women quickly responded to the establishment of a modem health
facility seven miles away, making the journey to the prenatal clinic and giving birth there. Caldwell
concludes ‘it is unlikely that such a complete break with the past would have happened in South Asian
societies with lower female educational levels’.
20 Britain granted internal self-rule to Sri Lanka in 1931 on the basis of representative government elected
through universal franchise.
21 Unfortunately, the latter docs not include Sri Lanka in its cross-section of countries. Nevertheless,
regression analysis of rhe maternal mortality ratio illustrates that the strongest correlate is female enrolment
in secondary school. Attendance at delivery by trained medical personnel is also important, as is the
percentage of GDP spent on health services. In the multivariate model, variation in countries’ income level
is insignificant. The study finds that the relationship between female education and trained delivery
attendance is strong (correlation coefficient is 0.709), suggesting that female education is an important
driver in the use of healthcare. Moreover, the correlation between female education and income is only
moderate, and that between trained delivery attendance and income even lower, confirming that it is
education rather than income that is more important. These results are consistent with writers’ arguments
about the significance of the Sri Lankan Government’s multisectoral approach to intervention.
22
The four other countries are Malaysia, Hong Kong, Mauritius, and Jamaica.
23 For other references to the literature linking Sri Lankan Government intervention with health
achievements see Drcze and Sen (1989).
24 Facing serious fiscal constraints in 1971, the government reintroduced user fees having abolished them
in 1951. However, as their rcintroduction was not matched by commensurate increases in funds, they were
abolished in 1977 (World Bank, 1998; Hsiao, 2000). Another funding option, the public financing of
private providers, has never been seriously considered by the government (Hsiao, 2000).
25 We discuss the complementarity between Sri Lanka’s public and private health systems in more detail
below.
26
The Survey shows that 17 per cent of such births occurred in private institutions.
27 Infants are excluded because they do not qualify for complete immunisation coverage before they reach
their ninth month, at which age the measles vaccination is prescribed. Complete immunisation coverage
refers to coverage by BCG, DPT, polio, and measles immunisations.
28 However, Hsiao (2000) also remarks that evidence suggests that the degree of redistribution appears to
have declined during the 1990s, although the reasons for this arc unclear.
29 Table 6 illustrates that the performances ofJamaica, Malaysia, Hong Kong, and Ireland arc similar to that
of Sri Lanka. Rannan-Eliya (2001a) attributes this to the fact that these countries’ health structures are
comparable to that of Sri Lanka, as just described in the main text. For more on the redistributive impact
of Sri Lanka’s health care system, sec Rannan-Eliya et al (2001).
30
The former characteristic is emphasised in the OECD country framework (Hurst and Jee-Hughes, 2001).
31 In this study, reproductive health services include pre- and postnatal care, care in childbirth, infant care,
obstetric and gynaecological services, and family planning. It excludes services intended to treat STDs,
including HIV/AIDS. However, Sri Lanka has been shown to be relatively free of H1V/AIDS (Hsiao,
2000).
52 This figure for Egypt is taken from the 2000 Demographic and Health Survey, accessed at
www.measurcdhs.com. It measures the percentage of births in die five years preceding the Survey that
took place in all types of health facility, so those tliat occurred in government institutions only are a smaller
proportion.
55 This relates to those illnesses which lead to liigh healthcare costs or hospitalisation. These costs often
lead to die selling of capital assets tlius increasing the poverty* of the poorest.
54 Health insurance contributes less than two per cent of total health sector financing in the in-patient
sector and issues of affordability and adverse selection limit access among high-risk groups such as the
poor and the ill (Hsiao, 2000).
35 As Rannan-Eliya (2001a) points out, the costs of illness include not only the costs of treatment, but also
costs incurred in caring for the patient as well as any loss of income experienced by both the patient and
the carer.
36 The opportunity to practise privately was revoked in the early 1970s, but restored later in the decade
(World Bank, 1998). Government supports the private health sector in otlier ways as well; for example, the
1998 Budget contained incentives in the form of duty exemptions to private health investors for import of
medical and dental equipment (ibid).
37 Hsiao (2000) explains diat although private sector incomes are substantially higher than public sector
ones, most Ministry of Health staff remain in the public sector in order to receive necessary specialist
training in the large government hospitals and to build up their reputations by rising through the ranks of
the public sector.
38
Government hospital records are die principal source of morbidity data in Sri Lanka.
39 The former is complicated by the fact that it is the Ministry of Higher Education that, without reference
to the Ministry of Health, determines the annual intake of medical students (Hsiao, 2000).
40 Government interventions aimed at addressing the problem of malnutrition have included nutritional
supplements to mothers and children and midday meals for school children (Gunatillekc, 2000).
41 The Nursing Homes Act of 1949 relates to the regulation of the private sector, but is outdated because it
only applies to nursing homes and does not cover private GPs or government doctors practising privately.
42 WHO agreed to develop a primary’ healthcare lens, during the 57th World Health Assembly, through
which economic and health policies could be viewed to determine their possible impact on health.
34
Save the Children
I St John's Lane
London
EC IM 4AR
UK
www.savethechildren.org.uk
Save the Children
gs-’
Position: 2601 (2 views)