AfDB 2014 www.afdb.org
E c o n o m i c
B r i e f
What policies should be implemented to address inequalities in health care in Tunisia?
CONTENTS
Summary p.2 1 – General Introduction p.4
Key Messages
2 – Indicators of Health Status and Use of Health Care Services p.5
• Despite the progress achieved, health inequalities remain considerable and relatively little known in Tunisia. In light of the analysis conducted, there is significant elbow room for reducing these inequalities. In Tunisia, there are significant inequalities in care consumption between governorates for similar needs (those related to reproductive health, for example). There are also significant differences in the health status of the population of these governorates. The life expectancy of 74.5 years in 2009 does not exceed 70 years in Kasserine and Tataouine, but reaches 77 years in the governorates of Tunis and Sfax. The analysis indicates that:
3- Territorial Inequalities in
- The overall inequality in health spending declined from 2000 to 2010. The breakdown of the Gini index shows that this movement is almost entirely explained by the decrease in inequality in pharmaceuticals spending, which accounted for 42.2% of health spending in 2010. This trend can be attributed to a greater availability of pharmacies throughout the national territory. - The items where inequality has worsened and that had an inertia effect were long-term illnesses (17% of expenditure), hospital stay and medical surgery (8.6%) and radio and scans (8% of health spending). Such spending is related to the demographic and epidemiological transition. - Dental care is characterized by unusually high levels of inequality and lack of access for the disadvantaged classes.
Health Care Facilities p.9 4- Trend in inequalities in health spending in Tunisia between 2000 and 2010 p.27
• The main recommandations in this context are as follow:
5- General Conclusion p.41
- From the supply side: (i) In the public sector, it is necessary to revitalize primary health care by improving the operation (ii) It is also important to strengthen Level II which seems to be the weak link in the system. Better coverage of the territory in terms of Level II beds should necessarily go hand-in-hand with the provision of more specialized physicians for the poorest regions in light of the demographic and epidemiological transition. (iii) Efforts should be made to ensure that at each level the system performs its assigned tasks under the best possible conditions. These tasks should be clearly defined. Each hospital institution should have a scheme of work that allows for coherent strategic management. (iv) The specific incentives that were introduced to encourage physicians to settle in deserted areas should be evaluated. Public-public and possibly private-public partnerships should be instituted. Also, it is important to negotiate with corporations an institutional framework to better regulate the opening of private practices. (v) It is necessary to determine measures that should be implemented to enhance health care delivery at local or regional level, as part of an overall regional development policy. - On the demand side: (i) It is important to reduce financial barriers to health care access by better targeting the poor who benefit from free medical assistance. (ii) Pharmaceuticals are a significant drain on the budgets of the poorest households and it is necessary to reduce this weight by ensuring good governance of public pharmacies. (iii) There is a need to ensure a better collective coverage of longterm illness, hospital stay and medical surgery, x-rays and scans. Knowing the profile of households that incur these expenses will make it possible to better target them, if need be. (iv) Dental care continues to be characterized by extremely high inequalities in expenses. Improved coverage of the territory in terms of availability of dental practices and greater public awareness of the importance of dental health should curb one of the causes of the inequality. Similarly, a special processing of reimbursement for dental expenses by health insurance, apart from the recurrent expenses, should contribute to reducing inequalities in dental care access. - On the institutional side: (i) It is necessary to aim at reducing social and regional inequalities in health care. (ii) There is a need to produce and monitor indicators for assessing the progress of specific categories not only at the national level but also at the local level. It is important to conduct periodic surveys on the status of health, health care use, or the failure to seek health care for financial reasons.
Bibliography p.42 Annexes p.44
Zondo Sakala Vice President
[email protected]
Jacob Kolster Director ORNA
[email protected] +216 7110 2065
This paper was prepared by Salma Zouari, Ines Ayadi and Yassine Jmal, under the supervision of Vincent Castel (ORNA) and Sahar Rad (ORNA) et Laurence Lannes (OSHD). Overall guidance was received from Jacob Kolster (Director, ORNA). Ahmed Rekik and Chokri Arfa suggested improvements to the preliminary version of this research. Asma Baklouti, Mariem Ellouze, Rahim Kallel and Abdessalem Gouider each made an input.
A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
Summary
Therefore, there is clearly a need to develop a strategy for strengthening and revitalizing primary health in the country as well as enhancing Level II.
I
n Tunisia, there are significant inequalities in care consumption between
governorates for similar needs (those related to reproductive health,
1-2- Regarding human resource allocations, the inequality between
for example). There are also significant differences in the health status
governorates has decreased, except for physicians whether in the
of the population of these governorates. The life expectancy of 74.5
public or private sector. Although there has been a significant drop
years in 2009 does not exceed 70 years in Kasserine and Tataouine,
in the number of inhabitants per physician from 2002 to 2010,
but reaches 77 years in the governorates of Tunis and Sfax.
the gaps have widened between the better endowed governorates and the less endowed ones, while the variation coefficients have
Three hypotheses were then made:
l
l
increased.
Households, whatever the level of their resources and even when they
The availability of free medical practitioners is characterized by high
benefit from social security coverage, have unequal access to care
levels of inequality; the relationship between the most endowed and
because of inequities in the provision of health care services in their
the least endowed governorates is 14.3. This is followed by dental
immediate environment.
practices (ratio of 11.3) and hospital beds (10.7). The most evenly
Despite the importance of social coverage, households assume an
distributed resources are pharmacies and paramedical staff.
average of 41% of health spending in the form of out-of-pocket
l
expenditures. Therefore, households have unequal access to care
It would be advisable to review the criteria for opening positions of
arising from inequalities in income distribution and illustrated by
public health physician at regional level and the institutional framework
unequal health spending.
governing private practices. Like the practice of pharmacy, the practice
Due to the importance of out-of-pocket health care spending, the
of dentistry and medicine on a free basis should be better regulated.
regressive (or progressive) nature of care spending and its inelasticity
Similarly, public-private and especially public-public partnerships
compared to income, can give them a potentially catastrophic and
(such as agreements between academic physicians and regional
impoverishing character that makes unequal access to care even
hospitals) that may make disadvantaged areas more attractive as is
more acute.
being considered for specialists could be a solution. However, the implementation of such partnerships should be accompanied by
These hypotheses were tested on the basis of available statistical
measures to ensure their effectiveness for all stakeholders.
data. Health policy recommendations have been made. 1-3- Lastly, since the status of health care facilities in a governorate 1- On the assumption that the availability of care provision, whether
cannot be analysed by reference to a single determinant, all the
public or private, and good coverage of the national territory in health
components of the sector and the complementarity between different
infrastructure contribute to the decline in inequality in access to care,
providers should be taken into account simultaneously. For this purpose,
we analysed the trend of provision indicators by governorate and the
we have integrated the various determinants of facilities (by category
dispersion of these indicators through the use of cross-sectional data
and overall) in order to arrive at relatively homogeneous groups (called
of the 2010 health map and various longitudinal indicators published
clusters) and calculated for each governorate, a composite indicator
in the Statistical Yearbook of the National Institute of Statistics for the
of care provision that measures its position compared to other
period 1997-2010. Three aspects were analysed: infrastructure, the
governorates as well as the progress that may be achieved over time.
availability of beds and the provision of human resources. Among the three components of health care facilities, the geographic 1-1- With regard to infrastructure and bed availability, it turned out
distribution of medical human resources stands out as the most unequal,
that only the availability of PHCs declined over the last decade. Level
with a significant concentration on the coast. Despite an increase in the
II, which is the reference for Level I, would not be very effective because
density of physicians, regional disparities have widened. Qualitatively,
it lacks adequate technical equipment and specialized physicians. We
the inequalities are even more blatant and more than 2/3 of specialists
suspect that patients are referred to Level III which takes the place
are found on the coast as regards not only rare specialties but also the
of Level II, thus causing inefficiencies.
most common such as gynaecology and paediatrics.
2 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
Economic Brief 2 0 1 4
Three governorates constantly fall within the most favoured cluster
•
AfDB
w w w . a f d b . o r g
l
the contribution of inequality in each SPY category or sub-category
l
the marginal effect - equalizer or non-equalizer – of the variation of a
whatever the aspect considered. They are Tunis, Sousse and Monastir.
to total health spending inequality;
Conversely, four governorates always fall within the most disadvantaged cluster: Jendouba, Kairouan, Kasserine and Sidi Bouzid. Between these
particular SPY on total health spending inequality.
two groups, the various governorates show more or less substantial deficits depending on the nature of the resources analysed. The scope
2-1- The overall inequality in health spending declined from 2000 to
of intervention required by each governorate may then be defined.
2010. The breakdown of the Gini index shows that this movement is almost entirely explained by the decrease in inequality in pharmaceuticals
Government intervention is necessary when there is a build-up of
spending, which accounted for 42.2% of health spending in 2010. This
inequalities. However, the choices related to the health sector and efforts
trend can be attributed to a greater availability of pharmacies throughout
to better allocate resources to priority areas can only be effective if they
the national territory.
form part of a comprehensive local development strategy in these areas. The reduction of economic, cultural and social differences between the
2-2- The items where inequality has worsened and that had an inertia
governorates can only facilitate and strengthen health reforms.
effect were long-term illnesses (17% of expenditure), hospital stay and medical surgery (8.6%) and radio and scans (8% of health spending).
2- Working from the assumption that inequalities in access to health
Such spending is related to the demographic and epidemiological
are linked to income inequalities, we assessed the inequality of out-of-
transition.
pocket household health expenditure and analysed the trends thereof and training through inequality indicators and their breakdown by
The reduction of the corresponding inequality requires specific
expenditure category. For this purpose, we referred to the individual
government policies that target the most vulnerable groups. The
data of the national surveys on household budget and consumption in
collective management of these expenditures still seems insufficient.
2000, 2005 and 2010. This data provides information on total health
Knowing the profile of households that incur these expenditures will
spending per person per year (SPY) and the various expenditure
help to better target them.
categories: routine medical care, special medical care, pharmaceuticals and medical devices or expenditure sub-categories (medical
2-3- Dental care is characterized by unusually high levels of inequality
consultations; dental care; radio, scanner and medical analysis; medical
and lack of access for the disadvantaged classes.
stay and surgery; special dental care; special radiology expenditure; childbirth; long-term diseases; drugs; other pharmaceuticals, etc.).
Improved coverage of the territory by dental practices and greater awareness of people about the importance of oral and dental health
This approach gave information about:
should curb one of the causes of this inequality. Similarly, a specific treatment for reimbursement made by health insurance that is non-
l
Overall inequality in health spending and its trends;
concurrent with current spending should contribute to the reduction of
l
inequality of SPY in each care spending item and sub-item;
inequalities in access to dental care.
3 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
1. General Introduction
S
ince 1956, the foundations for a universal system of health care
is heavily subsidized or free for the beneficiaries of free health coverage
delivery have been established in Tunisia. For three decades,
who are estimated at 27% of the population.
the resulting benefits have been improved over time and a social security system has been put in place for employees. However, public
Despite the important size of community coverage of medical care by
expenditure on the health sector has slowed since the 1990s, and
insurance or by the State budget, private spending still remains very
private practices have gradually substituted public health services that
high and is increasing. In 2010, health care was financed by public
have experienced some decline in their quality and availability. Household
budgets to the tune of 23.8%, by health insurance resources (CNAM
health spending has risen sharply, sometimes leaving a heavy dent on
with 27.7% and private insurance with 7%) and out-of-pocket
the household budget. The most vulnerable segments are not spared
household spending that covered 41.2%.
(Arfa, ElGazzar, 2013). Thus, the health system faces many challenges, and it is important to: Since January 2011, there is heightened awareness of inequalities in the health status of the population and health care use. More attention
l
Reduce regional disparities in the provision of health care services;
is paid to issues of equity in access to care and there is more concern
l
reduce inequalities in household health spending;
about a more egalitarian distribution of health services throughout the
l
limit the amount of out-of-pocket household spending.
country. Therefore, it is necessary to better define the situation of health inequalities and its recent trend, and to identify policies that can help
Indeed, although health infrastructure covers almost the entire 1
country , there are inequalities in availability between the different
address the above challenges.
regions. The health system is predominantly public, with 87% of bed capacity in public hospitals and 13% in private clinics. On average,
We will begin by recalling some household health status indicators
Tunisia has 123 physicians per 100 000 inhabitants. However,
and the use of health services (Section I). We will then analyse the
physician density is much lower in the poorest regions where most
inequities in health care provision (Section II). Lastly, we will also
of the beneficiaries of free health care are found.
2
address the spending inequalities and their sources (Section III). In this respect, we will mainly use data from the 2010 Tunisia Health
With regard to the financing of the demand for care, the National Health
Map published by the Ministry of Health as well as data relating to
Insurance Fund (CNAM) covers about 68% of the total population. It
the health sector published by the National Institute of Statistics (INS)
covers both public and private health care services in the country. The
in the Statistical Yearbook of Tunisia between 1997 and 2010. We
majority of physicians, laboratories, dentists and pharmacists are
will further use the individual databases of the National Surveys on
contracted with CNAM. There are three branches: the public branch,
Household Budget and Consumption conducted by the INS in 2000,
the private branch and the reimbursement branch. The public branch
2005 and 2010.
1
The health system includes: (a) primary health centres or primary health centres and local or district hospitals; (b) regional hospitals; and (c) university teaching hospitals. 2 Medical density in Tunisia is lower than the European average, which is more than 300 physicians per 100 000 people. It is the highest in the Maghreb (Algeria and Libya 120, Morocco 60, Mauritania 10) and occupies the ninth place in the EMRO region behind Lebanon (330), Bahrain (300), Qatar (280), Jordan (260), Egypt (240), Kuwait and Oman (180), Saudi Arabia (160) and ahead of Iran (90), Pakistan (80) , Syrian Arab Republic and Iraq (50 ), Sudan and Yemen (30) (MH, Tunisia Health Map 2010).
4 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
Economic Brief 2 0 1 4
•
AfDB
w w w . a f d b . o r g
2. Indicators of Health Status and Use of Health Care Services3
T
here are few studies and statistics on this issue. However, we
because reproductive health concerns the entire population in the
will refer to data known for their relevance and published regularly
same manner.
by the INS. The health status will be assessed through life expectancy
1-
and the infant mortality rate (IMR). These two indicators are particularly
Health status indicators
suited to the health inequalities study (Jusot, 2003). However, they are published in a systematic way only at the national level. The
Differences in life span may be seen as a synthetic indicator of social
mortality rate, which is a poor indicator of health status because it is
differences affecting health throughout the life cycle (Aïach, 2000).
sensitive to the structure of the population by age, is however available
There has been a remarkable increase in life expectancy at birth in
at the governorate level. It will only be used to assess the evolution
Tunisia (Figure 1). From only 58 years in 1956, life expectancy has
of its variation coefficient.4 Health service use will be analysed through
risen to 74.9 years in 2011. The continuous improvement of this
the data on reproductive health because of their availability and
indicator is applicable to both men and women.
Figure 1: Life expectancy by gender (1990-2011)
Male
2011
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
78 76 74 72 70 68 66 64
Female
The increase in life expectancy is due to a decrease in mortality
The mortality rate showed a decreasing trend. From 19.1 per thousand
rates in general, and the sharp decline in infant mortality in
in 1960, it dropped to 5.5 per thousand in 2011 (Figure 2). The infant
particular.
mortality rate fell from 120 per thousand births in 1966 to 14 per thousand in 2011 (Figure 2).
3
All the statistics in this section are derived from Tunisian Statistical Yearbooks published by the National Institute of Statistics. For a Y distribution of mortality rates with average Y, the coefficient of variation, noted CV, is derived from the variance. It is defined as the ratio of the standard deviation σ to the mean mortality rates: CV = σ /Y where σ2 = 1/ N ∑(Yi -Y)2 CV is used to compare the dispersions of distributions with different averages. The higher the CV the more dispersed the distribution will be. 4
5 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
Figure 2: Mortality rate
Figure 3: Infant mortality rate
25.0 20.0 15.0 10.0 5.0 0.0
200 150 100 50 1960 1964 1968 1972 1976 1980 1984 1988 1992 1996 2000 2004 2008
1990 1964 1968 1972 1976 1980 1984 1988 1992 1996 2000 2001 2008
0
Infant mortality is an indicator widely used in international comparisons.
The life expectancy of 74.5 years in 2009 did not exceed 70 years in
It is an indicator of robust health, revealing a country’s development
Kasserine and Tataouine, but reached 77 years in the Tunis or Sfax
level and the quality of its health care system. It depends on several
governorates (Map 1). Similarly, the decline in infant mortality has not
factors, including income, maternal educational level and the
been equally beneficial to all children regardless of their place of birth
effectiveness of preventive care provided to mother and child
(Map 1). In 2009, the infant mortality rate was 17.8 ‰ at the national
(Bouchoucha and Vallin, 2007). The decline in infant mortality rate may
level. In the South, it was 21 ‰, while in the Midwest it rose to 23.6 ‰
be attributed to both factors inherent in the health system (modernization
(Ministry of Regional Development, 2011). Map 1 shows two groups of
and better coverage of the country) and the evolution of Tunisian society
governorates at odds with each other: the first (Tunis, Sousse, Monastir
(improvement in the quality of life and an increase in the living standards
and Sfax) recorded the lowest infant mortality rates (IMR), while the
and the educational level of the population).
second (Kasserine, Sidi Bouzid and Kairouan) had the highest rates.
However, this overall positive trend hides significant disparities between
Lastly, the mortality rate by governorate (indicator sensitive to the age structure
rural and urban areas as well as between socioeconomic groups and
of the population) shows contrasting trends and especially an increase in
between the various governorates. We will focus on the regional aspect
the variation coefficient, indicating a rise in inter-governorate inequality and
of inequality and health status indicators.
greater heterogeneity of living conditions prevailing there (Figure 4).
Figure 4: Coefficient of variation in mortality rates by governorate (1978-2010)
0.25 0.20 0.15 0.10 0.05 0.00 1997 1998 1999 2000 2000 2001 2002 2003 2004 2005 2007 2008 2009 2010
6 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
Economic Brief 2 0 1 4
2-
Health care use indicators
•
AfDB
w w w . a f d b . o r g
women’s perception of health and medicine, and it depends on the alternative care available to them and their medical environment.
In general, health care use is related to health condition and inequality
However, it is often subject to social and family control (Gastineau,
in health care use primarily reflects unequal needs. Hence it is not
2003). Belonging to a family or community group, social habits and the
necessarily unfair. It is therefore necessary to analyse the inequalities
environment are likely to influence women's choices in this matter.
observed only when facing the same need. As such, the indicators related to reproductive health are well suited for such an analysis. They help to
With reference to the number of deliveries by governorate, we calculated
identify women who give birth and analyse inequalities between them.
three indicators: the home delivery rate, the hospital delivery rate and the clinic delivery rate.5
Tunisian Statistical Yearbooks each year publish statistics on the use of reproductive health care. We will try to see to what extent the use of
In 2010, the home delivery rate (or rate of medically unassisted childbirth)
such care is egalitarian.
was 7.6%. It varied greatly between governorates. Governorates with high home delivery rates include Monastir, Nabeul and Mahdia, all of which have relatively good health infrastructure (Figure 5).
However, it is important to note that reproductive health care use reflects
Figure 5: Home delivery rate in 2010 by governorate
Zaghouan Kasserine Manouba Sidi Bouzid Kairouan Siliana Mahdia Nabeul Monasr TUNISIA Medenine Tataouine Gaafsa Ben Arous Kebili Sousse Ariana Tozeur Sfax Gabes Le Kef Jendouba Beja Bizerte Tunis
50% 40% 30% 20% 10% 0%
Clinic delivery rate was 12.3%. With the exception of Monastir, it was
private health infrastructure (Figure 6).
generally higher in the major urban centres of the coast which have
Figure 6: Clinic delivery rate in 2010 by governorate
Ben Arous Sfax Nabeul Sousse Tunis Ariana TUNISIA Bizerte Médenine Le Kef Mahdia Monas r Gabes Kairouan Beja Gafsa Jendouba Manouba Sidi Bouzid Siliana Touzeur Kasserine Kebili Zaghouan Tataouine
50% 40% 30% 20% 10% 0%
5
The sum of these three rates is not a unit because some women do not report the place where they gave birth.
7 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
Lastly, hospital delivery is the dominant standard in Tunisia. On
are those where the private sector is important (Sfax, Tunisia...)
average, 67.5% of women give birth in hospital. Governorates
or where the use of medicine is relatively limited (Sidi Bouzid,
where the propensity to give birth in hospitals is lower (Figure 7)
Kasserine…).
Figure 7: Hospital delivery rate in 2010 by governorate
Jendouba Beja Tozeur Kebili Tataouine Gabes Le Kef Siliana Bizerte Gafsa Ariana Medenine Sousse Mahdia Kairouan TUNISIA Sfax Nabeul Tunis Manouba Sidi Bouzid Ben Arous Kasserine Monas r Zaghouan
100% 80% 60% 40% 20% 0%
By comparing the number of antenatal or postnatal visits in the public
show a lower frequency rate for these procedures. These governorates
sector to the total number of births registered in a year (assisted or
are better off in private infrastructure, and many women consult their
unassisted, in hospital or private clinic), the frequency of these procedures
gynaecologist. The number of public postnatal visits is much lower (0.57
may be determined. There is an average of three antenatal visits per
in 2010), but it is just as unevenly distributed among the governorates
delivery. Only Tunis, Sousse, Sfax, Monastir and Mahdia governorates
as the number of antenatal visits (Figures below).
Figure 9: Postnatal visits per delivery
Figure 8: Antenatal visits per delivery
1.5 8 6 4 2 0
1.0 0.5 Tozeur Kebili Zaghouan Kasserine Gabes Sidi Bouzid Gafsa Le Kef Bizerte Siliana Tataouine Ariana Beja Nabeul Kairouan Ben Arous Jendouba Medenine TUNISIA Mahdia Monasr Sfax Sousse Tunis
Kebili Gabes Zaghouan Kasserine Ariana Ben Arous Sidi Bouzid Nabeul Tozeur Mahdia Bizerte Jendouba Tataouine Gafsa Beja Siliana TUNISIA Le Kef Sfax Monasr Medenine Kairouan Sousse Tunis
0.0
•
More generally, the various governorates do not fall in the same
availability of funding or means of support to make the demand effective;
category for the two indicators, reflecting the importance of community •
and social determinants in health care use. It would be interesting to
existence of an offer or several offers to meet the need.
consider the factors that explain these differences in order to assess the fairness of the system (Fleurbaey and Schokkaert, 2011). The use
When one of the last two elements is absent, access to care becomes
of health care which determines the health status of individuals is
impossible and care will not be provided at the risk of leading to serious
mainly due to the interaction of three determinants:
vital and economic consequences. Disease causes a loss of income and can propel the individual into poverty. As such, these two dimensions
•
Existence of demand related to the expression of a need for health
deserve special attention because they have a determinant impact on
(disease prevention, disease treatment, reproduction, etc.);
access to care. We will devote the following sections to them.
8 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
Economic Brief 2 0 1 4
•
AfDB
w w w . a f d b . o r g
3. Territorial Inequalities in Health Care Facilities
he availability of a health service whether public or private and
hospitals represent approximately 50% of all public sector beds. The
good coverage of the national territory in health infrastructure
health system further includes the polyclinics of the National Social
contribute significantly to the reduction of inequality in access to
Security Fund, hospitals under the Ministry of National Defence and
health care (Gold Zeynep et al., 2009). After an overview of Tunisia’s
the facilities of the Ministry of the Interior and Local Development
health care system, we will analyse the availability of public and private
(Arfa and Elgazzar, 2013).
T
health resources in the 24 governorates and try to build a composite indicator of health care facilities that can help to assess inter-regional
In Tunisia, access to the various levels is open and not by referral. Hence,
inequalities and to monitor their trend.
the first line may start at the primary, secondary or tertiary level. The intense activity of emergency services in hospitals is an example.
1-
Overview of the health care system in 20106 The private health care sector has grown significantly: it accounts for
In Tunisia, the health care delivery system is primarily public although
about 14% of total bed capacity and 70% of advanced technology
there is a growing private sector (Arfa, 2007). Nationally, more than
services. In terms of human resources, it employs 48.3% of physicians
7
86% of hospital beds are in the public sector. The leading care
(55.6% of specialists and 42% of general practitioners), 77.5% of dentists
provider is the Ministry of Health. The public provision of health services
and 81.5% of pharmacists. Private clinics are mostly concentrated in
8
is structured in three levels of care. Primary health care is provided
major coastal urban areas (Arfa and Elgazzar, 2013).
by 2 085 primary health centres (PHC), with 2 923 district hospital beds (consisting of small facilities with an average of 27 beds per
Despite an equalizing trend and geographical accessibility deemed
facility) and maternity centres, which together account for about 15%
acceptable for front-line facilities, the distribution of health services
of public sector bed capacity. This care level implements preventive
in the country is characterized by a certain inequality that should be
health policy. It handles 60% of public sector medical outpatients and
evaluated and corrected.
more than 1.3 million reproductive health visits (perinatal consultations,
2-
contraception, STI, screening for female cancers, etc.). It manages
Trend in the distribution of health facilities
the health activities of all pupils and students at all levels (pre-school, primary, secondary, university, vocational training and others).
To study the trend in the distribution of health facilities in the country,
Level II health care is provided by 33 regional hospitals (RH), which
centres and hospital beds over the past decade. These two indicators
account for 35% of total bed capacity and medical specialists in the
characterize public health provision.9 We will complete them with
we will analyse the allocations of the 24 governorates in primary health
human resource indicators (public sector physicians and paramedical
public sector.
staff). We will also analyse the availability of private practices, Level III health care consists of a network of 24 hospitals and
pharmacies and dental offices in governorates. These three indicators
academic institutions with an average size of 405 beds. These
characterize private health provision.
6
The main source is the 2010 Health Map of Tunisia (Ministry of Health). In 2010, the theoretical public bed capacity is 19 565 beds, while private clinics account for only 3 029 beds. 8 When people need health care, they turn most often to primary health care services, which are the first point of contact with the system. In general, primary health care has a double function. First, it provides preventive and curative support for common diseases. Then it acts as an interface and when necessary, refers patients to higher levels; it facilitates their movement within the health system when more specialized care is needed. 9 We will not discuss any issues related to the efficient use of the infrastructure, (World Bank, 2008). Overall, the potential of district hospitals has been under-utilized because of the weakness of their technical facilities, which limits the scope of diagnostic and therapeutic care management. Regarding regional hospitals, despite generally satisfactory technical facilities, productivity is affected by the lack of specialists, who are more attracted to university hospital careers or private practice. Lastly, UTHs control most of the heavy equipment in the public sector. Skill levels are high, but the sector suffers from consultation congestion, due to the weakness of the second level, as well as an increasingly strong tendency for brain drain to private practice that offers significantly higher income levels (WHO, 2010). 7
9 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
For each indicator, we will refer to the per capita endowment or its
2003 (Figure 10). It increased from 4 795 in 2003 to 5 051 in 2010. This
inverse (the number of inhabitants per unit). We will see if, on average,
trend could mean a more efficient use of PHCs for the benefit of a denser
the indicator improves. In addition, the use of the coefficient of
population. It may also indicate less easy access to these centres. In
variation 10 will reveal whether, overall, the distribution of health
the latter case, the PHC facility, as the access point of the population
resources has become more or less unequal. Comparing the values
to the health system, is fulfilling its role of health prevention and curative
of the indicator at the beginning of the last decade and at its end, for
treatment of common diseases less than before.
each governorate, will allow description of the trend of the governorate. The data used are drawn from the Tunisian Statistical Yearbook for
However, the coefficient of variation of the number of inhabitants per
2006-2010 (Serial No. 53).
PHC according to the governorate shows a downward trend indicating
2-1
in inequalities is due primarily to the deterioration of the situation in
a reduction in inter-governorate inequality (Figure 10). The decrease
Trend in the distribution of public health care facilities
governorates such as Monastir, Bizerte, Sfax, Sousse, Nabeul, Ben 2-1-1. Primary Health Centres (PHCs)
Arous, Ariana and Tunis, as shown in Figure 12. These governorates are relatively well served by Level III and PHCs providing consultations
The number of inhabitants by PHC globally reflects a reversal trend from
almost daily.
Figure 10: Inhabitants per PHC (1998-2010)
Figure 11: CV inhabitants per PHC (1998-2010)
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
5100 5000 4900 4800 4700 4600 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
5100 5000 4900 4800 4700 4600
Figure 12: Inhabitants per PHC by governorate 25 000 20 000 15 000 10 000 5 000 0
2000 Tataouine Kebili Siliana Le Kef Tozeur Beja Zaghouan Mahdia Sidi Bouzid Gafsa Jendouba Kasserine Medenine Gabes Kairouan Ensemble Monasr Bizerte Sfax Sousse Nabeul Manouba Ben Arous Ariana Tunis
2010
10 For a Y expenditure distribution with average Y, the coefficient of variation (CV) is derived from the variance. It is defined as the ratio of the standard deviation σ to the average of expenses: CV = σ /Y where σ2 = 1/ N ∑(Yi -Y)2 CV is used to compare the dispersions of distributions with different averages. The higher the CV the more dispersed the distribution will be.
10 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB
Economic Brief 2 0 1 4
•
w w w . a f d b . o r g
In this respect, it should be noted that this indicator does not allow
improving the frequency of primary health care consultations rather
proper assessment of the availability of health services in regions
than the multiplication of small health centres.12
because it does not take into account the pace of consultations in 2-1-2. Hospital beds
the PHC. Indeed, primary health centres differ in their type and the pace of medical consultation observed there. In 2010, 1040 of the 2085 PHCs provided at most one day of consultation per week. The
Unlike the first indicator, the bed equipment rate or number of public
2 085 primary health centres are in fact equivalent to 870 full-time
beds per 1000 inhabitants shows a significant increasing trend in the
centres.11 In addition, the pace of consultations is not equally distributed
public provision of care (Figure 13). The number of public beds per 1000
among the PHCs. The proportion of PHCs providing medical
inhabitants increased from 1.74 in 2001 to 1.85 in 2010. Similarly, the
consultation six days per week is only 4.4% in Medenine, 4.8% in
coefficient of variation of the bed rate in governorates decreased (Figure
Tataouine, 9.3% in Tozeur, 8% in Mahdia, 8.6% in Kébili, 8.9% in Sidi
13). This therefore means reduced inter-governorate inequality.
Bouzid and 9.6% in Beja. Thus, in these governorates, most of the population does not have daily access to mobile, community primary
Despite these positive developments, the public bed equipment rate
health care services. Accordingly, efforts should focus more towards
varied in 2010 from 0.4 in Ben Arous to 4 in Tunis (Figure 15).
Figure 13: Bed availability rate (1998-2010)
Figure 14: CV bed availability rate
1.9
0.60 0.50 0.40 0.30 0.20 0.10 0.00
1.8 1.7 1.6 1.5
1998 1999 2000 2001 2002 2004 2005 2006 2007 2008 2009 2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
Figure 15: Public hospital beds per 1 000 inhabitants by governorate
2000 Ben Arous Ariana Sidi Bouzid Kairouan Kasserine Nabeul Jendouba Mahdia Bizerte Medenine Siliana Gabes Tataouine Beja Sfax Kebili Tunisie entère Le Kef Gafsa Monasr Sousse Manouba Zaghouan Tozeur Tunis
4.50 4.00 3.50 3.00 2.50 2.00 1.50 1.00 0.50 0.00
1999
1998
1.4
2010
11 The 2010 Health Map can be used to calculate for each governorate the number of PHC full-time equivalent and the number of inhabitants per PHC full-time equivalent, but this statistic is not available for earlier years. 12 Three evaluations of the primary health care system (1997, 2000 and 2004) were carried out by the Ministry of Health as part of the Health Districts National Development Programme. It is a programme developed since 1994 by the Directorate of Primary Health Care. The overall objective of PNDCS is to make all health units in the country able to manage the health status of the population through a set of preventive, curative, promotional and rehabilitation activities, and ensure coordination within and between sectors involving all health stakeholders. The PNDCS has two specific objectives: firstly, improving the (technical and relational) quality and efficiency of care at the primary health centres (PHC) and the district hospital and secondly, strengthening and involving the population in health management. The main recommendations were: (i) optimizing health delivery by moving from a logic of coverage with infrastructure (number of hospital beds, number of DHCs) to an approach of coverage with effective services (number of medical consultation days offered, range of hospital services with appropriate technical facilities); (ii) improving the dimensions of care quality, for example by adapting opening hours to the rhythm of the patient population.
11 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
Reducing inequality reflects both:
to ensure that at each of its levels, the tasks assigned are carried out in the best conditions. These missions should be clearly specified. Each
l
An improvement in bed equipment in many governorates among the
hospital institution should have a roadmap that allows strategic
least well off, notably: Sidi Bouzid, Kasserine, Siliana, Tataouine, Beja,
management consistent with the whole system.
Kébili, Le Kef, Gafsa, Tozeur... l
a worsening situation in other governorates. The number of beds per
2-1-3.
Public sector physicians
1 000 population decreased in Ariana, Sfax, Monastir, Sousse, Manouba and Zaghouan. These governorates are experiencing strong
There has been a significant decrease in the number of inhabitants per
population growth induced in particular by migration. Almost ten years
physician in the public sector in all governorates with the exception of
of stagnation of the bed equipment rate in Tunis or the reduction
Ariana. This decrease reflects an increase in public health care provision.
thereof in large cities such as Sfax, Monastir, Sousse, Ariana and
Overall, the number of inhabitants per physician dropped from 2176 in
Manouba are all the more disturbing trends since these cities are
2002 to 1569 in 2010 (Figure 16). Until 2008, this trend increased
university teaching hospital centres. The quality of training at the
inequality between governorates. However, since 2008, the inequality
patient’s bedside may be affected by the increasingly less favourable
gap is closing but is still significant (Figure 16). The number of inhabitants
conditions in which it takes place. As such, there is the risk of a vicious
per physician in the public sector varied, in 2010, between 493 in Tunis
circle that reproduces mediocrity. To decongest level III and allow it
and 3377 in Kasserine, a ratio of 1 to 6.8.
to devote more time to training and research missions, level II should be developed.
There is a clear improvement in physician availability in many governorates (Figure 18). However, at the same time, the situation
Indeed, the bed equipment rate analysed below takes into account beds
has changed very little in governorates like Kasserine, Medenine,
at the three levels. It therefore hides disparities within levels. In the absence
Nabeul and Kébili. Yet these governorates were initially less endowed
of detailed statistics for the study period, it was not possible for us to
with physicians.
appraise the paces of matching developments. However, interviews with stakeholders led us to conclude that the weak link in the system is level II.
Five governorates are better provided with public health physicians
Often this level is ineffective or non-existent and therefore needs to be
than the country as a whole. The number of inhabitants per physician
strengthened.
there is less than 1569. They are Tunis, Sousse, Monastir, Sfax (which
Finally, the quest for greater equity in the health system should not result
twice less the number of physicians; the number of physicians per
in levelling from the bottom, or in a substantial carrying forward of
inhabitant is higher than 3 000 in Kairouan, Jendouba, Sidi Bouzid,
activities from a certain level to a higher one. The system should be able
Medenine and Kasserine.
enjoy level III services) and Tozeur. Conversely, five governorates have
Figure 17: CV Inhabitants per public health physician
Figure 16: Inhabitants per public health physician
3000 2500 2000 1500 1000 500 0
A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
2010
2009
2008
k
n
2007
2005
a
12
2006
2004
2003
2002
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
0.37 0.36 0.35 0.34 0.33 0.32 0.31 0.30
Economic Brief 2 0 1 4
•
AfDB
w w w . a f d b . o r g
Figure 18: Inhabitants per public health physician by governorate 5000 4500 4000 3500 3000 2500 2000 1500 1000 500 0
2002
Tunis Sousse Monas r Sfax Tozeur Ensemble Zaghouan Manouba Mahdia Tataouine Ariana Siliana Bizerte Beja Gafsa Kebili Gabes Le Kef Ben Arous Nabeul Kairouan Jendouba Sidi Bouzid Medenine Kasserine
2010
in 2002 to 308 in 2010. The indicator has improved in all governorates
The overall improvement in the availability of physicians veils significant deficits in medical specialists (surgery, obstetrics, ophthalmology,
except Sousse, Monastir and Ariana. The trend of the coefficient of
orthopaedics, anaesthesiology ...).
variation across governorates indicates a gradual reduction of intergovernorate inequalities (Figure 19).
Better coverage of the national territory in level II beds is necessarily However, there are still significant inequalities. The number of
concomitant with better provision of these regions with physicians in general and specialist physicians in particular.
inhabitants per public sector paramedical staff varied in 2010 between 150 in Tunis and 720 in Ariana, a ratio of 1 to 4.8. The six governorates
2-1-4.
Public sector paramedical staff
best equipped with paramedical staff are Tunis, Tozeur, Sousse, Monastir, Gafsa and Kef. The six governorates least equipped are
The number of inhabitants per senior technician is a clear indication of
Zaghouane, Nabeul, Kasserine, Sidi Bouzid, Ben Arous and Ariana,
the increase in public health care provision (Figure 19). It went from 341
as shown in Figure 21.
Figure 19: Inhabitants per paramedical staff
Figure 20: CV Inhabitants per paramedical staff 0.60 0.50 0.40 0.30 0.20 0.10 0.00
360 340 320 300
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
280
13 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
Figure 21: Inhabitants per paramedical staff by governorate 1200 1000 800 600 400
2002
200
2010 Tunis Tozeur Sousse Monas r Gafsa Le Kef Kebili Beja Ensemble Tataouine Sfax Mahdia Bizerte SIliana Gabes Jendouba Manouba Kairouan Medenine Zaghouane Nabeul Kasserine Sidi Bouzid Ben Arous Ariana
0
Once more, better coverage of the territory with level II beds will involve
better. In these areas, even when the number of inhabitants per physician
better provision of these regions with senior health technicians and
is already low, this number has continued to decrease. By contrast, in
nurses.
the hinterland and in the least developed governorates, the propensity of physicians to settle there was low and the number of inhabitants per
2.2
Trend in the distribution of private health care facilities
physician remained high and/or was on the increase. To counteract this spontaneous location of physicians, it is important to negotiate with the
2-2-1.
Private practice offices
medical corps to revise the institutional framework governing the opening of private practice offices and/or to grant special incentives to physicians
The number of private practice offices has increased considerably
who settle in priority areas. The Order of 24 December 2009 in part
throughout the past two decades such that the number of inhabitants
enshrined the idea by providing for compensation for medical
per office has on average been divided by 1.66 in 10 years (Figure 22).
specialists13 practising in the private sector and contracted with health
Inequalities between governorates declined until 2004, but have since
facilities in priority areas (defined by the Order of 1 March 1995 issued
increased (Figure 22). Between 2004 and 2010, the situation improved
by the Prime Minister laying down priority health areas for the granting
in all governorates except Siliana, Sidi Bouzid and Tataouine.
of certain benefits). The effects could be assessed.14 In France, a solution to the “medical desert” phenomenon was the establishment of “medical home”, that is to say, a structure whose main advantage is that of
In 2004, the ratio between the governorate best provided with private practice offices (Tunis) and the least provided governorate (Siliana)
bringing together in one place many practitioners and several specialties
was 1 to 11.6. In 2010, this ratio rose to 14.3 (Figure 24). This problem
with the purpose of saving by pooling and sharing certain costs. However,
is not specific to Tunisia; the same situation prevails in several
these structures instead resulted in the creation of a wider geographical
developed countries, including France where it is called “medical
network of rural and “desert” areas. Proposals have been made to
desert” (Potvin Moquet, Jones, 2010; High Council of Public Health,
hamper the freedom of installation of physicians rather than encourage
2009 and Senate 2013).
them to open offices in areas where medical facilities are scarce. In particular, this means excluding from health insurance physicians who
Thus, in a context where physicians are free to choose their location,
choose to settle in already saturated areas. As a result, since their patients
there was a craze for major urban centres and areas where the
are not reimbursed by social security, it would be impossible for a young
purchasing power of the patient base is higher and the quality of life
physician to build a patient base (Senate, 2013).
13
TND 500 for specialists in surgery and obstetrics and gynaecology, TND 400 for all other specialties. There is concern that physicians may abuse this situation by diverting patients from the hospital to their private practice or by using hospital equipment for private purposes. Ethical standards and rules of governance should be enacted. Very strict controls must be implemented. 14
14 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB
Economic Brief 2 0 1 4
Figure 22: Inhabitants per private practice office
•
w w w . a f d b . o r g
Figure 23: CV Inhabitants/Private practice office
2010
2009
2008
2007
2006
2005
2000
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
0
2004
1 000
2003
2 000
2002
3 000
2001
0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00
4 000
Figure 24: Number of inhabitants per private practice office by governorate 16000 14000 12000 10000 8000 6000 4000 2000 0
2001
Tunis Sfax Ariana Sousse Ben Arous Ensemble Nabeul Medenine Monas r Bizerte Manouba Gabes Mahdia Beja Gafsa Zaghouan Le Kef Kairouan Kebili Jendouba Tozeur Tataouine Sidi Bouzid Kasserine Siliana
2010
2-2-2.
Dental offices
per office was divided on average by 1.76 between 2001 and 2009 (Figure 25). Inequalities between governorates have significantly
The number of dental offices has increased significantly over the
decreased, as shown by the trend of the variation coefficients
last two decades to the extent that the number of inhabitants
(Figure 25).
Figure 26: CV Inhabitants per dental office
Figure 25: Inhabitants per dental office
14 000 12 000 10 000 8 000 6 000 4 000 2 000 0
0.80 0.60 0.40 0.20
15 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
0.00
AfDB 2 0 1 4
Economic Brief
•
w w w . a f d b . o r g
Figure 27: Number of inhabitants per dental office by governorate 60000 50000 40000 30000 20000 10000 0
2004
Tunis Ariana Sousse Sfax Ben Arous Nabeul Bizerte Ensemble Monas r Manouba Mahdia Medenine Gafsa Le Kef Gabes Ensemble Jendouba Kairouan Beja Siliana Zaghouan Kebili Sidi Bouzid Tataouine Kasserine Tozeur
2009
Between 2001 and 2009, the situation improved in all governorates
2-2-3.
Pharmacies
and is particularly striking in Zaghouan, Siliana, Sidi Bouzid and Kasserine (Figure 27).
Overall, the number of pharmacies has increased faster than the country’s population, such that the number of inhabitants per pharmacy
In 2004, the ratio between the governorate most provided with dental
has been divided by 1.2 in 10 years (Figure 28). Accordingly, the
offices (Tunis) and the least provided governorate (Zaghouan) was 1 to
inequalities between governorates decreased significantly as shown by
30. In 2009, this ratio dropped to 11 (between Tunis and Kebili). So
the variation coefficient (Figure 28). The situation improved in all
there are still margins to reduce inequalities in dental office availability.
governorates between 2001 and 2010 (Figure 30).
Figure 28: Inhabitants per pharmacy
Figure 29: CV Inhabitants per pharmacy
8 000
0.50
6 000
0.40 0.30
4 000
0.20
2 000
0.10
0
60000
Figure 30: Inhabitants per pharmacy by governorate
50000 40000 30000 20000
2004
10000
2009 Tunis Ariana Sousse Sfax Ben Arous Nabeul Bizerte Ensemble Monas r Manouba Mahdia Medenine Gafsa Le Kef Gabes Ensemble Jendouba Kairouan Beja Siliana Zaghouan Kebili Sidi Bouzid Tataouine Kasserine Tozeur
0
16 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
0.00
Economic Brief 2 0 1 4
The improvement has been very significant in governorates that were
•
AfDB
w w w . a f d b . o r g
increased. Hence, it is necessary to develop an appropriate strategy
15
less provided with pharmacies. Despite this, in 2010, the governorate
for addressing this challenge. It would be advisable to review the
most provided with pharmacies (Tunis) had per capita 2.63 times
criteria for opening positions of public health physician at regional
more pharmacies than that least provided (Kasserine). It should be
level and the institutional framework governing practices at the private
noted that the inequalities in the number of pharmacies nationwide
level. Similarly, public-public partnerships (such as agreements
are much lower than inequalities in private practice offices due to very
between academic physicians and regional hospitals) or, failing that,
strict regulations.
public-private partnerships likely to make disadvantaged areas more attractive, could be considered for medical specialists.
2.3
Summary The availability of private medical practitioners is characterized by high
The analysis shows a generally favourable trend that however hides
levels of inequality; the ratio between the most endowed and the least
internal distortions (Table 1 below). The availability of PHCs is declining.
endowed governorates is 14.3. They are followed by dental offices (ratio
Level II, which is the reference for Level I, would not be very effective
of 11.3) and hospital beds (10.7). The most fairly distributed resources
because it is poorly resourced. Therefore, there is an important carry
are pharmacies and paramedical staff. Like in pharmacy, the private
over to Level III, which thus replaces Level II, thereby causing
practice of dentistry and medicine should be better regulated.
inefficiencies. Obviously, there is a need to develop a strategy that strengthens and revitalizes primary health care in the country and
Lastly, the status of health care facilities in one governorate cannot be
enhances Level II.
analysed by reference to a single determinant. As such, all components of the sector and complementarity between the various providers should
Similarly, it has been shown that the trend of all dispersion indicators
be considered simultaneously. Consequently, it seems worthwhile to
is favourable, with the exception of the indicator for physicians,
conduct an analysis of health care provision that integrates all the
whether in the public or private sector. Although the number of
determinants of the provision so as to end up with relatively homogeneous
physicians per inhabitant witnessed a significant drop between 2002
groups. It would further be interesting to develop for each governorate
and 2010 (28.5% for the public health sector and 21% for private
a composite indicator of health care facilities so that a governorate could
practice), the gaps have widened between the better endowed and
gauge its position in relation to other governorates as well as the progress
less endowed governorates, and the variation coefficients have
it may achieve over time.
15 The operation of pharmacies is strictly subject to a numerus clausus, which is established on the basis of five areas according to delegations. It is regularly reviewed to adapt to the new realities of the profession and demographics.
17 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
Table 1: Level and distribution of major health care provision indicators Indicator
Inhabitants /PHC
Hospital beds/ 1 000 inhabitants
Inhabitants/ Physician (public sector) Inhabitants/ Paramedical staff Inhabitants/ Private practice (2004-2010) Inhabitants/ Dental office (2002-2009)
Inhabitants/ Pharmacy
Statistics
2002
2010
2010/2002
Average
4807
5051
+5%
CV
0.94
Max/min
Disadvantaged governorates
Appraisal Unfavourable
Tunis
0.89
Favourable
Ariana
8.63
8.66
Moderate
Ben Arous
Average
1.7
1.85
Favourable
Ben Arous
CV
0.49
0.41
Favourable
Ariana
Max/min
47.4
10.7
High
Sidi Bouzid
Average
2196
1569
Favourable
Kasserine
CV
0.32
0.35
Unfavourable
Medenine
Max/min
5.27
6.85
Moderate
Sidi Bouzid
Average
341
308
Favourable
Ariana
CV
0.44
0.3
Favourable
Ben Arous
Max/min
6.68
4.8
Low
Sidi Bouzid
Average
2128
1681
Favourable
Siliana
CV
0.5
0.6
Unfavourable
Kasserine
Max/min
11.6
14.3
High
Sidi Bouzid
Average
8847
5774
Favourable
Tozeur
CV
0.66
0.58
Favourable
Kasserine
Max/min
24
11.3
High
Tataouine
Average
6756
5604
Favourable
Tozeur
CV
0.39
0.27
Favourable
Kasserine
Max/min
4.4
2.6
Low
Tataouine
+8.8%
-28.5%
-8.7%
-21%
-34.7%
-17%
3-
Health care facilities in 2010: cluster analysis
1. Average distance to get to a regional hospital
3.1
Indicators
3. Inhabitants by PHC
2. Average distance to get to a general hospital 4. Proportion of PHCs providing medical consultation 6 days of 6 To analyse the distribution of health care facilities between the 24 governorates
5. Inhabitants per PHC full-time equivalent (FTE)
in Tunisia, we will refer to a broad set of indicators that characterize such
6. Inhabitants per primary care physician
facilities. These are indicators relating to health infrastructure, human resources
7. Frontline bio-medical laboratory unit per 100 000 inhabitants
in public and private health facilities, and equipment. These indicators are
8. Frontline radiology unit per 100 000 inhabitants
drawn from the Tunisia 2010 Health Map published by the Ministry of
9. Frontline dental chairs per 100 000 inhabitants
Health and/or from the Tunisia Statistical Yearbook published by the INS.
10. Inhabitants per day pharmacy 11. Inhabitants per night pharmacy
3-1-1.
Health infrastructure indicators
12. Private bio-medical laboratories per 100 000 inhabitants 13. Haemodialysis machines per 100 000 inhabitants (public and
These are indicators on:
private)
18 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
Economic Brief 2 0 1 4
3-1-2.
Common equipment indicators16
•
AfDB
w w w . a f d b . o r g
This cluster covers 626 PHCs, 21 district hospitals (341 beds), 13 regional hospitals (1 315 beds) 22 university hospitals (92% of the overall with 9 032
1. Hospital bed equipment rate (public and private)
beds) and 4 590 private medical practices out of 6 273 (general practitioners
2. Public hospital bed equipment rate
and specialists). It is characterized by good positioning in terms of access
3. Private bed equipment rate (in clinics)
to hospitals and backed by sustained availability of alternative types of
4. General surgery bed equipment rate
infrastructure and also by poor positioning in terms of the number of
5. Gynaecology and obstetrics bed equipment rate
inhabitants per Primary Health Centre (PHC). Hence, we have:
6. Paediatric bed equipment rate 7. Ophthalmology bed equipment rate
l
The best location in terms of access to regional hospitals and general
l
The largest proportion of PHC providing medical consultation 6 days
11. Anaesthesiology bed equipment rate
l
The highest number of inhabitants per PHC on average;
12. Psychiatric bed equipment rate
l
The highest number of inhabitants per PHC in full time equivalent;
l
A moderate number of inhabitants per primary care physician;
l
The lowest number of primary biomedical laboratory units for
1. Density of physicians (per 100 000 inhabitants)
l
The lowest number of primary radiology units for 100 000 inhabitants;
2. Density of general practitioners (per 100 000 inhabitants)
l
The lowest number of primary health care dental chairs for 100 000
inhabitants)
l
The highest number of inhabitants per day-time pharmacy;
4. Density of general practitioners in the private sector (per 100 000
l
Moderate number of inhabitants per overnight pharmacy;
inhabitants)
l
The highest number of medical analysis laboratories for 100 000
8. ENT bed equipment rate
hospitals (except Médenine due to its low density);
9. Orthopaedic bed equipment rate
a week;
10. Cardiology bed equipment rate
3-1-3.
Human resource indicators
100 000 inhabitants;
inhabitants;
3. Density of general practitioners in the public sector (per 100 000
inhabitants;
5. Density of specialists (per 100 000 inhabitants) 6. Density of public sector specialists (per 100 000 inhabitants)
The highest number of haemodialysis machines for 100 000
l
inhabitants.
7. Density of private sector specialists (per 100 000 inhabitants) 8. Medical density per specialty (all physicians) 9. Density of pharmacists (per 100 000 inhabitants)
Besides Médenine, this cluster is composed of university hospital
10. Density of dentists (per 100 000 inhabitants)
cities.18 It can be observed that there is a predominance of tertiary
11. Density of nurses, nursing aides and senior technicians (per 100 000
care, including emergency services that are particularly in demand
inhabitants).
and are overriding the PHC.19 The question then is not so much
3-2
the motives underlying the people's preference for emergency room
whether or not to increase the density of PHCs but also to understand
Heath Infrastructure Distribution
services to PHCs. Should it be blamed on the overly broad primary Based on health infrastructure indicators in the 24 governorates observed
health network or the discrepancy between its temporal accessibility
in 2010 (Tables 15, 16, 17 and 18 in Annex 1), a dynamic clusters analysis17
and the quality of care it provides?20 These two aspects certainly
was conducted in four clusters (Map 3).
deserve special consideration and it is appropriate to both standardize the availability of infrastructure and upgrade the operation of all
3-2-1- The first cluster (Table 13 in Annex 1) includes the Tunis, Ariana, Ben
structures at all levels. The certification of hospitals would be entirely
Arous, Manouba, Sousse, Monastir, Sfax and Médenine governorates.
appropriate. In this context, an agency for the accreditation and
16 We did not consider indicators for equipment that has a regional scope and serves several governorates, such as the equipment rate for public beds with university status by major region (north, centre and south); the MRI equipment rate by major region; the scanner equipment rate and the equipment rate for other heavy equipment. 17 The method for classifying dynamic clusters is essentially based on the distribution of a population into homogeneous groups (classes or clusters) using the core concept associated with each class. It may involve, as in our study, for example, discovering the main governorates with the closest health facilities.
19 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
certification of health services was established by Decree No. 2012-
an intensive care unit, with the technical equipment needed to support
1709 of 06/09/2012.
different types of emergencies, district hospitals currently perform only a single real hospital function: carrying out eutocic deliveries. Regional
3-2-2- The second cluster (Table 14 in Annex 1) includes the Bizerte,
hospitals that are supposed to provide Level II care, lack specialists (the
Nabeul and Kebili governorates. This cluster is characterized by a
cluster is very poorly staffed with specialist physicians) and equipment.
very average positioning with respect to all criteria relating to public
Furthermore, quality certification will help bring these structures up to
infrastructure (hospitals, PHC ...) or private facilities (pharmacies,
standard. Lastly, it would be appropriate to develop specific incentives
dental offices, laboratories, etc.). It includes 273 PHC, 9 CH and 6
to induce private stakeholders to settle in these governorates.
RH (2 135 beds i.e. 11% of the overall nominal capacity), 709 private practices, 286 pharmacies, 221 dental offices...
3-2-4- The fourth cluster (Table 16 in Annex 1) includes the Jendouba, Kairouan, Kasserine and Sidi Bouzid governorates. This cluster is
3-2-3- The third cluster (Table 15 in Annex 1) includes the Béja, Gabès,
characterized by even low rate of access to hospitals and available
Gafsa, Le Kef, Mahdia, Siliana, Tataouine, Tozeur and Zaghouan
basic infrastructure (with 27 CH, 470 PHC and only 4 RH). Furthermore,
governorates.
the most reduced availability of private facilities can be observed (350 private practices). The shortfall of private health services is probably
These nine governorates have a rather low hospital access rate (due
due to the low standard of living in these governorates and the lack of
to low population density) and limited availability of various types
effective demand for health services.
of infrastructure, especially those that are private-owned. This shortcoming is partly offset by proper positioning in terms of the number
Hence, this fourth cluster comprises all priority governorates in terms
of inhabitants per PHC and per primary care physician.
of infrastructure wherein an intervention to enhance public health coverage would allow coverage similar to the rest of the country, and
Yet these governorates have 46 district hospitals out of a total of 109,
seems to be a necessary step to boost private coverage through the
10 regional hospitals (about 33 throughout Tunisia) and two university
ripple effect. In this regard, it would be wise to develop specific incentives
hospitals (in Mahdia and Zaghouan), with a nominal capacity of 4 133
to induce private stakeholders to settle in these governorates.
beds (21 % of national capacity). The table below summarizes the specificities of each cluster with respect However, while a hospital should have at least one surgical ward and
to health infrastructure.
18
Tunis, Sousse, Monastir and Sfax. In a study on the reasons for recourse to the emergency services of major hospitals in Greater Tunis (Ben Gobrane et al., 2012), the major reasons given by patients are quick and easy access to emergency services, the availability of equipment as compared to PHCs and, for the populations, inappropriate working hours of primary care facilities that work only in the morning. Hence, recourse to emergency services is partly due to the shortcomings of primary care medicine. 20 In most of these structures, consultations are carried out only in the morning. In rural areas, the length of consultations is notoriously reduced given the number of consultations conducted. In urban areas, opening hours do not match the time users are available for consultation. The result is threefold: either unwarranted recourse to hospital emergency services at different levels, or delay in recourse or forced and costly recourse to private primary care facilities. 19
20 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
Economic Brief 2 0 1 4
•
AfDB
w w w . a f d b . o r g
Table 2: Distribution of health infrastructure by clusters Characteristics of Allocations Governorates
Cluster 1
Tunis, Ariana, Ben Arous, Manouba, Sousse, Monastir, Sfax and Medenine.
Cluster 2
Bizerte, Kebili et Nabeul
Cluster 3
Beja, Gabes, Gafsa, Le Kef, Mahdia, Siliana, Tataouine, Tozeur et Zaghouan
Cluster 4
Jendouba, Kairouan, Kasserine and Sidi Bouzid
Other front-line structures
Private medical structures
Hospitals
PHC
+
-
-
+
+/-
+/-
+/-
+/-
-
+
+
+/-
+/-
+/-
+/-
-
The analysis below focused on hospital availability in terms of average
located. Moreover, the quality of private technical equipment in Tunis
distance access to a hospital, but did not provide information on the
and the concentration of specialists observed there have fostered the
availability of the said hospitals. Furthermore, to better clarify this aspect,
export of health services. Standardization and certification procedures
we will consider some indicators of bed availability rates.
would allow a greater development of this sector.
3-3
Bed Availability Rate
3-3-2- The second cluster (Table 18 in Annex 1) includes the Sfax, Mahdia, Le Kef and Médenine governorates. This cluster is characterized
Based on the bed availability rate in the 24 governorates observed in 2010
by its leadership in private infrastructure. It ranks first in terms of bed
(Table 19, 20, 21 and 22 in the annex), a dynamic cluster analysis was
availability rates in clinics. However, the number of its public beds is
conducted in four clusters (Listing 2: Map 4).
relatively small (3 567 out of 19 565 beds, i.e. 18%). Overall, it ranks
3-3-1- The first cluster (Table 17 in Annex 1) includes the Tunis,
orthopaedics, ORL, ophthalmology, paediatrics and anaesthesiology.
Sousse, Monastir, Tozeur and Manouba governorates. This cluster is
However, this cluster has the highest number of beds in psychiatric
characterized by:
wards and ranks immediately after the first cluster with respect to
third in terms of hospital beds as well as specialties in general surgery,
specialized beds in gynaecology and cardiology wards.
• The best positioning in terms of bed availability rates and hospital bed availability rates (public beds). This positioning is observed for
In this cluster, Sfax is a university hospital city. Its two university hospital
all specialties except ORL and psychiatry; and
centres21, which are adjacent to each other, suffer various shortcomings.
• The second positioning in terms of clinic beds (private rooms) as
They are congested and overwhelmed by a workload exceeding their nominal capacity and thereby leading to long waiting periods. They
well as public beds in ORL and psychiatry.
service not only the Sfax governorate but also the southern population In this cluster, Tunis, Monastir and Sousse are three university hospitals
of more than 4 million inhabitants, which implies overuse of equipment.
with a total of 6 664 beds (34% of the national capacity), including 2
Because of such anomalies, these UH increasingly face difficulty in
961 by specialty. These UH are regional in scope and assigned to the
meeting academic training, specialization, high-level care and medical
North and Centre, and even national for certain specialties. They service
research needs in good condition. For over a decade, a new UH has
a much larger population than that of the governorate in which they are
been programmed for Sfax but has not yet been implemented.
21 The Habib Bourguiba Hospital (506 beds) is home to the surgical specialty services; the Hedi Chaker Hospital (889 beds), which is older, provides medical pathology services.
21 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
However, Sfax is characterized by a significant private health sector,
previous investment code provided for some tax incentives 23 for
with clinics (mono-disciplinary or polyclinic), radiology centres, medical
equipment.
laboratories and pharmacies. It serves as a medical platform for the South and for the export of health services primarily for Libyans.
3-3-4- Cluster 4 (Table 20 in Annex 1) comprises the less affluent
3-3-3- The third cluster (Table 19 in Annex 1) comprises the Béja, Gafsa,
Nabeul, Jendouba, Kairouan, Kasserine, Sidi Bouzid and Gabes. There
Kébili and Tataouine governorates. This cluster is relatively very well
have, on average, the lowest public bed availability rates and, to a lesser
equipped in terms of public beds but is poorly equipped with respect to
extent, low private bed availability rates. This is true for almost all specialties.
private beds. It ranks second with respect to public infrastructure (1 831
However, the Northern governorates (Ariana, Ben Arous, Zaghouan,
governorates in terms of beds: Ariana, Ben Arous, Zaghouan, Bizerte,
hospital beds in Levels 1 and 2). The significance of the public sector is
Bizerte and Nabeul) seem to be of lower priority due to their proximity to
highlighted by the specialized beds availability rate in ORL, surgery,
Tunis and the importance of the private sector. In the other governorates
orthopaedics, ophthalmology, paediatrics and anaesthesiology. However,
(Jendouba, Kairouan, Kasserine, Sidi Bouzid and Gabes), not only is the
there is a shortfall of equipment for gynaecology, cardiology and psychiatry.
per capita bed ratio low, but a significant proportion of the beds are district hospital beds, and district hospitals are often poorly equipped and poorly
In addition, this cluster is characterized by a relatively underdeveloped
staffed with specialists. The lack of equipment and resources in DHs has
private sector. It is the most underprivileged group in terms of
transformed these centres into intermediate facilities incapable of resolving
clinic beds. The establishment of clinics is governed by clinical
the problems they encounter, thereby causing users to bypass this level
specifications22, considered stringent as compared to that of public
and resort to the private sector or the secondary and tertiary levels, usually
structures (e.g. it is required to have a nurse for two intensive care
in other governorates. Table 3 below summarizes the distribution of beds
beds, whereas there are no set standards for the public sector). The
between the public and private facilities for each cluster.
Table 3: Bed distribution by cluster Governorates
Public Beds
Private Beds
+
+/-
Cluster 1
Tunis, Sousse, Monastir, Manouba and Tozeur
Cluster 2
Sfax, Mahdia, Le Kef and Medenine
+/-
+
Cluster 3
Beja, Gafsa, Kebili et Tataouine
+/-
-
Cluster 4
Ariana, Ben Arous, Zaghouan, Bizerte, Nabeul, Jendouba, Siliana, Kairouan, Kasserine, Sidi Bouzid and Gabes
-
+/-
22 Decree No. 93-1915 of 31 August 1993 to determine structures and specialties, and standards in terms of capacity, equipment and staffing of private health institutions, as supplemented and amended by Decree No. 99-2833 of 21 December 1999 and Decree No. 2001-1082 of 14 May 2001. 23 Decree No. 94-1056 of 9 May 1994, establishing a list of equipment needed for health and hospital institutions that may qualify for the tax incentives under Section 49 of the Investment Incentives Code and the conditions for granting these benefits, as amended and supplemented by Decree No. 98-967 of 27 April 1998 and Decree No. 2006-382 of 6 February 2006.
22 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
Economic Brief 2 0 1 4
3-4
Human Resources
•
AfDB
w w w . a f d b . o r g
Tataouine and Tozeur (Table 24 in Annex 1). In the public hospitals o f the 10 governorates, there are only 1 134 physicians (general
Based on human resource indicators from the 24 governorates, reported
practitioners and specialists), i.e. 17% of public health physicians. The
in 2010 (Tables 23, 24, 25 and 26 in Annex 1), a dynamic cluster analysis
private sector has only 601 physicians (308 general practitioners and
was conducted in four clusters (Listing 3: Map 5).
221 specialists) and relatively few dentists.
3-4-1- The first cluster (Table 21 in Annex 1) comprises the Tunis, Ariana,
3-4-3- Ben Arous, Bizerte, Gabes, Mahdia, Manouba, Médenine,
Sousse, Monastir, and Sfax Governorates. Overall, these five
Nabeul and Zaghouan constitute an intermediate group with moderate
governorates have the highest rates of human resource allocation both
human resource allocation for almost all categories. 25% of doctors
in the public sector (3957 to 6723 physicians, i.e. 59% of all physicians
and 27% of paramedical personnel in the public and private sector
in the public sector) and the private sector (47% of physicians in the
have settled in these 8 governorates (Table 22 in Annex 1).
private sector practice in these five governorates), regardless of the area of specialization. 56% of private dentists and 43% of pharmacies
3-4-4- Béja (Table 23 in Annex 1) is a governorate hinged between the
are located in the governorates of this cluster.
second cluster and the cluster of human resource-deficient governorates.
3-4-2- The cluster of human resource-deficient governorates comprises
The table below compares the distribution of human resources in both
Gafsa, Jendouba Kairouan, Kasserine, Kebili, Le Kef, Sidi Bouzid, Siliana,
sectors for the 4 clusters.
Table 4: Distribution of health workers by cluster and by sector Governorates
Public Beds
Private Beds
+
+
Cluster 1
Tunis, Ariana, Sousse, Monastir, Sfax
Cluster 2
Ben Arous, Bizerte, Gabes, Mahdia, Manouba, Medenine, Nabeul and Zaghouan
+/-
+/-
Cluster 3
Beja
+/-
+/-
Cluster 4
Gafsa, Jendouba, Kairouan, Kasserine, Kebili, Le Kef, Sidi Bouzid, Siliana, Tataouine and Tozeur
-
-
Overall, among the three types of indicators of health care provision,
of health care provision. Hence, it is important to summarize
the geographic distribution of health human resources turns out to be
infrastructure, equipment and human resource indicators into a single
the most unequal and reveals a significant concentration on the coast.
indicator.
Despite an increase in physician density, regional disparities have
4Profile of governorates and composite indicator of health care facilities by governorate.
widened. Qualitatively, the inequalities are even more egregious and more than 2/3 of the specialists are concentrated in the coast, not only for rare specialties but also for the most common ones such as gynaecology and paediatrics.
4-1
Profile of governorates
Such regional breakdown reflects the geographic dichotomy that shows
The ranking of the various governorates into homogenous groups with
a clear regional imbalance in favour of the coast to the detriment of the
respect to their medical infrastructure, their availability in terms of beds
North West and Central West of Tunisia. It is important to consider the
and human resources, help to develop a profile for each governorate
impact of this inequality in human resource allocation on the inequality
according to the groups to which it belongs (see Table below).
23 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
Table 5: Assignment of governorates to clusters (Rji) Human Resources
Governorates
Infrastructure
Equipements
MONASTIR
1
1
1
SOUSSE
1
1
1
TUNIS
1
1
1
SFAX
1
1
2
MANOUBA
2
1
1
ARIANA
1
1
4
MEDENINE
2
1
2
BEN AROUS
2
1
4
TOZEUR
4
3
1
MAHDIA
2
3
2
BIZERTE
2
2
4
NABEUL
2
2
4
GABES
2
3
4
KEBILI
4
2
3
LE KEF
4
3
2
ZAGHOUAN
2
3
4
BEJA
3
3
3
GAFSA
4
3
3
TATAOUINE
4
3
3
SILIANA
4
3
4
JENDOUBA
4
4
4
KAIROUAN
4
4
4
KASSERINE
4
4
4
SIDI BOUZID
4
4
4
24 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
Economic Brief 2 0 1 4
•
AfDB
w w w . a f d b . o r g
The table shows that three governorates still belong to the most
This indicator is the average of the three scores awarded to each
advantaged cluster irrespective of the criterion applied. They are Tunis,
governorate.
Sousse and Monastir. By contrast, four governorates still belong to the most disadvantaged cluster. They are Jendouba, Kairouan, Kasserine and Sidi Bouzid (Map 6).
If rji is the ranking of the cluster to which governorate j belongs with respect to resource i, score sji may be defined as the inverse of rji
Between these two groups, the other governorates have more or less
sji = 1/rji given that rji = 1, 2, 3, 4
significant shortfalls depending on the type of resources being analysed. The table defines the scope of intervention required by each governorate.
Thus, the score of governorate j for resource i (sji) is considered a unit
Hence, Sfax, for example, requires greater bed availability rates in order
score where the governorate belongs to the higher class. The score
to measure up to Tunis, Sousse and Monastir.
decreases as the governorate moves away from this class.
4-2
The composite indicator of a governorate’s health care provision is equal
Composite indicator of health care provision
to the arithmetic average of its scores. Ij = ∑ (sji)/3 Each governorate, depending on its profile, may be assigned a composite indicator of health care provision.
The outcomes are presented in the table below.
Table 4: Score by resource and composite indicator of health care facilities by governorate. Governorates
Human Resources
Infrastructure
Equipements
Total
MONASTIR
1.00
1.00
1.00
1.00
SOUSSE
1.00
1.00
1.00
1.00
TUNIS
1.00
1.00
1.00
1.00
SFAX
1.00
1.00
0.50
0.83
MANOUBA
0.50
1.00
1.00
0.83
ARIANA
1.00
1.00
0.25
0.75
MEDENINE
0.50
1.00
0.50
0.67
BEN AROUS
0.50
1.00
0.25
0.58
TOZEUR
0.25
0.33
1.00
0.53
MAHDIA
0.50
0.33
0.50
0.44
BIZERTE
0.50
0.50
0.25
0.42
NABEUL
0.50
0.50
0.25
0.42
GABES
0.50
0.33
0.25
0.36
KEBILI
0.25
0.50
0.33
0.36
LE KEF
0.25
0.33
0.50
0.36
ZAGHOUAN
0.50
0.33
0.25
0.36
BEJA
0.33
0.33
0.33
0.33
GAFSA
0.25
0.33
0.33
0.31
TATAOUINE
0.25
0.33
0.33
0.31
SILIANA
0.25
0.33
0.25
0.28
JENDOUBA
0.25
0.25
0.25
0.25
KAIROUAN
0.25
0.25
0.25
0.25
KASSERINE
0.25
0.25
0.25
0.25
SIDI BOUZID
0.25
0.25
0.25
0.25
25 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
The analysis is based primarily on quantitative indicators. However,
while it is true that the choice of the location of pharmacies, given
qualitative aspects are still essential. The same is true for those relating
that it is highly regulated, resulted in a fairly equal space coverage by
to the efficiency of structures. It is obvious that it is equally important
pharmacies, it is worth noting also that such was not the case for
to improve the functioning of the existing infrastructure to make the
dental practices and especially for private medical practices which,
most out of them, as it is to further increase the density of infrastructure
instead of correcting the deficiencies resulting from the lack of public
(World Bank, 2008).
health services, exacerbate existing inequalities. It is then clear that in the absence of a regulatory framework for the opening of private
To this end, a certification of existing structures should be considered
medical practices, it is important to introduce special incentives for
(at least the most important). The missions and responsibilities of each
young doctors to settle in priority governorates, despite the absence
level of care should be laid down, and resources should be allocated
of an adequately solvent demand. It may include conventional
accordingly. Establishment projects should be stopped for each
cooperation between regional hospitals and specialists in private
structure.
practice for the missing specialties.
The analysis revealed that, despite the reduction in public budgets
However, the choices relating to the health sector and the efforts to
allocated to health, there is reduced inequality in the allocation of
better allocate resources to priority areas can be effective only if they
infrastructure to the various regions. This decrease in inequality is partly
form part of a comprehensive local development strategy for these
due to the concentration of government efforts on the most resource-
areas. The reduction of economic, cultural and social gaps between
deprived areas. However, it stems mostly from a reduction of resources
the governorates can only facilitate and strengthen the health reforms.
allocated to the major urban centres with high population growth dynamics. This movement is synonymous with a decline in access to
To better support our recommendations, we would have liked to test a
health care for vulnerable populations in these regions. It is also
panel model which explains the population’s health status by governorate
symptomatic of reduced resources allocated to structures responsible
(life expectancy and infant mortality rate) and by the prevailing health
for the training of future doctors and paramedical personnel.
status in each governorate. We were not able to do so because health indicators are not published at the governorate level. Such work would
The dwindling resources allocated to the public sector and the ensuing
be very informative and could be conducted later.
reduced access to care would be offset by an increase in private sector resources and a greater availability of its services. In fact, there
Reduction of health inequalities may be achieved only by understanding
is an interesting private sector dynamics in its three components:
the determinants, setting the corresponding objective, informing health
private medical practices, dental offices and pharmacies. However,
care professionals and monitoring the achievements.
26 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
Economic Brief 2 0 1 4
•
AfDB
w w w . a f d b . o r g
4. Trend in inequalities in health spending in Tunisia between 2000 and 2010
n recent years, the State’s effort in favour of health has been on the
The analysis of the contribution of the various out-of-pocket health
decline. Public health expenditures accounted for 2.7% of GDP in
expenditure items to the total health expenditure inequality and the
1995 and for only 2.3% in 2011. By contrast, private spending by
trends of this contribution between 2000 and 2010 should help to
I
households rose sharply. In 2010, household spending amounted to
better understand some of the causes of the inequality in out-of-
51% of total health expenditures, with approximately 80% being out-
pocket health spending and to possibly identify policies that should
of-pocket payments and 20% corresponding to health insurance
be implemented to reduce this inequality. In this regard, we will use
premiums. In fact, out-of-pocket payments made by households for
the breakdown of the inequality index by source25 (or component).
health purposes account for 41.2% of national health expenditures
This technique indicates which expenditure categories contributed
(Arfa et al., 2013).
most to the formation of inequality. Any policy aimed at reducing health care inequalities should primarily target these categories (Wagstaff and Van Doorslaer, 2000).
The importance of out-of-pocket health expenditure in national health accounts (Arfa et al., 2007 and 2008) implies that the search for greater equity in health should entail efforts to achieve more equitable
Traditionally, inequalities in expenditures are captured through the
distribution of the out-of-pocket spending of households. Therefore,
Lorenz curve and the Gini index that analyse the distribution of health
it is important to evaluate the inequality of out-of-pocket health
expenditures in themselves. However, despite its merits, this approach
expenditures of households and to analyse their trends using inequality
presents only a partial picture of the inequalities in health expenditures.
indicators.
A more comprehensive picture may be obtained through the concentration curve and index, which analyse the distribution of health
To that end, we will refer to data from the national surveys on budget
expenditures in relation to the distribution of the population’s living
and household consumption of 2000, 2005 and 2010. These surveys
standards ranked from the poorest to the richest.
were conducted on a representative sample of households across the country.24 They provide information on individual consumption of
We will start off by focusing on the Gini index, considered as an indicator
goods and services, and therefore make it possible to study the trends
of inequality in health spending, to see how its breakdown by source
in the living standards of households through their expenditures.
can provide information on the formation and trend of inequality. Next,
These data provide information on expenditures per person per year
we will follow the same approach for the concentration index.26 Lastly,
(SPY) and their structure by various expenditure items, especially
we will analyse the inequality in health spending and its trend in light of
those related to health. The data also facilitates the assessment of
the available data.
the degree of inequality of these expenditures and the analysis of their trends. Various health spending items were covered in the
For estimates of the Gini and concentration indices, we use the STATA
consumer surveys (see nomenclature in Annex 1: Health indicators
12 and DASP27 version 2.2 software, which enable us to calculate and
by cluster 2). They correspond to disbursements made by households.
break down the various indices.
24
For the three surveys, the initial sample is drawn from a stratified random sampling conducted in two stages in each governorate. The sample frame consisted of the data files of the general population census (1994 and 2004 respectively). Regarding the 2000 survey, 12 960 households were sampled and 12 249 responded (representing a response rate of 95%). Concerning the 2005 survey, 13 392 households were sampled and 12 317 responded (that is, a response rate of 92%). As for the 2010 survey, of the 13 392 households initially sampled, 11 291 responded (representing a response rate of 85%). 25 The breakdown of inequality indices was introduced in health economics by Wagstaff et al. (2003). 26 The literature on inequalities shows that there is a range of relevant inequality indicators, including the Theil indicator and the log deviation. In addition, other breakdown techniques, such as groups, are interesting and very informative. In subsequent research, we plan to break down the inequalities in health expenditures using various geographic (environment, region, governorate) and socioeconomic (occupational status of the household head, household size, vulnerability. etc.) criteria, and to identify the relative importance of intra-group and inter-group inequalities and their trend. 27 Distributive Analysis Stata Package, developed by Araar and Duclos (2007). PEP, World Bank, UNDP and Laval University. http://dasp.ecn.ulaval.ca/
27 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
1-
•
Economic Brief w w w . a f d b . o r g
Lorenz curve and the Gini index
1.3.
Breakdown of the Gini index by expenditure categories
To describe the distribution of health expenditures as well as the
Analytically, the Gini index has several expressions. Lerman and Yitshaki
degree and origins of their inequality, we start by adopting the Lorenz
(1985) showed that:
classification method and then the method that consists in calculating acceptable inequality indicators.
G = 2 cov (Y ; F(Y)) / Y
1.1.
Y being the expenditure vector Y : (Y1, Y2, .... Yn) and F(Y) representing
Lorenz Curve
(Equation 1)
the cumulative distribution considered as random variable uniformly distributed Inequality in the distribution of a variable (health expenditure, for
between [0, 1].
example) may be highlighted using the Lorenz curve (Figure 31) which matches the proportion of the population classified by increasing order
The breakdown adopted makes it possible to analyse the contributions
of health expenditures with the share of total health expenditures
to total inequality of the various expenditure categories (consultations,
incurred by that population.
radiological procedures, drug purchases, etc.). It also facilitates the measurement of their specific inequality contribution to the total inequality (Lerman and Yitshaki 1985).
Figure 31: Lorenz Curve for health spending
Let Y= (Y1, Y2, .... Yn) be the distribution of total health expenditures and Yik the expenditure of the person i in item or category k where k=1....K. Y1,Y2,...,Yk are therefore the expenditure components and Y = ∑ Yk (Equation 2) Given the properties of the covariance, we have: G = 2 ∑cov (Yk ; F(Y)) / Y
(Equation 3)
where cov (Yk ; F(Y)) represents the covariance of the expenditure in item k with the cumulative distribution of total expenditure.
In the case of two distributions X and Y, X dominates Y if the Lorenz curve relating to distribution “X” constantly lies above the curve relating to distribution “Y”. Distribution “X” is then more egalitarian than “Y”.
When cov (Yk , F(Yk)) Yk, is multiplied and divided by cov(Yk ;F(Yk ))
Nevertheless, when two curves intersect, comparison of the inequality
and by Yk, we obtain the rule of breakdown according to the source
trend becomes impossible. Therefore, the digital indicators for evaluating
(or component), that is:
inequality must be calculated as the Gini coefficient. k] * (Yk /Y) ] G = ∑ [ [cov (Yk ; F(Y)) / cov (Yk ; F(Yk))] * [2cov (Yk ; F(Yk)/ Y
1.2.
GINI coefficient
(Equation 4) Let’s note:
The Gini coefficient seems to be the most popular of the various inequality indices. It is derived from the Lorenz curve (concentration curve) in that it is the ratio of the area between this curve and the first
- the Gini correlation between the component k and the total
diagonal line and the half-square in which the curve lies. The Gini
expenditure:
index lies within the range [0, 1]. The more it tends towards 1, the Rk = cov (Yk ; F(Y)) / cov (Yk ; F(Yk))
more unequal is the distribution of expenditures. On the other hand,
(Equation 5)
when the indicator declines, the distribution of expenditure becomes - the Gini coefficient related to the component k: G k = 2cov
more egalitarian.
28 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
Economic Brief 2 0 1 4
(Yk ; F(Yk) / Yk
(Equation 6)
•
AfDB
w w w . a f d b . o r g
living, ranked from the poorest to the richest (horizontal axis). In other words, it links the share of health expenditures to the quintiles of total expenditures (or other variable indicating the standard of living).28
- the proportion of the component k in the total expenditure : k /Y Sk = Y
(Equation 7)
Then G = ∑ Rk Gk Sk
(Equation 8)
Figure 32: Concentration curve for health spending in 2010
Cumulative distribution of Heath expenditure
The relative contribution of an expenditure source k to the total inequality is: Pk = Rk Gk Sk / G.
(Equation 9)
The sum of the relative contributions of various sources (components) is equal to the unit. The breakdown of the Gini index also helps to determine the marginal effect of variation in each expenditure category k on total expenditure inequality. Let ek be a scalar slightly superior
Population in order of increasing Heath expenditure
to the unit, an increase in expenditures derived from the source k results in the passage to vector ekYk and will involve a variation of
2.2.
G. The variation in the value of G brought about by this change at
Concentration index
the margin of the expenditure category is obtained from the partial The concentration index (Kakwani, 1980), which derives directly from
derivative of G in relation to ek. We show that:
the concentration curve, quantifies the degree of socioeconomic ∂G/ ∂ek = Sk (Rk Gk - G)
(Equation 10)
inequality related to the variable being examined (health expenditure). The concentration index is equal to twice the area between the
∂G/ ∂ek ek is the marginal contribution of source k to total inequality.
concentration curve and the first bisector. If there is no socioeconomic
The relative marginal effect is obtained by dividing the above expression by
and the concentration index is equal to zero. By convention, the
G, that is to say:
concentration index is negative when the concentration curve lies
inequality, the concentration curve is confounded with the bisector
above the first bisector. In this case, health expenditures are highly (∂G/ ∂ek)/ G = ( Rk Gk Sk / G ) - Sk
(Equation 11)
concentrated among the poor. Conversely, the concentration index is positive when the concentration curve lies below the first bisector.
It is clear that the sum of the relative marginal effects is nil, multiplying
Consequently, it is important not to focus on the inequality of health
all the sources of income by e leaves Gini’s global index unchanged.
expenditures as reflected by the Gini index for these expenditures,
2.
Concentration curve and index
the concentration index.
2.1.
Concentration Curve
2.3. Breakdown of the concentration index by expenditure categories
but instead on the inequality of health expenditures as revealed by
The concentration curve (Figure 32) represents the cumulative percentage of health expenditures (vertical axis) associated with the
According to Kakwani (1980), the expenditure concentration index
percentage of the population classified by increasing standards of
X is:
28
The Gini curve appears as a special case of the concentration curve. For the Lorenz curve, the vertical axis and the horizontal axis refer to the same variable. For the concentration curve, the variables are different.
29 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
C = 2 cov (X ; F(Y)) / Y
3.
(Equation 12)
Kakwani index
X being the health spending vector and Y the vector of total
The comparison of the health expenditures concentration curve with
expenditure: (Y1, Y2, .... Yn) and F (Y) representing the cumulated distribution considered as a random variable evenly distributed
the Gini curve of total expenditures gives an indication of the progressivity of the variables being studied, which, in our case, are health expenditures.
between [0 and 1].
If health expenditures are proportional to the standard of living, then the
The Gini index appears as a special case of the concentration index
and the concentration curve of health expenditures is confounded with
where X=Y.
that of living standards (or Lorenz curve). When the poor spend
inequality in health expenditures is similar to that of living standards,
proportionately less on their health, their share of health spending is That is, X1,X2,...,Xk the expenditure components with X = ∑ Xk
(Equation 13)
lower than their share in the total expenditures. In this case, the Lorenz curve dominates (or is above) the concentration curve. The opposite
The concentration index C may be written as follows (O’Donnell et al.,
phenomenon is observed when the system is regressive.
2008): The Kakwani index is equal to twice the area between the Gini curve C = ∑ Sk Ck
(Equation 14)
and the concentration curve. It is equal to the difference between the health spending concentration index and the Gini index of total
Where Sk = Xk /X represents the specific budget coefficient relating to expenditure item k or the expenditure elasticity Xk in relation to total
expenditures.
expenditure X.
I=C–G
Ck is the concentration index specific to the category k
The value of I lies between -2 and 1. A negative value means that health spending is regressive; the Gini curve is below the concentration
The contribution of category k to total inequality is:
curve. A positive value implies the progressivity of health spending; the Gini curve is above the concentration curve. There is uncertainty
Pk = SkCk/C
(Equation 15)
when the curves intersect. In which case, it becomes necessary, in addition to the graphical analysis, to use the Kakwani index to
This leads to: ∑ Sk Ck /C = 1
distinguish between these cases.
Figure 33: Concentration curves for the health SPY and total SPY in 2010
The sum of the relative contributions of various sources (components) is equal to the unit. The breakdown of the concentration index also helps to determine the contribution of variation in each expenditure category k to total expenditure inequality. Thus, the breakdown of the Gini index as proposed by Lerman and Yitshaki (1985) and that of the concentration index as presented by Kakwani (2000) makes it possible to measure the contribution of an expenditure category to the total inequality. The breakdown also helps to gauge the impact of a marginal increase in a particular expenditure category on total inequality. We will use these breakdown techniques to analyse how the various items of health expenditure have contributed to the inequality in private expenditures on health in Tunisia.
30 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
Economic Brief 2 0 1 4
4. Trend in and sources of inequality of health SPY: 2000-2005-2010 4.1.
•
AfDB
w w w . a f d b . o r g
Table 8: Trend in the concentration index and Kakwani index
Overall inequality trends
Gini indices show that inequalities in health expenditures are much
Years
2000
2005
2010
C
0.444437
0.448276
0.410342
G
0.408558
0.414008
0.384674
I
0.035879
0.034268
0.025668
greater than inequalities in total household expenditures (Table 7). In 2010, the Gini curve of health spending revealed a high concentration of health expenditures (Figure 31). Twenty per cent (20%) of the population accounts for over 75% of health spending. The high levels of inequality stem from unequal health status of the population, given that health expenditures are closely linked to the need for care, which is expressed only by those in poor health.
The concentration index shows that inequality in health spending increased between 2000 and 2005, but dropped sharply between
Table 7: Trend in Gini index for total SPY and the health SPY29
2005 and 2010 (Table 5). Despite this decline, inequality in health spending continues to be higher than inequality in total expenditures
Years
2000
2005
2010
(Figure 33).
Total SPY
0.408558
0.414008
0.384674
The Kakwani index is positive. Health expenditures are progressive
Health SPY santé
0.744939
0.715508
0.711574
(Figure 34). The Kakwani index declined between 2000 and 2010, the progressivity of health spending was constrained.
Therefore, it is important to consider not the inequality in health
The scope of the inequalities in health care spending explains why
spending as reflected by the Gini index for these expenses, but the
they should be paid close attention and the need to identify ways and
inequality in health spending as shown by the concentration index.
means of reducing these inequalities.
Figure 34: Budget coefficient of health spending by decile of total expenditures in 2010
7% 6% 5% 4% 3% 2% 1% 0% d1
d2
d3
d4
d5
d6
d7
d8
d9
29
d10
The Gini index for total SPY 2000 is equal to that published in the report of the 2000 consumption survey (p.27). The 2000 and 2005 indices differ from those published in 2012 in the poverty report (p. 23). However, the 2010 index does not differ from the latter.
31 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
4.2. Contribution of expenditure items to total inequality in health spending
special radiology services (scans and MRI), child delivery, medical treatment abroad, long-term care (long-term consultations and medication) and cosmetic surgery (since 2010);
In the national budget and household consumption surveys of 2000, 2005 and 2010, health expenditures are classified into four categories:
•
Pharmaceuticals (drugs) and such other pharmaceuticals as baby products (talc, soap, etc.), adhesive bandages, etc.;
•
Routine medical care which includes medical consultations, dental care, paramedical services (x-rays, analyses and nursing
•
Medical devices, including optical glasses, blood pressure
services) in both public and private institutions. This item also
measuring devices, hearing aids, etc.
includes the use of traditional medicine (healers and medicinal plants);
Therefore, one may wonder to what extent each of these items
Special medical care which includes stays in the hospital
question, we will proceed to break down the health inequalities by
or clinic, surgical procedures, special dental surgery procedures,
category (Table 9).
contributed to changes in health spending inequalities. To answer this •
32 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
Economic Brief 2 0 1 4
•
AfDB
w w w . a f d b . o r g
Table 9: Breakdown of health spending inequality by expenditure category
Year
2000
2005
2010
Item
Yk in D
Gini Index
Sk
Concentration Index
Kakwani Index
Gk
RkSkGk/G
Marginal effect
Gk
SkCk/C
Ck-G
Routine medical care
15.471
21.8%
0.877%
22.6%
0.008
0.471
23.14%
0.062442
Special medical care
27.974
39.4%
0.872
40.6%
0.012
0.456
40.78%
0.050442
Pharmaceuticals
26.351
37.2%
0.799
35.3%
-0.019
0.408
34.10%
-0.000558
Medical equipment
1.130
1.6%
0.984
1v5%
-0.001
0.550
1.97%
0.141442
Total
70.927
100%
0.745
100.0%
0.0000
0.444
100.0%
0.035442
Routine medical care
30.287
27.1%
0.846
28.6%
0.015
0.486
29.45%
0.071992
Special medical care
32.054
28.7%
0.888
30.4%
0.017
0.453
29.01%
0.038992
Pharmaceuticals
47.510
42.6%
0.740
39.5%
-0.031
0.417
39.57%
0.002992
Medical equipment
1.742
1.6%
0.977
1.5%
-0.001
0.565
1.97%
0.150992
Total
111.672
100%
0.716
100.0%
0.0000
0.448
100.0%
0.033992
Routine medical care
39.062
27.1%
0.843
28.6%
0.015
0.433
26.04%
0.048326
Special medical care
42.067
29.2%
0.903
32.1%
0.029
0.493
39.99%
0.108326
Pharmaceuticals
60.790
42.2%
0.724
37.8%
-0.044
0.331
31.91%
-0.053674
Medical equipment
2.210
1.5%
0.986
1.5%
-0.000
0.625
2.06%
0.240326
Total
144.251
100%
0.730
100.0%
0.0000
0.410
100.0%
0.025326
33 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
Inequality of the SPY for routine medical care (G = 0.843 and C =
were progressive as evidenced by the Kakwani index and budgetary
0.433 in 2010) was higher than the overall inequality in health
coefficients by decile of total SPY (Figure 35).
expenditures (G = 0.730 and C = 0.410 in 2010). However, on account of their average share in these expenditures (the Sk stood at
Of all health expenditures, pharmaceutical expenditures were the most
27.1% in 2010), their relative contribution to inequality was average
substantial and the least unequal component (Table 6). In 2010, they
(28.6% in 2010). From 2000 to 2010, the relative marginal effect of
accounted for 42.2% of total health expenditures, while their contribution
spending on routine medical care was positive throughout the period,
to the formation of inequality stood at 37.8%, according to the Gini index,
meaning that this source of inequality instead had an inertia effect on
and at 42.2%, according to the concentration index. From 2000 to 2010,
inequalities in health spending. The Kakwani index was positive and
the Gini index specific to pharmaceutical expenditures declined from
routine health care expenditures were progressive. Indeed, in 2010,
0.799 to 0.724.30 The marginal effect of the growth of these expenditures
the budget coefficient of these expenses decreased for the 9th and
on inequality was negative throughout the period analysed, reflecting an equalizing impact. In 2010, pharmaceutical expenditures were regressive
10 deciles (Figure 35).
from the fifth decile (Figure 35) and the Kakwani index was negative. The inequality of the SPY for special medical care was on the uptrend
Expenditures on medical equipment were the smallest component
(GK2000 = 0.872; Gk2010 = 0.903 and = 0.459 CK2000; Ck2010 =
of health expenditures (Table 6), yet they were the most unequal
0.493). Throughout the period, their relative marginal effect remained
(Gk201=0.986 and Ck2010=0.625). In 2010, they accounted for
positive. This development shows that spending on special medical
1.5%, while their contribution to the formation of inequality stood at
care is a source of inequality and its increase helped to worsen the
1.5%, according to the Gini index. These expenditures were highly
inequality in overall care expenditures, given that these expenditures
progressive and had the highest Kakwani index.
Figure 35: Budget coefficients of health spending by decile of total SPY, by category, in 201031
0.03 0.025 0.02 0.015 0.01 0.005 0
Rou ne medical care Special medical care Pharmaceu cals
d2
30 31
d3
d4
d5
d6
d7
d8
Medical equipement
d9 d10
The concentration index went from 0.408 to 0.331 Our calculations from the pooled budget data on the 2010 survey published on site of the National Institute of Statistics (INS) http://www.ins.nat.tn/indexfr.php
34 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
Economic Brief 2 0 1 4
4.3. Contribution of routine medical care to the total inequality
•
AfDB
w w w . a f d b . o r g
(G2010 = 0.993 and C2010 = 0.600). This situation reflects the difficulty that poor people have in accessing health care other than the one provided free of charge. The often exorbitant cost of these procedures and care
Routine care expenses include medical consultations, dental care,
would lead to high elasticity and discourage low-income patients from
radiology and biological procedures in both public and private sector
using the services.32 In actual fact, the expenses entailed amount to a
facilities, and the use of traditional medicine (Table 7).
high proportion of their income. The budget coefficient for these
Between 2000 and 2010, the decline in inequalities in routine medical
index is very high, compared to other items of expenditure. There is
care expenses was driven by the decline in inequalities in medical
therefore considerable elbow room for improving the access of poor
consultations (Gk2000 = 0.877; Gk2010 = 0.849 and Gk2010 = 0.443; Ck2010
people to such care.
expenses increases exponentially with income (Figure 36). The Kakwani
= 0.433). In 2010, the specific budget coefficient of these expenditures decreased for the 9th and 10th deciles (Figure 36). The Kakwani index
The trends are similar to those recounted in the literature.
was positive but very small.
Socioeconomic status is identified as a key determinant of health expenditure. Similarly, people of higher socio-economic status generally
The decline in inequalities in spending on medical consultations may be
use more health care services, especially the specialized services, than
attributed to the greater availability of physicians (reflected in the
other people (Allin et al. 2006). In contrast, when people of lower socio-
reduction in the ratio of the number of inhabitants per physicians). The
economic status are not entitled to free care, they are more likely, than
decline is observable despite an increase in inequality in the availability
other people, to forgo health expenditures.
of physicians nationwide. This suggests that when patients do not have easy access to medical services and physicians, they tend to move
For De Looper and Lafortune (2009), despite having higher rates of
towards the latter.
illness, disease and death, poorer or less educated persons often have
Spending on radiological and biological procedures continued to be
services. They make less use of these goods and services, some of
highly unequal (G2010 = 0.947 and C2010 = 0.464). The same is true for
which are very expensive compared to their income. We would expect
dental care, which is generally not considered a priority by the poor
the same to be true for special medical expenses.
difficulties in accessing appropriate specialists and preventive health
Figure 36: Budget coefficients of routine health spending by deciles of total SPY, and by sub-item, in 201033
0.015 Medical consulta ons
0.01
Dental
0.005
X-Rays, Scans and Medical analyses
0 d1 d2 d3 d4 d5 d6 d7 d8 d9 d10
32 33
For these types of care, there is balance and the demand nil. Our calculations from the pooled budget data on the 2010 survey published on site of the National Institute of Statistics (INS) http://www.ins.nat.tn/indexfr.php
35 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
Table 10: Breakdown of inequality in routine health spending by expenditure sub-category
Year
2000
2005
2010
Item
Yk in D
Gini Index
Sk
Concentration Index
Kakwani Index
Gk
RkSkGk/G
Marginal Effect
Gk
SkCk/C
Ck-G
Medical consultations
11.571
16.3%
0.877
16.3%
0.0000
0.443
16.26%
0.03444
Dental care
0.806
1.14%
0.996
1.3%
0.0016
0.607
1.55%
0.19844
X-rays, scans and medical analyses
3.037
4.28%
0.98
4.9%
-0.0062
0.544
5.24%
0.13544
Traditional medicine
0.056
0.08%
0.99
0.08%
-0.0000
0.525
0.09%
0.11644
ROUNTINE MEDICAL CARE
15.471
21.8%
0.877
22.6%
0.008
0.471
23.14%
0.06244
Medical consultations
21.177
19.0%
0.849
19.5%
0.005
0.600
19.66%
0.18599
Dental care
1.506
1.3%
0.995
1.6%
0.003
0.464
2.02%
0.04999
X-rays, scans and medical analyses
7.515
6.7%
0.954
7.4%
0.007
0.822
7.66%
0.40799
Traditional medicine
0.089
0.08%
0.999
0.09%
0.000
0.578
0.10%
0.16399
ROUNTINE MEDICAL CARE
30.287
27.1%
0.846
28.6%
0.015
0.486
29.45%
0.07199
Medical consultations
25.749
17.9%
0.849
18.2%
0.003
0.406
16.08%
0.02133
Dental care
1.640
1.1%
0.993
1.2%
0.001
0.600
1.65%
0.21533
X-rays, scans and medical analyses
11.515
8.0%
0.947
8.9%
0.009
0.464
8.23%
0.07933
Traditional medicine
0.157
0.1%
0.999
0.1%
0.000
0.822
0.07%
0.43733
ROUNTINE MEDICAL CARE
39.062
27.1%
0.843
28.6%
0.015
0.433
26.04%
0.04833
36 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
Economic Brief 2 0 1 4
4.4.
Contribution of special medical expenditures
•
AfDB
w w w . a f d b . o r g
Increasing inequalities in special health care services are observed in all expenditure items with the exception of special expenditures for
Les dépenses de soins exceptionnels regroupent les maladies de
radiological services (MRI and scans). Whenever radiological services
longue durée, les dépenses de séjour et chirurgie médicale, les soins
are called for, the special expenditures for the services are most often
dentaires exceptionnels, les dépenses exceptionnelles de radiologie et
waived in hospitals or covered by health insurance, and the patients
les accouchements (Tableau 11).
themselves seldom have to make any payments.
Table 11: Breakdown of inequality in special care spending by expenditure sub-category
Year
2000
2005
Item
Yk in D
Concentration index
Gini index
Sk
Kakwani index
Gk
RkSkGk/G
Marginal effects
Gk
SkCk/C
Ck-G
Hospital stay and medical surgery
8.703
12.3%
0.94
13.5%
0.0123
0.509
14.05%
0.1004
Special dental care
0.710
1.0%
0.99
11.7%
0.0017
0.768
1.73%
0.3594
Special radiology expenses
0.671
0.9%
0.99
1%
0.0008
0.503
1.07%
0.0944
Child Delivery
1.110
1.6%
0.98
1.4%
-0.0017
0.341
1.20%
-0.0676
Treatment abroad
0.215
0.3%
0.99
0.4%
0.0008
0.769
0.53%
0.3604
Long-term illnesses
16.565
23.4%
0.90
23.1%
-0.0024
0.423
22.20%
0.0144
SPECIAL MEDICAL CARE
27.974
39.4%
0.872
40.6%
0.012
0.459
40.78%
0.0504
Hospital stay and medical surgery
11.244
10.1%
0.96
11.3%
0.0125
0.470
10.56%
0.0560
Special dental care
0.837
0.7%
0.99
0.7%
0.00015
0.575
0.96%
0.1610
Special radiology expenses
1.474
1.3%
0.98
1.4%
0.0005
0.430
1.27%
0.0160
Child Delivery
1.793
1.6%
0.98
1.7%
0.00045
0.448
1.61%
0.0340
Treatment abroad
0.034
0.03%
0.99
0.04%
0.0001
0.750
0.05%
0.3360
Long-term illnesses
16.673
14.9%
0.93
15.2%
0.0034
0.437
14.57%
0.0230
SPECIAL MEDICAL CARE
32.054
28.7%
0.888
30.4%
0.017
0.453
29.01%
0.0390
Hospital stay and medical surgery
12.472
8.6%
0.977
10.3%
0.0174
0.515
12.42%
0.1303
Special dental care
1.156
0.8%
0.996
0.8%
0.0006
0.665
1.42%
0.2803
Special radiology expenses
2.101
1.5%
0.981
1.5%
0.0004
0.396
1.55%
0.0113
Child Delivery
1.870
1.3%
0.987
1.3%
0.0004
0.444
0.98%
0.0593
Treatment abroad
0.000
0.0%
0.999
0.0%
-0.000
0.392
0.00%
0.0073
Long-term illnesses
24.464
17.0%
0.933
18.0%
-0.0101
0.485
23.61%
0.1003
SPECIAL MEDICAL CARE
0.009
0.0%
0.999
0.0%
0.000
0.411
12.42%
0.0263
SOINS MEDICAUX EXCEPTIONNELS
42.067
29.2%
0.903
32.1%
0.029
0.493
39.99%
0.1083
2010
37 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
Long-term illnesses had the most substantial marginal effect, accounting
coefficients were strongly on the uptrend, reflecting the magnitude of the
for 17% of total health expenditures in 2010, including expenses
inequalities (Figure 37). The highest Kakwani index observed was for
for hospital stay and medical surgery. The corresponding budget
special dental care and treatment abroad in 2000 and 2005.
Figure 37: Budget coefficients of special health spending by deciles of total SPY and by sub-item, in 201034
0.014 0.012 0.01 0.008 0.006 0.004 0.002 0
Hospital Stay and Surgery Special Dental Care Special Radiology Expenditures Child Delivery Long-term Illnesses
d1 d2 d3 d4 d5 d6 d7 d8 d9 d10
This trend reflects the demographic and epidemiological transitions35
inequality in pharmaceutical expenditures and the weight of these
experienced by the country. The health indicators in Tunisia show
expenditures on the poorest populations, although they are entitled to
an overall decline in the incidence of infectious diseases and a
free or virtually free care and medicines.37
simultaneous increase in the incidence of non-communicable diseases The Kakwani indices were negative or extremely low.
and chronic diseases among aging populations. Analysis of mortality data36 shows that the main causes of death are diseases of the circulatory system, metabolic diseases and cancer. These diseases may
The decline in inequality in pharmaceutical expenditures was due to the
be avoided through primary (healthier lifestyles) or secondary prevention
decrease in inequality relating to these two types of expenditures. Their
(screening, early diagnosis). Their prevention and, above all, treatment
contribution to inequality diminished and their marginal effect was
call for increased spending by households and the community.
negative, pointing to an equalizing effect.
4.5.
Contribution of pharmaceutical expenditures
The decline in inequality of pharmaceutical expenditures may be attributed to the increasingly equal availability of pharmacies nationwide.
The pharmaceutical expenditures include expenses relating to purchase
Faced with the difficulty of accessing physicians, patients - especially
of drugs and those relating to the purchase of pharmaceutical products
the poorer ones and those of the middle class - often resign themselves
(Table 9). In 2010, the corresponding budget coefficients were on the
to self-medication. This is also true for patients who live in areas that are
downtrend depending on the deciles (Figure 38), reflecting the low
under-served in terms of health infrastructure and doctors.
34
Our calculations from the pooled budget data on the 2010 survey published on site of the National Institute of Statistics (INS) http://www.ins.nat.tn/indexfr.php The infant mortality rate dropped from 51.4 ‰ in 1984 to 16 ‰ in 2011. Life expectancy at birth was 74.9 years in 2011 (source: National Institute of Statistics - INS). 36 National Statistics on Medical Causes of Death. Tunis 2009). Research Unit on Aging and Medical Causes of Death – National Public Health Institute. 37 It would appear patients entitled to free drugs are sometimes forced to buy their drugs from private pharmacies due to shortages in hospital pharmacies or Primary Health Centres (PHCs). These shortages are caused by governance problems and would be resolved. 35
38 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB
Economic Brief 2 0 1 4
•
w w w . a f d b . o r g
Figure 38: Budgetary coefficients for pharmaceuticals by deciles of total SPY, and by sub-items, in 201038
0.025 0.02 Drugs
0.015 0.01
Other Pharmaceu cals
0.005 0 d1 d2 d3 d4 d5 d6 d7 d8 d9 d10
Table 12: Breakdown of inequality in spending on pharmaceuticals by expenditure sub-category Concentration Index
Gini Index Year
2000
2005
2010
38
Item
Yk in D
Sk
Kakwani Index
Gk
RkSkGk/G
Marginal Effect
Gk
SkCk/C
Ck-G
Medicines
24.448
34.4%
0.813
33.1%
-0.013304
0.404
31.31%
-0.004558
Other pharmaceuticals
1.904
2.7%
0.963
2.1%
-0.005545
0.462
2.79%
0.053442
PHARMACEUTICALS
26.351
37.2%
0.799
35.3%
-0.019
0.408
34.10%
-0.000558
Medicines
39.783
35.6%
0.782
34.7%
-0.008982
0.435
34.59%
0.020992
Other pharmaceuticals
7.728
6.9%
0.857
4.7%
-0.02157
0.322
4.98%
-0.092008
PHARMACEUTICALS
47.510
42.6%
0.740
39.5%
-0.031
0.417
39.57%
0.002992
Medicines
47.795
33.1%
0.792
32.3%
-0.008436
0.366
28.69%
-0.018674
Other pharmaceuticals
12.995
9.1%
0.788
5.5%
-0.035153
0.202
3.23%
0.182674
PHARMACEUTICALS
60.790
42.2%
0.724
37.8%
-0.044
0.331
31.91%
-0.053674
Our calculations from the pooled budget data on the 2010 survey published on site of the National Institute of Statistics (INS) http://www.ins.nat.tn/indexfr.php
39 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
The analysis focused on the trends in the inequality in health spending
spending in 2010. Medical consultations (16.3% of health spending)
in Tunisia and the contribution of the various items of expenditure to the
and medical devices (1.5% of health spending) also had equalizer
formation of this inequality. The analysis is based on data on total health
effects. These changes may be attributed to the increased availability of
spending per person per year (SPY), as established by the national
physicians and the improved national coverage in terms of pharmacies.
budget and household consumption surveys for 2000, 2005 and 2010. The items where inequality has worsened and produced an inertia effect The trends and sources of inequality are captured mainly through the
were long-term illness (17% of spending), the spending on hospital stay
calculation and breakdown of the Gini index and the concentration index
and medical surgery (8.6%) and paramedical or x-ray, scanner and
by expenditure category. This approach is likely to furnish us with
biology services (8% of health spending). These expenses are related to
information on:
the demographic and epidemiological transition. To reduce the corresponding inequality, government authorities need to adopt specific
•
The overall inequality in health spending and its trends;
policies targeting the most vulnerable groups among the young and the
•
the inequality of the SPY in each health care expenditure item
elderly.
and sub-item; •
the contribution of the inequality of each SPY to the total inequality;
Dental care continues to be characterized by unusually high levels of
•
the marginal effect (equalizer or non-equalizer) of the variation of
inequality and lack of access suffered by disadvantaged classes.
a particular SPY on the total inequality in health expenditure.
Improved coverage of the territory by dental practices and increased public awareness of the importance of oral health should curb one of the
The outcomes showed that overall inequality declined from 2000
causes of this inequality. Similarly, special processing of their dental
to 2010. The breakdown of inequality indicators reveals that this
expenditure refund by health insurance, with no competition with
trend was almost exclusively due to a decrease in inequality in
recurrent expenditures, should contribute to reducing inequalities in
pharmaceutical expenditures, which accounted for 42.2% of health
dental care access.
40 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
Economic Brief 2 0 1 4
•
AfDB
w w w . a f d b . o r g
5. General Conclusion
D
2- On the demand side:
espite the progress achieved, health inequalities remain considerable and relatively little known in Tunisia. In light of the
analysis conducted, there is significant elbow room for reducing these
2-1- It is important to reduce financial barriers to health care access by
inequalities.
better targeting the poor who benefit from free medical assistance.
1- From the supply side:
2-2- Pharmaceuticals are a significant drain on the budgets of the poorest households and it is necessary to reduce this weight by
1-1- In the public sector, it is necessary to revitalize primary health
ensuring good governance of public pharmacies.
care by improving the operation 2-2- There is a need to ensure a better collective coverage of long1-2- It is also important to strengthen Level II which seems
term illness, hospital stay and medical surgery, x-rays and scans.
to be the weak link in the system. Better coverage of the territory in
Knowing the profile of households that incur these expenses will make
terms of Level II beds should necessarily go hand-in-hand with the
it possible to better target them, if need be.
provision of more specialized physicians for the poorest regions in light of the demographic and epidemiological transition.
2-3- Dental care continues to be characterized by extremely high inequalities in expenses. Improved coverage of the territory in terms of
1-3- Efforts should be made to ensure that at each level the system
availability of dental practices and greater public awareness of the
performs its assigned tasks under the best possible conditions.
importance of dental health should curb one of the causes of the
These tasks should be clearly defined. Each hospital institution should
inequality. Similarly, a special processing of reimbursement for dental
have a scheme of work that allows for coherent strategic
expenses by health insurance, apart from the recurrent expenses,
management.
should contribute to reducing inequalities in dental care access.
1-4- The specific incentives that were introduced to encourage
3- On the institutional side:
physicians to settle in deserted areas should be evaluated. Publicpublic and possibly private-public partnerships should be instituted.
3-1- It is necessary to aim at reducing social and regional inequalities
Also, it is important to negotiate with corporations an institutional
in health care.
framework to better regulate the opening of private practices. 3-2- There is a need to produce and monitor indicators for assessing the It is necessary to determine measures that should be
progress of specific categories not only at the national level but also at the
implemented to enhance health care delivery at local or regional level,
local level. It is important to conduct periodic surveys on the status of health,
as part of an overall regional development policy.
health care use, or the failure to seek health care for financial reasons.
1-5-
41 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
AfDB 2 0 1 4
•
Economic Brief w w w . a f d b . o r g
Bibliography
•
Aïach P. (2000), « De la mesure des inégalités : enjeux sociopolitiques et théoriques » (Measurement of inequality: Socio-Political and Theoretical Challenges), in Leclerc A et al., Les inégalités Sociales de Santé (Social Inequalities in Health), La Découverte/INSERM.
•
Aïach P., Fassin D. (2004), « L’origine et les fondements des inégalités sociales de santé » (The origins and foundations of social inequalities in health), Article in French published in La Revue du Praticien, 54.
•
Allin S., Masseria C., Mossialos E., (2006), “Inequality in health care use among older people in the United Kingdom: an analysis of panel data”. Working Paper No. 1/2006. http://eprints.lse.ac.uk/19262/1/LSEHWP1.pdf
• •
Arfa Ch. and Elgazzar H. (2013), Consolidation and Transparency: Transforming Tunisia’s Health care for the Poor. UNICO Studies Series 4. World Bank. Arfa Ch. and Achouri H. (2008), “Tunisia: Good Practice in Expanding Health care Coverage: Lessons from Reforms in a Country in Transition”, in World Bank, Good practices in health financing lessons from reforms in low– and middle–income countries.
•
Arfa Ch., Souiden A., Achour N. (2007), National Health Accounts in Tunisia: Resullts for Years 2004 and 2005, Ministry Public Health.
•
World Bank (2008) Diagnostic Study. Performance of Public Health Institutions in Tunisia. http://documents.banquemondiale.org/curated/fr/2008/10/16262037/etude-diagnostique-performance-des-%C3%A9tablissementspublics-de-sant%C3%A9-en-tunisie
•
Ben Gobrane H.L et al. (2012), « Motifs du recours aux services d’urgence des principaux hôpitaux du Grand Tunis ». (Reasons for using emergency departments of major hospitals in Greater Tunis), Eastern Mediterranean Health Journal, vol. 18.
•
Bouchoucha I. and Vallin J. (2007), "Regional inequalities in infant mortality in Tunisia”. The 5th African Population Conference. http://uaps2007.princeton.edu/papers/70240
•
De Looper M. Lafortune, G. (2009), “Measuring disparities in Health Status and in Access and Use of Health care in OECD Countries”. DELSA/HEA/WD/HWP (2009)2. http://search.oecd.org/officialdocuments/displaydocumentpdf/?cote=DELSA/HEA/WD/HWP(2009)2&doclanguage=en
•
Fleurbaey M., Schokkaert E. (2011), “Equity in Health and Health Care” CORE DISCUSSION PAPER 2011/26 http://www.ecore.be/DPs/dp_1309869083.pdf
•
Gastineau, B. (2003), "Social and Economic Patterning of Health among Women / Les facteurs sociaux et économiques de la santé des femmes," in Arber S. and Khlat M.(eds), Paris, CICRED. http://www.cicred.org/Eng/Publications/Books/TunisHealthWomen/TunisGastineau.pdf
•
Hajem S et al (2011), « Statistique Nationale sur les Causes médicales de Décès -2009 » Institut National de Santé Publique. (National Statistics on Medical Causes of Death – 2009)" Research Unit on Aging and Medical Causes of Death – National Public Health Institute
•
Jusot F. (2003), « Inégalités Sociales de Mortalité : Effet de la Pauvreté ou de la Richesse » (Social Inequalities in Mortality: Effect of Poverty or Wealth). http://epee.univ-evry.fr/EPEE/colloques/jusotevry200312021.PDF
42 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
Economic Brief 2 0 1 4
•
AfDB
w w w . a f d b . o r g
•
Kakwani, N. C. (1980), Income Inequality and Poverty: Methods of Estimation and Policy Applications. New York: Oxford University Press.
•
Ladner J. Bailly L. Pitrou I, Tavolacci M.P (2008), « Les patients auto-référés dans les services hospitaliers d'urgence: motifs de recours et comportements de consommation de soins » (Self-referred patients in hospital emergency departments: reasons for use and patterns of health care consumption), in Pratiques et Organisations de Soins (Health Care Practies and Organization), vol. 39,1.
•
Lerman, R.I. and Yitshaki S. (1985), “Income Inequality Effects by Income Source, a New Approach and Applications: the United States.” Review of Economics and Statistics, Vol. 61.
•
Maurey H. (2013), Rapport d’information Sénat n°335, Session ordinaire (Senate information report No. 335, Regular Session), 2012-2013 «Présence médicale sur l’ensemble du territoire» (Nationwide Medical Presence). http://www.senat.fr/rap/r12-335/r12-3351.pdf
•
Ministry of Regional Development (2011), Livre Blanc (White Paper) http://eeas.europa.eu/delegations/tunisia/documents/more_info/livreblanc_devreg_nov11_fr.pdf
•
Ministry of Health (2010), Health Map 2010 http://www.santetunisie.tn/msp/images/CSfinale2010.pdf.
•
National Office for the Family and the Population/UNICEF (2007), Multiple Indicator Cluster Survey. MICS Tunisia.
•
Haut Conseil de la santé publique (Public Health Council) (2009), Les inégalités sociales de santé : sortir de la fatalité. (Social Inequalities in Health: Overcoming the Fatality). www.hcsp.fr/explore.cgi/hcspr20091112_inegalites.pdf
•
O’Donnell O. et al. (2008), Analyzing Health Equity Using Household Survey Data. A Guide to Techniques and Their Implementation. WBI Learning Resources Series The Wold Bank
•
WHO (2010), Country Cooperation Strategy for WHO and Tunisia (2010-2014) www.emro.who.int/docs/CCS_Tunisie_2010_FR_14489.pdf, Consulted on 1-3-2101
•
Or Zeynep et al. (2009), «Inégalités de recours aux soins en Europe. Quel rôle attribuable aux systèmes de santé ? » (Inequalities in health care use in Europe. What role for health systems?), Revue économique, 2009/2 Vol. 60, p. 521-543. DOI: 10.3917/reco.602.0521.
•
Potvin L., Moquet M.-J., Jones C. (2010), Réduire les inégalités sociales en santé (Reducing Inequalities in Health). Saint-Denis : INPES, coll. Santé en action, 2010.
•
Van Doorslaer E, Koolman X. (2004), “Explaining the differences in income-related health inequalities across European countries”, Health Economics, 13, 7.
•
Van Doorslaer E., Koolman X, Jones A. (2004), “Explaining income-related inequalities in doctor utilisation in Europe”, Health economics 13,7.
•
Van Doorslaer, E., Koolman X, Jones, Puffer F. (2002). “Equity in the use of physician visits in OECD countries: has equal treatment for equal need been achieved?” In Measuring Up: Improving Health Systems Performance in OECD Countries
•
Van Ourti T. (2004), “Measuring horizontal inequity in Belgian health care using a Gaussian random effects two part count data model”. Health economics, 13, 7.
•
Wagstaff A., van Doorslaer E., Watanabe N. (2003), “On decomposing the causes of health sector inequalities with an application to malnutrition inequalities in Vietnam”. Journal of Econometrics 112.
•
Wagstaff A., van Doorslaer E. (2000), “Income inequality and health: what does the literature tell us?” Annual Review of Public Health, 21. 43 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
A
f
r
i
8
BEN AROUS
c
a
n
D
e
v
e
l
o
p
m
e
n
t
45
B
a
n
k
1
1
Rank of cluster depending on the criteria
22
37
2
20
28
19
1
1
0.263
0.226
0.25
0.044
0.3
0.551
0.76
Proportion of PHC offering medical consultation 6 days per week
4
20414
10640
13144
8425
15030
14847
18529
22299
Inhabit. /PHC FTE (Full-Time Equivalent)
2
5588
6118
5677
4816
9498
7090
6144
5242
Inhabit. per primary care physician
1
0.2
1.14
0.86
2.33
1.54
0.54
0.52
0.6
Frontline medical laboratory unit per 100000 inhabit.
1
0.2
1.14
0.86
2.52
0.88
0.54
0.52
0.2
Frontline radiology unit per 100000 inhabit.
1
0.7
2.94
1.5
4.66
1.54
1.9
1.39
1.61
Frontline dental chairs 100000 per inhab.
1
3877
5883
5575
5923
5845
7229
5203
6385
Inhab. /daytime pharmacy
3
45468
33989
46550
17769
50656
36870
41250
45273
Inhab. /Nighttime pharmacy
1
6.9
3.6
5.6
2.3
0.8
1.9
3.8
4
Medical Laboratory per 100000 inhab.
1
30.1
18.3
23.9
19.4
27.6
19
24.6
23.7
Haemodialysis machines per 100 000 inhab. (Total, public and private)
•
4
20414
6180
6006
5368
4035
9218
11786
19920
Inhabit. per Primary Centre (PHC)
2 0 1 4
TUNIS
SOUSSE
SFAX
14
219
MEDENINE
MONASTIR
20
17
22
Average distance from a General Hospital (GH)
MANOUBA
20
14
Average distance from a Regional Hospital (RH)
ARIANA
Governorate
Economic Brief
AfDB
w w w . a f d b . o r g
Annex 1: Indicators of Health care Facilities by Cluster Table 13: Health infrastructure indicators: Cluster 1.1
A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
Ranking
NABEUL
KEBILI
BIZERTE
2
24
40
23
3
84
299
86
3
6022
1784
6073
Inhabit. per Primary Centre (PHC)
3
14%
9%
14%
Proportion of PHC offering medical consultation 6 days per week
2
14858
4687
14197
Inhabit. /PHC FTE (Full-Time Equivalent)
3
7682
4432
7699
Inhabit. per primary care physician
2
1.46
1.99
2.01
Frontline medical laboratory unit per 100000 inhabit.
2
1.46
1.99
2.01
Frontline radiology unit per 100000 inhabit.
3
2.79
1.99
2.74
Frontline dental chairs 100000 per inhab.
2
5791
10047
6833
1
31367
30140
36440
Inhab. Inhab. /daytime /Nightpharmacy time pharmacy
2
2.5
1.3
1.8
Medical Laboratory per 100000 inhab.
3
18.5
19.2
24.2
Haemodialysis machines per 100 000 inhab. (Total, public and private)
•
Average distance from a General Hospital (GH)
2 0 1 4
Governorate Average distance from a Regional Hospital (RH)
AfDB Economic Brief
w w w . a f d b . o r g
Table 14: Health infrastructure indicators, Cluster 1.2
46
B
a
n
k
A
f
r
i
c
a
n
D
e
v
e
36
44
GAFSA
LE KEF
l
o
p
m
e
n
t
B
a
n
59
42
29
3
TOZEUR
ZAGHOUAN
Ranking
40
TATAOUINE
SILIANA
44
32
GABES
MAHDIA
22
Average distance from a Regional Hospital (RH)
BEJA
Governorate
47
k
4
1
3
20%
6%
5%
18%
8%
19%
20%
25%
10%
Proportion of PHC offering medical consultation 6 days per week
1
7809
8295
8141
8116
7979
7998
9316
9726
11067
Inhab. /PHC FTE (Full-Time Equivalent)
1
4736
3996
4177
4500
4833
5832
6379
6821
5777
Inhab. per primary care physician
4
2.93
5.8
4.1
3.85
2.78
3.51
2.66
2.49
1.96
Frontline medical laboratory unit per 100000 inhab.
4
2.35
3.86
3.42
3.85
2.52
2.73
3.25
2.21
1.96
Frontline radiology unit per 100000 inhab.
4
2.93
4.83
4.1
3.42
2.78
3.51
2.66
3.04
2.61
Frontline dental chairs 100000 per inhab.
3
8974
7962
7310
11700
9007
8848
9138
6694
9279
2
28417
34500
29240
29250
30485
36657
42263
36150
38275
Inhab. Inhab. /daytime /Nightpharmacy time pharmacy
3
1.2
0
0.7
0.4
1
0.8
0.6
1.7
1
Medical Laboratory per 100000 inhab.
2
28.2
30.9
25.3
16.7
19.4
21
24.8
16.3
14.4
Haemodialysis machines per 100 000 inhab. (Total, public and private)
•
3480
4709
2358
2629
3538
2759
3635
3929
3257
Inhab. per Primary Centre (PHC)
2 0 1 4
69
333
307
143
44
206
226
168
127
Average distance from a General Hospital (GH)
Economic Brief
AfDB
w w w . a f d b . o r g
Table 15: Health infrastructure indicators, Cluster 1.3
A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
Ranking
SIDI BOUZID
KASSERINE
KAIROUAN
4
44
50
46
36
JENDOUBA
2
157
257
98
157
Average distance from a General Hospital (GH)
2
3327
4238
4305
3712
Inhab. per Primary Centre (PHC)
2
9%
16%
17%
11%
Proportion of PHC offering medical consultation 6 days per week
3
13130
12530
13596
12432
Inhab. /PHC FTE (Full-Time Equivalent)
4
3438
5720
8354
8637
Inhab. per primary care physician
3
2.67
2.78
1.97
1.42
Frontline medical laboratory unit per 100000 inhab.
2
2091
2.08
1.97
1.65
Frontline radiology unit per 100000 inhab.
4
2.91
2.31
1.97
1.89
Frontline dental chairs 100000 per inhab.
4
11458
11376
9995
10076
4
41250
54038
43054
70533
Inhab. Inhab. /daytime /Nightpharmacy time pharmacy
4
0.2
0.2
0.5
0.5
Medical Laboratory per 100000 inhab.
4
13.6
11.6
8.8
15.8
Haemodialysis machines per 100 000 inhab. (Total, public and private)
•
Average distance from a Regional Hospital (RH)
2 0 1 4
Governorate
AfDB Economic Brief
w w w . a f d b . o r g
Table 16: Health infrastructure indicators, Cluster 1.4
48
k
A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
49
B
a
n
k
Ranking
TUNIS
TOZEUR
1
52.7
30.9
26.6
23.9
MONASTIR
SOUSSE
26
MANOUBA
Governorate Beds per 10 000 inhab (pub + private)
1
40.4
30.9
23.9
22.5
26
Hospital beds / 10 000 inhab.
1
12.3
0
2.8
1.4
0
Clinic beds / 10 000 inhab
1
5.03
4.83
1.63
2.93
0
General surgery beds / 10 000 inhab.
1
15.69
10.03
8.13
8.31
0
Gyneco beds/ 10000 WCBA
1
21.6
11.41
7,28
8,81
0
Paediatrics beds/ 10000 children below 15 years
1
1.96
1.93
0.98
0.66
0
Ophthalmology beds /10000 inhab.
2
0.66
0.97
0.67
0.74
0
1
1.73
0
0.98
0.54
3.25
ENR beds Ortho/10000 paedics inhab. beds per 10000 inhab.
1
2.22
0.97
1.18
1.47
0
Cardiology bed per 10000 Inhab.
1
0.99
0.97
0.75
0.66
0.79
1
0
0
0.49
0.62
16.41
Anaesthe- Psychiatric sia and in- beds/10000 tensive inhab. care beds /10000 inhab.
Economic Brief
2 0 1 4
Table 17: Bed availability indicators, Cluster 2.1 •
AfDB
w w w . a f d b . o r g
A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
Ranking
MEDENINE
MAHDIA
LE KEF
2
20.9
16.5
21.7
3
16.6
14.9
21.2
2
4.3
1.6
0.6
Clinic beds / 10 000 inhab.
3
2.68
1.94
2.73
General surgery beds / 10 000 inhab.
2
7.99
6.97
8.32
Gyneco beds/ 10000 WCBA
3
8.89
4.94
7.04
Paediatrics beds/ 10000 children below 15 years
3
0.72
0.25
0.58
Ophthalmology beds /10000 inhab.
3
0.61
0.5
0.58
3
0.77
0.81
0.58
ENR beds Ortho/10000 paedics inhab. beds per 10000 inhab.
2
0.66
0.86
0.17
Cardiology bed per 10000 Inhab.
3
0.53
0.25
0.31
Anaesthesia and intensive care beds /10000 inhab.
2
0.13
0.63
0.58
Psychiatric beds/10000 inhab.
•
Hospital beds / 10 000 inhab.
2 0 1 4
Governorate Beds per 10 000 inhab (pub + private)
AfDB Economic Brief
w w w . a f d b . o r g
Table 18: Bed availability indicators, Cluster 2.2
50
B
a
n
k
A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
Ranking
TATAOUINE
KEBILI
GAFSA
BEJA
Governorate
3
17.7
19.8
23.2
17.9
Beds per 10 000 inhab (pub + private)
2
17.7
18.5
22
17.9
Hospital beds / 10 000 inhab.
4
0
1.3
1.2
0
Clinic beds / 10 000 inhab.
2
2.05
3.98
2.96
2.45
General surgery beds / 10 000 inhab.
3
6.52
7.11
5.95
9.1
Gyneco beds/ 10000 WCBA
2
8.33
8.4
6.66
6.03
Paediatrics beds/ 10000 children below 15 years
2
0.68
0.53
1.33
0.65
Ophthalmology beds /10000 inhab.
1
0.68
0.8
0.33
0.65
2
0.82
0
0.86
0.98
ENR beds Ortho/10000 paedics inhab. beds per 10000 inhab.
4
0
0
0.8
0
Cardiology bed per 10000 Inhab.
2
0.68
0.53
0.18
0.33
Anaesthesia and i ntensive care beds /10000 inhab.
4
0
0
0
0
Psychiatric beds/10000 inhab.
Economic Brief
2 0 1 4
Table 19: Bed availability indicators, Cluster 2.3
51
n
k •
AfDB
w w w . a f d b . o r g
A
f
r
i
c
a
n
D
e
v
e
l
19.1
15.8
GABES
JENDOUBA
o
p
m
e
n
t
52
B
a
n
k
27.9
ZAGHOUAN
4
16.7
SILIANA
Ranking
10.7
14.9
12.8
SIDI BOUZID
NABEUL
KASSERINE
12.8
17
BIZERTE
KAIROUAN
5.5
BEN AROUS
4
27.9
16.7
10.7
12.7
12.3
12.1
14.3
17
16.5
3.8
8.3
Hospital beds / 10 000 inhab.
3
0
0
0
2.2
0.6
0.6
1.5
2.1
0.5
1.7
4.3
Clinic beds / 10 000 inhab.
4
1.47
1.79
1.45
1.46
1.16
1.14
1.77
1.66
2.47
0.45
0
General surgery beds / 10 000 inhab.
4
5.22
4.84
4.57
6.74
4.89
3.77
3.31
7.91
6.83
2.06
3.89
Gyneco beds/ 10000 WCBA
4
4.49
3.95
3.69
3.94
4.08
3.97
2.58
5.17
8.47
0
0
Paediatrics beds/ 10000 children below 15 years
4
0
0.43
0.39
0.27
0.46
0.36
0.35
0.97
0.37
0
0
Ophthalmology beds /10000 inhab.
4
0
0.43
0.29
0.27
0.46
0.36
0.35
0.55
0.37
0
0
4
0
0
0
0.8
0.65
0.54
0
0
0.55
0
0
ENR beds Ortho/10000 paedics inhab. beds per 10000 inhab.
3
0
0.77
0.48
0.4
0
0.54
0.71
0.41
0.73
0
0.6
Cardiology bed per 10000 Inhab.
4
0
0
0.15
0.27
0.23
0.27
0
0.83
0.11
0.21
0.7
Anaesthesia and intensive care beds /10000 inhab.
3
0
0
0
0
0
0.71
0.57
0
0
0
0
Psychiatric beds/10000 inhab. •
12.6
Beds per 10 000 inhab (pub + private)
2 0 1 4
ARIANA
Governorate
AfDB Economic Brief
w w w . a f d b . o r g
Table 20: Bed availability indicators, Cluster 2.4
A
f
r
41.1
33.8
60.5
100.6
1
MONASTIR
SFAX
SOUSSSE
TUNIS
Ranking
i
c
a
n
D
e
v
e
l
o
p
m
e
n
1
Ranking
53
t
B
a
n
k
17.5
22.3
20.1
19.8
28.1
21.5
1
ARIANA
MONASTIR
SFAX
SOUSSSE
TUNIS
ARIANA
Ranking
Governorate Pharmacists / 100000 inhab. private
10.3
3.3
TUNIS
SOUSSSE
4.5
3.3
MONASTIR
SFAX
2.4
ARIANA
Governorate General surgeons /100000 inhab.
29.9
ARIANA
1
202.7
109.7
68.0
74.9
48.0
1
7.5
12.8
10.5
1
29.0
40.9
30.2
1
25.5
35.8
27.0
25.5
18.0
21.3
Dentists /100000 inhab. private
1
1
11.3
3.9
3.7
2.5
6
1
8.0
7.4
8.3
2.4
16.9
4.8
1
33.5
43.2
35.3
27.8
34.9
26.1
1
391.6
666.5
436.3
315.4
400.9
139.2
Paramedical Dentists /100000 Dentists /100000 inhab. staff/100 000 inhab. public total inhab. public
1
7.6
3.6
4.4
2.1
2.4
1
1
21.8
42.6
10.3
22.3
9.7
24.1
Paramedical staff/100 000 inhab. private
10.3
3.9
4.1
2.7
2
1
1
413.4
709.1
446.6
337.7
410.6
163.3
Paramedical staff/100 000 inhab. total
0.5
1.07
1.61
1.62
0
•
23.3
31.0
19.7
1
13.6
5.4
5.7
3.3
3.8
1
209.0
93.8
94.6
55.3
67.5
Psychiatrists/ 100000 inhab.
1
158.9
98.1
80.6
69.3
64.9
General pract Specialists/ 100000 inhab. 100000 inhab. (total). (total)
2 0 1 4
3.2
8.7
2.2
8.2
6.6
6.6
7.6
17.4
Pharmacists/ Pharmacists/ 100000 inhab. 100000 inhab. public total
1
6.85
3,78
2,84
2,3
2,83
1
367.9
191.9
175.2
124.6
132.3
Physicians / 100000 inhab. (total)
Cardiologists Anaesthetists/ 100000 inhab. /100000 inhab.
1
165.2
82.2
107.2
49.7
84.3
Physicians/ 100000 inhab. (private)
Surgeons/ 100000 inhab.
1
106.9
44.6
60.5
21.5
49.4
Specialists/ 100000 inhab (private).
Ophthalmologists/ 100000 inhab.
1
58.3
37.6
46.7
28.1
34.9
Physicians/ General pract. 100000 inhab. 100000 inhab. (public (private) sector)
Gyneco-obstetriPodiatrists cians /100000 /10000 women of childbea- children ring age (15-49 years)
1
102.2
49.2
34.2
33.8
18.1
Governorate General pract Specialists/ 100000 inhab. 100000 (public) inhab. (public sector)
Economic Brief
AfDB
w w w . a f d b . o r g
Table 21: Human resource indicators: Cluster 3.1
A
f
r
i
c
a
n
D
36.1
31.7
21.7
22.6
45.2
MAHDIA
MANOUBA
MEDENINE
NABEUL
ZAGHOUAN
e
v
e
l
o
p
m
e
n
t
B
54
a
n
k
1.8
0
3.5
2.7
2.3
2
MANOUBA
MEDENINE
NABEUL
ZAGHOUAN
Ranking
2.5
2
1.6
General surgeons /100000 inhab.
MAHDIA
GABES
BIZERTE
BEN AROUS
Governorate
4
29.0
GABES
Ranking
29.6
BIZERTE
3
62.8
34.1
30.1
51.8
51.7
40.9
45.0
35.3
2
1.25
2.17
1.55
1.04
1.5
1.12
1.82
2.36
2
2.9
4.4
3.1
4.6
3.3
2.8
5.7
5.7
Gyneco-obstetriPodiatrists cians /100000 /10000 women of childbearing children age (15-49 years)
2
17.6
11.6
8.3
20.1
15.6
11.9
15.4
9.5
2
2
2.3
3.7
3.1
1.1
2
2.5
3.3
4.2
2
7.0
27.1
24.6
9.0
11.4
20.2
22.5
30.0
2
25.8
55.0
50.7
36.9
34.6
36.8
48.3
68.2
2
1.8
1.9
2
1.4
1
2.2
2
2.6
2
88.6
89.1
80.7
88.7
86.3
77.7
93.3
103.5
3
1.2
2.1
2
0.5
0.5
1.4
2.4
1.6
2
63.9
50.5
47.8
59.7
59.3
45.6
55.4
64.1
2
0
1.2
0.7
0.3
1
0.8
0.9
0.7
2
1.17
0.53
0
13.29
1.21
0.55
0.18
0.17
Psychiatrists/ 100000 inhab.
2
24.6
38.7
32.9
29.0
27.0
32.1
37.9
39.5
General pract Specialists/ 100000 inhab. 100000 inhab. (total). (total)
Anaesthetists/ 100000 inhab.
Physicians/ 100000 inhab. (total)
Cardiologists /100000 inhab.
Physicians/ 100000 inhab. (private)
Surgeons/ 100000 inhab.
Specialists/ 100000 inhab (private).
Ophthalmologists/ 100000 inhab.
18.8
27.9
26.1
27.9
23.2
16.6
25.8
38.3
Physicians/ General pract. 100000 inhab. 100000 inhab. (public (private) sector) •
25.8
General pract Specialists/ 100000 inhab. 100000 inhab. (public) (public sector)
2 0 1 4
BEN AROUS
Governorate
AfDB Economic Brief
w w w . a f d b . o r g
Table 22: Human resource indicators. Cluster 3.2
A
f
r
i
c
a
n
D
e
v
e
20.7
18.0
12.6
BIZERTE
GABES
MAHDIA
l
o
p
m
e
n
t
B
a
Ranking
BEN AROUS
ZAGHOUAN
NABEUL
MEDENINE
2
18.8
15.8
20.5
21.5
16.5
24.8
BEN AROUS
MANOUBA
Pharmacists / 100000 inhab. private
Governorate
2
2.4
1.8
2.5
2.4
3.8
3.3
2.5
1.6
1.4
2
21.2
17.6
23.0
23.9
20.3
15.9
20.5
22.3
26.1
Pharmacists/ Pharmacists/ 100000 inhab. 100000 inhab. public total
2
13.4
8.2
14.9
12.7
9.0
11.9
11.9
17.6
21.1
Dentists /100000 inhab. private
3
3.7
5.3
4.5
2.9
3.8
3.3
3.6
3.3
2.8
2
17.1
13.5
19.4
15.6
12.7
15.1
15.5
20.9
23.9
4
259.5
261.6
221.7
266.1
272.3
302.0
283.0
294.4
175.2
Dentists /100000 Dentists Paramedical inhab. public /100000 inhab. staff/100 000 inhab. public total
2
9.9
1.2
14.2
5.7
10.8
11.1
10.2
12.4
13.5
Paramedical staff/100 000 inhab. private
Economic Brief
2 0 1 4
Table 22: Human resource indicators. Cluster 3.2
55
n
k
•
AfDB
w w w . a f d b . o r g
A
f
r
i
c
3
31.7
a
n
D
e
v
e
l
o
p
m
4
1.3
e
n
t
B
a
56
n
k
Ranking
BEJA
3
12.7
Governorate Pharmacists / 100000 inhab. private
Ranking
BEJA
Governorate General surgeons /100000 inhab
Ranking
BEJA 2
44.7
2.0
Pharmacists/ 100000 inhab. public
3
1.42
3
14.7
Pharmacists/ 100000 inhab. total
3
2.9
Gyneco-obstetricians Podiatrists /100000 women of /10000 children childbearing age (15-49 years)
3
13.1
3
2
3
9.8
4
3.9
Dentists Dentists /100000 inhab. /100000 private inhab. public
3
2.6
3
74.1
Physicians / 100000 inhab. (total)
3
13.7
Dentists /100000 inhab. total
2
1.6
2
329.5
Paramedical staff/100 000 inhab. public
4
0
3
4.6
2
334.1
Paramedical Paramédical staff/100 000 /100 000 inhab. private hab. total
4
0.33
3
29.4
Psychiatrists/ 100000 inhab.
3
44.7
General pract Specialists/ 100000 inhab. 100000 (total) inhab. (total)
Cardiologists/ Anaesthetists/ 100000 inhab. 100000 inhab.
3
29.4
Physicians/ 100000 inhab. (private sector)
Surgeons/ 100000 inhab.
3
16.3
Specialists/ 100000 inhab. (private sector).
Ophthalmologists/ 100000 inhab.
3
13.1
Physicians/ General pract. 100000 100000 inhab. inhab. (public (private) sector)
•
Specialists/ 100000 inhab. (public sector).
2 0 1 4
Governorate General pract./ 100000 inhab. (public)
AfDB Economic Brief
w w w . a f d b . o r g
Table 23: Human resource indicators. Cluster 3.3
A
f
r
i
c
a
n
D
e
v
23.9
24.3
25.9
37.2
32.3
24.7
37.2
45.8
52.2
2
JENDOUBA
KAIROUAN
KASSERINE
KEBILI
LE KEF
SIDI BOUZID
SILIANA
TATAOUINE
TOZEUR
Ranking
e
l
o
p
m
e
B
a
n
0.7
0.7
2.3
0.7
1.3
0.7
1.9
3
KASSERINE
KEBILI
LE KEF
SIDI BOUZID
SILIANA
TATAOUINE
TOZEUR
Ranking
1.8
t
KAIROUAN
1.7
JENDOUBA
n
1.8
57
GAFSA
Gouvernorat Chirurgiens Généralistes /100000 hab
35.5
GAFSA
k
4
0.67
0.43
4
3.9
2.1
1.7
4
1.9
1.4
1.3
1
1.2
1.3
0.7
1.4
0.9
1.8
4
1.9
0.7
0.9
0.2
1.6
2
0.7
0.7
0.9
0.9
4
1
0.7
0.9
0.5
0.4
0.7
0.2
1.1
0.9
1.2
3
0
0
0
0
0
0.7
0.2
0.9
0.2
0.3
3
0.97
0
0
0.5
0.39
0
0.19
0.82
0.71
0.3
Psychiatre/ 100000 hab
4
81.2
64.3
57.3
46.3
64.3
63.0
45.3
57.2
51.5
70.4
Physicians/ 100000 inhab. (total)
•
0.3
1.7
2.3
0.7
1.6
2.1
2.1
1.5
Anesthésistes/ 100000 hab
4
16.4
8.2
10.3
9.7
18.3
13.3
9.7
18.2
15.6
19.8
Specialists /100000 inhab. (total)
2 0 1 4
0.5
0.97
0.67
0.49
0.82
0.91
1.09
4
64.7
56.1
47.0
36.6
46.0
49.8
35.6
38.9
35.9
50.6
General PR actioners/ 100000 inhab. (total)
Cardialogues/ 100000 hab
4
17.4
16.4
11.5
16.0
24.2
20.6
15.7
23.9
19.4
26.3
Physicians/ 100000 inhab. (private)
Chirurgiens/ 100000 hab
4
4.8
6.2
1.7
4.1
10.5
8.0
6.0
9.3
7.3
11.2
Specialists/ 100000 inhab. (private)
Ophtalm./ 100000 hab.
4
12.6
10.3
9.8
11.9
13.6
12.6
9.7
14.7
12.1
15.1
General practitioners /100000 inhab. (private)
Pédiatres/ 10000 enfants
4
63.8
47.9
45.7
30.3
40.1
42.5
29.6
33.2
32.1
44.1
Physicians/ 100000 inhab. (public)
Gynéco-obstétriciens /100 000 femmes en âge de procréer (15 – 49 ans)
4
11.6
2.1
8.5
5.6
7.8
5.3
3.7
8.9
8.3
8.6
Specialists Gouvernorat General pract./ 100000 100000 inhab. (public) inhab. (public)
Economic Brief
AfDB
w w w . a f d b . o r g
Table 24: Human resource indicators. Cluster 3.4
A
f
r
i
c
a
n
D
e
v
12.3
12.5
KAIROUAN
KASSERINE
e
l
o
p
m
e
B
a
n
Ranking
4
17.4
t
TOZEUR
19.2
10.9
14.0
TATAOUINE
n
15.8
58
SILIANA
SIDI BOUZID
LE KEF
15.9
11.3
JENDOUBA
KEBILI
15.1
GAFSA
4
7.7
4.8
7.3
6.8
6.2
8.6
5.1
9.5
8.0
7.1
Pharmacists / 100000 inhab. public
3
5.8
3.4
5.1
3.2
4.3
4.0
2.8
2.9
2.6
3.8
Pharmacists/ 10 000 inhab. total
4
13.5
8.2
12.4
9.9
10.5
12.6
7.9
12.3
10.6
10.9
Dentists/100000 inhab. private
3
492.8
318.1
287.6
199.8
385.4
382.2
215.1
270.3
273.6
397.5
Dentists/100000 inhab. public
4
4.8
1.4
1.7
2.7
9.0
4.6
0.7
2.0
1.7
5.9
Dentists/100000 inhab. total
3
497.6
319.4
289.3
202.4
394.4
386.9
215.8
272.3
275.3
403.4
Paramedical staff/100 000 inhab. public
•
Pharmacists/ 100000 inhab. private
2 0 1 4
Gouvernorat
AfDB Economic Brief
w w w . a f d b . o r g
Table 24: Human resource indicators. Cluster 3.4
k
Economic Brief 2 0 1 4
•
AfDB
w w w . a f d b . o r g
Annex 2: Health Expenditure Nomenclature 2000 in 2005 (source : www.ins.nat.tn )
41 ROUTINE MEDICAL CARE
423 MRI SCAN 4231 Mri scan in a public institution 4232 Mri scan in a private institution 4239 Mri scan in a with no indication
411 MEDICAL CONSULTATIONS 4111 Consultations by public institutions 4112 Consultations by private institutions 4119 Consultation with no indication
424 CHILD DELIVERY 4241 Child delivery in a public institution 4242 Child delivery in a private institution 4249 Child delivery with no indication
412 DENTAL CARE 4121 Dental care in public institutions 4122 Dental care in private institutions 4129 Dental care with no indication
425 MEDICAL TREATMENT ABROAD 426 LONG-TERM CARE 4261 Consultations for long-term illnesses 4262 Drugs for long-term illness 4263 Non-classified expenditure on long-term illnesses
413 X-RAY AND ANALYSES 4131 X-ray and medical analyses in public institutions 4132 X-ray and medical analyses in private institutions 4139 X-ray and medical analyses with no indication
43 PHARMACEUTICALS
416 TRADITIONAL HEALER 4161 Traditional healer
431 DRUGS, NURSING 4321 Baby products 4322 Other pharmaceuticals
42 SPECIAL MEDICAL CARE 432 AUTRES PRODUITS PHARMACEUTIQUES 4321 Produits pour bebe 4322 Autre produit pharmaceutique
421 STAY AND SURGERY 4211 Stay and surgery in a public institution 4212 Stay and surgery in a private institution 4219 Stay and surgery with no indication
44 MEDICAL EQUIPMENT 441 MEDICAL EQUIPMENT 4411 Medical eyeglasses
422 SPECIAL DENTAL SURGERY 4221 Special dental care in a public institution 4222 Special dental care in a private institution 4229 Special dental care with no indications
4412 Hearing aids 4419 Other medical equipment
59 A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
A
f
r
i
c
a
n
D
e
v
e
l
o
p
m
e
n
t
B
a
n
k
© 2010 - AfDB - Design, External Relations and Communication Unit/YAL