Page 1
1
Ruprechts-Karls University of Heidelberg
Department of Tropical Hygiene
and Public Health
Postgraduate Master of Science Degree Course
“Community Health and Health
Management in Developing Countries”
The influence of the factor distance on utilisation of a Health
Center in rural Guinea and its implications for the associated
health insurance scheme
Mai – June 2000
Author: Götz Huber
Field Study Tutor: Dr. Michael Marx
Local Tutor : Dr. Franz von Roenne
Page 2
2
1. EXECUTIVE SUMMARY ................................................................................. 5
2. INTRODUCTION ............................................................................................. 6
2.1 Background of the study .............................................................................. 6
2.2 The context of the health insurance scheme “Maliando” .............................. 7
2.3 Guinea: Geography and socio-demograhic data .......................................... 9
2.4 History ........................................................................................................ 10
2.5 Socio-economic situation ........................................................................... 10
2.6 Health sector analysis ............................................................................. 11
2.6.1 The Guinean Health system................................................................ 11
2.6.2 Bamako Initiative (BI) in Guinea ......................................................... 11
2.6.3 Implementation of core BI strategies ................................................... 12
2.7 Background Information of the study area ................................................. 14
2.7.1 Profile of the natural region „Guinee forestière“ ................................. 14
2.7.2 German Agency for Technical Cooperation (GTZ)/ PSR .................... 14
2.7.3 PRIMA / “Maliando” ............................................................................. 15
2.7.4 District Health Sector with the Health Center Yende ........................... 16
2.8 Purpose of the study .................................................................................. 16
2.9 Research question ..................................................................................... 17
2.10 Research objectives ............................................................................... 17
2.10.1 General objective ................................................................................ 17
2.10.2 Specific objectives .............................................................................. 18
3 LITERATURE REVIEW ................................................................................. 19
3.1 Introduction ................................................................................................ 19
3.2 Health service utilisation ............................................................................ 19
3.3 Health care decision process ..................................................................... 21
3.3.1 Illness concepts ........................................................................................ 22
3.3.2 Utilisation of traditional medicine .............................................................. 22
3.4. Utilisation in Guinea ................................................................................... 23
3.4.1 Curative care ........................................................................................ 23
Page 3
3
3.4.2 Antenatal care and EPI ....................................................................... 24
3.5. Utilization related to distance (geographical accessibility) ......................... 24
3.6. Utilization related to costs (financial accessibility) ...................................... 25
3.6.1. Household expenditure for health Guinea ........................................... 25
3.6.2 Costs incurred of overcoming distance ............................................... 26
3.7 Utilization related to Quality of care ............................................................ 28
3.7.1 Perceived quality and reliability ........................................................... 28
3.7.2 Donabedian’s classification of quality of care ....................................... 29
3.8. Mutual health insurance schemes ............................................................. 30
3.8.1 Data on utilisation rates "Maliando" previous Studies ......................... 33
3.9 Methodological aspects .............................................................................. 33
3.9.1 Quantitative techniques used in the study .......................................... 33
5.9.1.3 Concept of distance decay ................................................................ 35
5.9.1.4 Population Census ............................................................................. 35
5.9.1.5 Global Positioning System (GPS) ...................................................... 36
5.9.1.6 Catchment area population ................................................................ 36
5.9.1.7 Coverage/ utilisation data .................................................................. 37
5.9.1.8 Calculating accessibility ..................................................................... 38
5.9.1.9 Calculating Utilisation ........................................................................ 38
3.9.2 Qualitative methods .................................................................................. 39
3.9.2.1 Combination of qualitative and quantitative approaches .................... 39
4 METHODOLOGY .......................................................................................... 41
4.1 Study area description and study population ............................................ 41
4.2 Time frame ................................................................................................. 41
4.3 Study type and design ................................................................................ 41
4.4 Finance and logistics. ................................................................................. 41
4.5 Data collection............................................................................................ 42
4.5.1 Data collecting tools .......................................................................... 42
4.5.2 Selection, staff training and Pretest ...................................................... 43
4.5.3 Data collection procedure/ Computer Data entry .................................. 44
Page 4
4
4.6. Data analysis .............................................................................................. 46
4.7 Validation .................................................................................................... 46
4.8 Limitations .................................................................................................. 47
4.9 Ethical considerations ............................................................................... 47
5. RESULTS ...................................................................................................... 48
5.1 Introduction ................................................................................................ 48
5.2 Population census ...................................................................................... 48
5.2.1 Global figures ...................................................................................... 48
5.2.3 Relating distance to the population figures ......................................... 49
5.3.1 Assessment of Outpatient register (CPC) and of preventive Services .... 49
5.3.1.1 Origin of attendees ........................................................................... 49
5.3.1.2 Age groups ........................................................................................ 50
5.3.1.3 General features attendance curative care (CPC) ............................. 50
5.4.1 Summary of the Findings .................................................................... 56
6. Discussion ................................................................................................... 58
6.1 Introduction ................................................................................................ 58
6.2 Institutional versus administrative catchment area population/Health ........ 58
6.3 Results from the “Monitorage” December 1999 .......................................... 59
6.3.1 Accessibility ........................................................................................ 59
6.3.2 Curative Care Utilisation from within the Yende sub-district ............... 60
6.4 Member and Non-member utilisation rate ................................................. 61
6.5 “Maliando”/ PRIMA ..................................................................................... 62
7. CONCLUSIONS AND RECOMMENDATIONS ................................................ 65
8. ANNEXE ........................................................................................................ 66
Page 5
5
1. EXECUTIVE SUMMARY
With the introduction of the Bamako Initiative in 1988 in Guinea, user fees were
introduced in a strategy of community financing mechanisms. A major problem is
the occurrence of permanent or temporary exclusion from health care. Locally
developed and district based health insurance schemes were considered to
provide a solution to this problem. The health insurance scheme “Maliando”
established in rural Guinea which is concerned by this study has only achieved
very low coverage rates. In a retrospective study using quantitative methods the
outpatient register of a rural health center was assessed to examine the influence
of the factor distance on its utilisation and the implications for the “Maliando”
health insurance scheme. To provide the basis for utilisation rate calculations, a
population census comprising 19.961 persons living in 2834 households was
conducted. Precise utilisation rate calculations for concentric distance bands were
done and relationships between distance and utilisation rates established. People
living in the 2 km range consulted 4,3 times as much the health center than people
living 6 km away.
Members of the insurance scheme living in a distance of 11-15 km use the health
center half as much as members living close (1-5 km). This may provide the
organizers of the scheme with arguments to introduce gradients in fees (sliding
scales according to distance) for the premium level. General observations about
the problems of “Maliando” were made during community meetings. The difference
in the perceived quality of care was a major reason for misunderstandings
between the health services and its users. There were unrealistic expectations of
the health services, Injections were considered high quality treatment and
standard drugs like Chloroquine, Cotrimoxazol and Aspirin were not appreciated.
People perceive the variety of drugs offered at the health center as too restrictive.
Considerable efforts by the promoters of the scheme have to be done to overcome
the misconceptions, that providers and users speak a common language.
Page 6
6
2. INTRODUCTION
2.1 Background of the study
In the framework of health system reform in Guinea, a major emphasis in the
attempts to improve health system performance has been to strengthen the health
information system (HIS). HIS capacity has been strengthened on the national and
the district level. Important progress has been achieved. Epidemiological and
service data are regularily compiled into reports and the results published in
regional and national sanitary reports. Clinicians in first line health services spend
a considerable time on data collection and report writing. The reports consist of
morbidity and health facility attendance as well as performance statistics.
Routinely collected service statistics give information on how many people were
served. The translation of the absolute figures into population based rates is done
by relating the figures to the population served, the clinic catchment area. These
data of utilisation and coverage, even if somewhat imprecise, are important for
comparative and monitoring purposes.
The collection of data of health service utilisation and coverage is useful to give
information on the prevalence of health problems and to monitor the effectiveness
and efficiency of health facilities. Low coverage may point to the existence of a
problem and may give rise for interventions to reduce unmet needs.
Doubts have been raised regarding the accuracy of the data, especially on
utilisation and coverage, for the following reasons:
1. The clinic catchment area is usually defined as the administrative area.
The institutional catchment area often does not correlate with the
administrative division (mismatch).
2. The data from large national surveys are often outdated and of dubious quality,
because the population often does not give true figures in a census for fear that
it may be used for tax purposes. In other occasions before elections the
numbers may be inflated.
Page 7
7
3. The numbers are only rarely updated at sub-district level and population
density figures are mostly unavailable at small administrative units. It is
therefore difficult for local health staff to produce reasonable accurate
estimates. It would often require a basic census to provide denominator data to
calculate coverage.
These problems with data from the existing HIS all add up so that the actual
figures presented may far divert from the truth and distort performance statistics.
This study intends to provide reliable data of the population living in the radius of
15 km and tries to investigate through the assessment of the patient register from
which area surrounding the rural health center Yende patients are drawn. The
utilisation of the different services of the HC is investigated and the effect of
distance demonstrated. This health center is the first line health care provider for
the district based health insurance scheme ”Maliando“.
2.2 The context of the health insurance scheme “Maliando”
After the introduction of community financing mechanisms with the Bamako
Initiative (BI), 1988 in Guinea, it was observed that certain groups within the
population were excluded from medical care because they could not afford to pay
the treatment fees at health center level. Another group delays treatment (in
average 2,75 days) when money has to be searched for borrowing (Roque, 1995).
At least 30% face temporary exclusion in the third trimester of the year as rural
household’s during that time face an important drop in income. 4 % of the
population, the indigents, could at no time afford paying for health care.
Excemptions to protect the poor from the full burden of fees were not effective. In
addition it was observed that utilisation rate of the health centers reduced because
of degrading quality of services delivered at health center level (Sylla, 1998).
There was also a growing dissatisfaction amongst the users because of the rigidity
of the system, the “Minimum" Package of Activities, the rigid use of flowcharts and
drugs at the health center’s curative consultation and because of the poor
Page 8
8
perceived quality of the public health care sector services. People increasingly
perceived the existing (public) offer as poor value for their money (Criel, 1999).
To find a solution to the problems, the Ministry of Health Guinea together with the
German Technical Cooperation-Guinea launched the action research project
“PRIMA” (Projet de Recherche sur le Partage des Risques maladies) in
September 1996 in the district (Prefecture) Kissidougou to study the feasibility and
potential of local health insurance schemes in Guinea/ Conakry.
The concept of a Mutual Health Organisation was developed and called MUCAS
(Mutuelle Communautaire d’Aire de Sante). The first MUCAS, called ““Maliando””
(Solidarity), started in 1998 in the sub-district (Sous-Prefecture) Yende. The
scheme targets the catchment area of the Yende health centre and covers
medical treatment at the primary level and at secondary level care for specified
surgical conditions (Cesarean Section, incarcerated hernia, extra-uterine
pregnancy, operable uterine tumor and appendectomy) and pediatric inpatient
care in two contracted hospitals including transport fees. ““Maliando”” takes as well
free of charge the care of a limited number of indigents.
Members of the ““Maliando”” insurance scheme once a year pay a premium
depending on family size, which covers for all illness episodes experienced by
family members for treatment at the health center including drugs. In addition a
small amount (“ticket moderateur”) has to be paid to reduce abuse (Sylla, 1998).
After two completed years of functioning the MUCAS ““Maliando”” has reached a
critical point because the membership rate had decreased dramatically. Having
dropped from 1352 to 1014 adherents, less than 600 people enlisted again. The
following observations were made during community meetings for the possible
reasons of the low acceptance of the health insurance scheme. In summary the
main reasons are the following:
The insurance premium is considerd as too high
There were unrealistic expectations of the health services, injections are
considered high quality treatment, “Strong medicine“, common drugs like AAS,
Page 9
9
Chloroquine, and Cotrimoxazole are not appreciated and considered low
quality treatment
People do not value the comprehensive services offered at the health center,
the medical act is seen as dispensing drugs and people compare the quantity
of drugs and the price they pay for it at the health center with the cost of the
same drugs if purchased directly at the street vendor
There was a lack of understanding and comprehension of the basic elements
of health insurance to prevent abuse and reduce moral hazards, the main
advantages of health insurance were insufficiently known
This study hopes as well to contribute to the reasons and possible solutions to the
low acceptance of the insurance scheme.
2.3 Guinea: Geography and socio-demograhic data
The Republic of Guinea is a tropical West African country of approximately 7.52
million inhabitants with a total land area of 246,000 km2. Guinea is located on the
Atlantic coast between Guinea- Bissau and Sierra Leone and borders inland with
Senegal, Mali, Ivory Coast and Liberia. Guinea is divided into four natural regions
(Lower, Middle, Upper and Forest Guinea) and/but administratively into seven
administrative regions with 33 Districts (“Prefectures”).
The capital Conakry has approximately 1.5 million inhabitants and contains over
20 % of the national population.
Important progress has been achieved in the health sector through the introduction
of Bamako initiative (BI) policies since 1988. Demographic and health indicators
improved. From 1985 to 1999 the following indicators have reduced: crude birth
rate from 47 to 41, crude death rate from 23 to 17, total fertility from 6 to 5.5,
infant mortality rate from 162 to 137 (1997), Under Five mortality rate from 259 to
215 (1995). (MSP, 1997). The life expectancy at birth is 47 years and the average
number of children per women is 5.7 (MSP, 1997). 47 % of the population are
under the age of fifteen. 18 percent of Guinean’s are under five years old (CIHI,
Page 10
10
1995). The population consists of various major ethnic groups (last data of from
1965) including the Fulanis, Peuls (40%), Malinke (Mandingos, 35%), Soussou,
Kissiens, and other “forest ethnicity’s" . An estimated 85 % of Guinean’s are
Muslim just 5% are Christians, mainly Roman Catholics.
2.4 History
Guinea gained independence from its colonial power France in 1958 in a
referendum rejecting Charle de Gaulle‘s offer to be member of the French
community . During 26 years Guinea was ruled as a one party socialist state by
the President Sekou Toure. In 1993 after an interim military government the
country held it‘s first multi parti presidential election. Guinea has experienced a
massive influx of refugees from Liberia (in 1996 an estimated 600.000) and from
Sierra Leone (200.000) due to civil wars in these countries (EIU, 1995).
2.5 Socio-economic situation
Although Guinea is generally considered to be one of the world's poorest nations,
the country's level of gross national product (GNP) per-capita was about $460 in
1991, above the median for sub-Saharan nations. The rural sector employs 80%
of the working population but produces only 30% of the GDP and a mere 4% of all
exported goods (Marx, 1995).
Mining is the major source of national income and foreign exchange revenue.
Guinea is the world's second-largest producer of bauxite, which accounted for 60
percent of export value in 1991, and also has significant reserves of diamonds,
gold, iron and uranium. Bauxite and aluminum production have declined drastically
in the 1990s. Traditionally the major source of public finances, mining revenues
dropped from over two-thirds of total government revenues in 1989 to just 30 % in
1994. Since the mid 1980s the government liberalized its highly regulatory
economic policies which resulted in some macroeconomic improvements. The
informal sector is expanding but the formal sector continues to lag behind. (EIU,
1995)
Page 11
11
In the UNDP's Human Development Index, a composite score based on various
indicators of education, health and economic conditions, Guinea finished last for
1992. One quarter of all live born fail to reach the age of five.
Only about thirty percent of adult Guinean’s are literate, including just 19 % of
adult women. The World Bank estimates that in 1991 only 37% of the nation’s
school age population were enrolled in primary school while the rate for secondary
education was just 10% (CIDI, 1995).
2.6 Health sector analysis
2.6.1 The Guinean Health system
The 1997 budget of the health sector was in 51 Billion Guinean Francs (approx.
42.420.000 US$) which made up 8% of the state budget. 57% of the health
budget is provided by donors (MSP, 1997).
The organisation of Public Sector Health Services is decentralized. A medical
officer coordinates the activities at the regional level (Inspection Regionale de la
Santé; IRS). In each of the country's prefectures, a director of health (Directeur
Prefectorale de la Santé, DPS) supports PHC programs in health centers and is
assisted by the head of the district hospital. The officers in charge of health
centers are responsible for supervising personnel at the sub-district level. Public
health facilities in 1997 included two universities, 7 regional hospitals, 26 district,
hospitals 349 health centers, 298 postes de santé (MSP, 1997). The distribution of
health centers follows the administrative division, one Health Center (HC) per Sub-
District. For more distant zones Health Posts (HP) exist.
2.6.2 Bamako Initiative (BI) in Guinea
The government of Guinea was one of the first to adopt the Bamako initiative (BI)
in 1987, with strong support from UNICEF. The model centered around community
co-management, cost sharing and decentralization of the health sector, as means
to the broader objective of universal access to primary health care. Guinea
together with Benin were the core PHC countries in this strategy to ensure
Page 12
12
sustainability of health services, improve their effectiveness in terms of coverage
of essential care, minimize operating costs and develop viable financing schemes.
The other major aspects of health system reform were the creation of district
health teams, the decentralization of health services together with the involvement
of the communities in the health service administration and an essential drug
policy with the establishment of the district pharmacies (“cellule de medicaments
essentielles”).
2.6.3 Implementation of core BI strategies
In a macroeconomic evaluation of Guinea, Camen (1992) stated that Guinea's
national program could claim significant success in creating a PHC "umbrella"
through local participation and financing. Knippenberg (1997), in a review of BI
activities in various African countries, found, that the introduction and
implementation of policies following the recommendations of the Bamako Initiative
has improved accessibility, utilisation and coverage of basic health. He states that
the major remaining problem is the equity issue. The exclusion of certain elements
of the population from curative care for financial reasons poses as well for Criel
(1998) a significant problem. However exclusion from curative care is not
associated with exclusion from preventive care (Knippenberg, 1997).
2.6.3.1 Rationalization of Service delivery
The minimum care packages (MCP) focusses on most cost-effective health
interventions (Primary curative consultation; CPC, Expanded Program of
vaccination; EPI, Pre- and post-natal consultation and Family Planning at health
center level). This includes outreach activities for EPI and Antenatal care to ensure
better access to preventive components of MCP. Continuity of care is ensured
through defaulter tracking.
Essential drugs (generics) are available at affordable prices at PHC level through
the establishment of district pharmacies.
Page 13
13
To rationalize service delivery standardized diagnostic and treatment flow charts
were developed. Flow charts help to reduce costs by limiting over-prescription and
facilitate the management of drugs and equipment and facilitates supervisory
control of drug consumption.
Funds generated by community financing are locally retained and managed by
health professionals in collaboration with health management committees.
Community financing is able to recover a substantial part (up to 50%) of recurrent
expenditures of primary health facilities (Knippenberg, 1997).
The MoH introduced very advanced Health Information and Management
Systems. The National HIMS was completely revised and standardized at the
national level. The health facilities prepare their own budget and regular control
mechanisms are built in the system to check for efficiency and good use of
financial and other resources (MSP, 1994). With monthly reporting forms Monthly
indicators are checked and performance evaluated. Guidelines hame been
elaborated for the use of the flow charts and reporting form. Hardcopy patient
cards for preventive services (EPI, ANC,FP) for health center use and for
distribution to clients (“carnet de sante“) exist.
A core element of the HIS is the six monthly monitoring (“monitorage”) for all the
health structures. The team is made up of the DPS, senior doctors of the district
which meets with the officer in charge of the center and the department heads and
members of the management committees. The objective is to make a critical
review of the activities. For each service percentages or figures for availability,
accessibility, utilisation, adequate and effective coverage are established. Figures
for target population are commonly available.
The monitoring provides local staff and members of the health committee with the
opportunity to assess the community coverage of the minimum care package’s
components, determine scores of performance, to identify problem areas and
potential solutions, to micro-plan, to define budgets, to enhance effective
Page 14
14
management and to monitor current trends and results. Goals for the next
monitoring are set and previously set goals verified.
Once local monitoring is completed and budgets are established in all regional
health centers the results are collated, presented and discussed at higher levels.
2.7 Background Information of the study area
2.7.1 Profile of the natural region „Guinee forestière“
Guinee Guinee forestière covers 22% of the total surface of Guinea and is a fores
area in the South-Eastern extreme of the country. Part of it is mountainous area
(massif Fountain). With a rainy season covering eight months of the year it is the
country‘s source of agro-industrial plantation crops (coffee, tea, cacao, palm oil)
and forest products.
2.7.2 German Agency for Technical Cooperation (GTZ)/ Projet Sante Rurale
The Rural health project (RPH) is a project in the framework of German Guinean
cooperation between the Ministry of Health (MoH) and GTZ.
Since 1983 the GTZ supports the health services in the Prefectures of
Kissidougou, Gueckedou and Faranah covering 600 000 inhabitants and
280 000 refugees from neighbouring Sierra Leone and Liberia. In 1997 two other
prefectures in the subregion, Dinguiraye and Dabola were included in the project
zone. The project supported the district Management team (DHMT) in
implementing the Bamako Initiative for preventive and curative services and the
introduction of “monitorage“ activities. Rural health centers, health posts and 2
district hospitals were constructed or renovated.
A maintenance system and prefectoral Pharmacies (PP) were established in the
districts. More recently technical assistance was more directed to the regional
level to transfer capacity in health financing and (global budgeting, monitoring of
expenses). The activities involved more assistance to refugees prioritizing
reproductive health services. Activities to improve quality were introduced.
Page 15
15
2.7.3 PRIMA (Projet de Recherche sur le Partage des Risques maladies) and
the MUCAS “Maliando”
The technical follow up of PRIMA was entrusted with the NGO Medicus Mundi
Belgium (MMB). Initially GTZ was the main funding agency, additional funding
came later on from the Belgium Agency for Development Co-operation. The
objectives of the PRIMA project were to design and test a model of a community-
based organisation risk sharing in order:
To reduce the phenomenon of financial exclusion through strenghtening of
solidarity mechanisms (permanent and temporary)
To develop more genuine forms of community participation in the management
and supply of district based health services (to have better quality of care)
To create a stable source of revenue for the health services
To create a counter-power (contre pouvoir) to the health services and thereby
a leverage for more quality of health care
To create the basis for a sustainable (local and national) capacity in terms of
promotion of community based health insurance schemes (Criel, 1999)
Several studies about the problems of financial exclusion had been conducted
(Rocque, 1995). A rapid rural appraisal (MARP) study investigated the
populations perception of the existing public health care delivery system, the
people's demands and expectations in terms of health care delivery explored pre-
existing (endogenous) mechanisms of mutual aid (Sylla, 1998). In a retrospective
analysis of existing data from the existing routine Health Information System,
health services utilisation patterns of primary and secondary care institutions were
determined (Criel, 1999).
The MUCAS model is in essence a partnership between the health services and
the population which reconciles two different models important in health financing:
The self-governed “syndicate” model, with the general tendency to subordinate
technical imperatives to the social demands of the users and the “technocratic”
Page 16
16
model (Type HMO) with the tendency to subordinate social needs of the users to
technical considerations expressed by health staff (Sylla, 1998).
2.7.4 District Health Sector with the Health Center Yende
The health center Yende with its two associated health posts Walto and Faindou in
the sub-district Yende is part of a two-tier system of the district Kissidougou
(administrative region Faranah), with 15 health centers, 9 health posts and one
district hospital. The HC Yende is located in the very south of the sub-district.
2.7.5 Health Center Yende
The HC has 15 staff from which 5 have health qualifications. It offers fixed services
(CPC, CPN, PEV, deliveries, FP and TB treatment) outreach activities for CPN,
PEV and laboratory services (Roenne, 2000). In average 680 new patients consult
for CPC monthly. Yende HC is the only first level care institution contracted by the
MUCAS ““Maliando””. The total budget for the health center is around 20.600 US$,
from which 12.800 US$ are state subsidies (salaries, equipment etc) and 7.800
US$ are generated by the HC. 30% of it’s revenue is generated by the contract
with ““Maliando””.
2.8 Purpose of the study
This study will contribute to give answers to the following questions:
From where do people come to obtain medical care at the health center
Yende?
Until which distance from the health center do people still may see benefits
from the insurance scheme?
How does the aspect distance influence the peoples decision to attend or not
to attend the health center or to join or not to join the health insurance
scheme?
Which distance distance represents a real barrier for the household to obtain
medical care?
Page 17
17
As the health center is the first line health care provider for the district based
health insurance scheme „”Maliando”“ the results may be useful for the managers
and organizers of the scheme to have realistic figures about the target population
for the health insurance scheme in order to better target promotional and other
activities. It may as well serve as basis for thoughts weather the attractiveness of
the insurance scheme can be increased for the population living further away from
the health center by offering reduced cheaper insurance packages (e.g. only for
certain services or for catastrophic events).
Possible further uses of the study could be to provide decision makers and health
planners on district and national level with realistic figures about the geographical
accessibility of Yende HC, which could be used for strategic planning.
2.9 Research question
How does the factor distance influence utilization of the health center Yende?
What can be the implications for the finding for the health insurance scheme
“Maliando”?
2.10 Research objectives
2.10.1 General objective
1. To obtain exact figures for the population living in villages surrounding the
health center Yende, as denominator for utilisation rate and coverage
calculations
2. To document the general utilisation pattern from existing service data in
different subgroups of the population (sex, age groups, insured/uninsured)
3. To quantify the utilisation of the HC Yende in relation to distance by analysing
the patient register for the period of one year (June 1999 – Mai 2000)
Page 18
18
4. To determine the group of villages surrounding the HC Yende patients which
are sufficiently using its services (institutional catchment area of the health
center Yende)
5. To compare the figures obtained in the study with existing figures from the
existing health information system (“Monitorage”, Coverage/utilization figures)
2.10.2 Specific objectives
To conduct a population census of villages surrounding HC Yende until 15 km
distance
1. To establish per capita utilisation rates of the villages living within 15 km
distance of the HC Yende
2. To extract from the existing patient register of the health facility information
about utilisation pattern of different subgroups of the population (vulnerable
groups, insured/ uninsured)
3. To establish whether the utilisation pattern is different in age groups and
between the sexes and how the utilisation pattern is different in the insured and
noninsured group
4. To determine how distance affects the utilisation for different illness categories
and services (EPI, antenatal care (ANC), Family Planning)
5. To calculate curative care, immunization, antenatal care and Family Planning
coverage with the newly developed institutional catchment area population
6. To quantify the difference of utilisation and coverage data using administrative
(sub-district Yende) or catchment area population
Page 19
19
3 LITERATURE REVIEW
3.1 Introduction
It seems to be paradox that in developing countries like in Guinea in many
instances the coverage with public health facilities is not sufficient and a large
unmet demand for health exists, but at the same time the existing health facilities
are under-utilized. This review of literature introduces to the basic concepts of
access to health care, tries to show how the decision making process takes place
in a household in a given illness episode and presents the possible options for
health care which are available. It introduces and describes the factors that
influence health care utilisation and especially looks at what hinders patients to
use government health facilities. Concrete average figures for utilisation and
coverage of different services in rural health centers in developing countries are
given to relate the concrete data obtained in the study. Experiences with health
insurance schemes will be mentioned.
3.2 Health service utilisation
Health service utilisation can be expressed as the proportion of people in need of
a service who actually receive it in a given period (Parks, 1995). It can be
measured as the Health service utilisation rate which is a rough indicator and
reflects the availability, accessibility, affordability and acceptability of the service
and the perception of the quality of the activities offered.
Before a person can or will use health services he must perceive a need for them.
He must be aware of his condition, and feel it warrants medical intervention; the
appropriate services must be available (within a reasonable time/ distance); the
services must be acceptable (he must have confidence in the technical
competence and humanness of the facility and its provider); and he must have the
ability to obtain the services (the necessary income and/or insurance and time).
Each of these prerequisites may comprise a barrier for utilisation. The use of
services is therefore not equated with the access to services( Matomora, 1997).
Page 20
20
Access to health services
The technical definition by WHO defines access to health services as the estimate
of the proportion of the population that can reach appropriate local health services
by local means of transport in no more than one hour. Recently WHO has revised
its definition to the proportion of the population having treatment for common
diseases and injuries and a regular supply of the essential drugs on the national
list within one hour’s walk or travel (CIHI, 1995). Access is the “actual use of
personal health services and everything that facilitates or impedes this use“
(Matomora, 1997).
A relationship may exists between utilisation of health services and health needs
and status. Utilisation rates give some indicators of the care needed by the
population and therefore the health status of a population, but it does not fully
explain health care behavior. Even if demand exists and is expressed by a
population there can be important barriers to utilisation by factors such as
availability and accessibility of health services (Parks, 1995).
Accessibility to health services is in practical terms the extend to which the
population in need can use the services, the patients ability to get to the services.
According to the operationalisation model of Canales accessibility to health care
can be divided into three dimensions, geographical, financial and cultural
accessibility (Matomoro, 1997).
Geographical accessibility
Geographical accessibility is defined as road distance and the cost of overcoming
that distance by public transport opposed to institutional accessibility which is
concerned with the question whether a facility is fully available or if social ,
economic and cultural barriers constrain its use (Ayeni, 1987).
Page 21
21
3.3 Health care decision process
The decision what to do in a particular illness episode is an outcome of the
patient’s health care decision making process which is influenced by the attitude of
an individual but often more important by the household’s attitude towards health
and the health care systems (Parks, 1995). Health care behavior does depend to
a large extend on the individuals and the household’s own concepts of health,
disease and healing (Bichmann, 1991). The household production of health
concept looks at the household’s options and choices and their criteria for
decision-making (Bermann, 1994). When symptoms appear the patient may adopt
an active or passive attitude towards the disease. This usually depends upon the
gravity of the illness, socio-cultural variables, the quality of existing services and
by previous experiences with the different structures/forms of care (Matomora,
1997).
If an active attitude is chosen it can be directed towards Family Care, self-
treatment, towards traditional structures or official medical structures. In general
patients during an illness will start by using home remedies and self-medication
relying on the advice of the extended family. Self-care is the first step in the search
for treatment of mild and well known symptoms. Only if this fails the patient turns
for professional advise (Mwabu, 1986).
If the illness is perceived as more serious and if sufficient funds are available there
is a very high likelihood of a patient consulting more than one provider for advice
or treatment. This sequential decision-making process can be repeated several
times. The visit pattern will vary greatly according to type and stage of illness, the
household’s perception about the illness and on its ability to afford a visit. This so
called “healer shopping“ is a characteristic feature of health care choice. It is
based on trial and error. The utilisation of health care facilities depends on the
actual success of the treatment (Hielscher and Sommerfeld, 1985).
Page 22
22
3.3.1 Illness concepts
Most usage is sequential although some is simultaneous. In the latter cases the
patient assumes several differential causational possibilities. Indigenous illness
concepts are unstable and are adopted or discarded in the course of the search
for treatment. If no relief is obtained, then it is considered that the illness has been
wrongly classified. Consequently follows a new classification and new treatment
follows. (Hielscher and Sommerfeld,1985)
3.3.2 Utilisation of traditional medicine
Modern medical practitioners are few in number compared to the total population
of the developing world. Moreover, they are unevenly distributed, concentrated
primarily in the urban areas. Often they are physically inaccessible to many rural
people. Whereas traditional healers are widely available in rural areas. But
widespread utilisation of traditional healers or other indigenous health resources
can not be simply explained by this. Often they are culturally, socially, and
environmentally closer to the people and so more accessible. They are more
sympathetic and often less expensive. They use local resources (herbs, etc), local
technology and local labor.
Even where modern services are well available, traditional healers will continue to
be used. It is the result of the belief that some categories of illness like mental
disorders or bone fractures are believed to be associated with witchcraft and for
which Western medicine will have no effective answer (Paul, 1985).
This belief cuts across educational lines, although it is less predominant in more
“educated” households. In the study by Okafor (1983) between 65 % and 70% of
the less literate and the illiterate households have a positive attitude to “native
medicine”. High prices prevent people from utilizing modern resources. Folk
medicine is usually available to everybody at a low prize. Payment in barter is also
accepted (Hielscher and Sommerfeld, 1985).
Page 23
23
Access to good quality care and medicine free of charge does not ensure that
people seek proper treatment. Only having enough money for hospital or clinic
treatment is no sufficient reason for choosing it. However much the likelyhood of a
patient choosing/getting/paying for effective treatment is affected by the cost and
acceptability of the available options. It depends at least as importantly on other
things happening at the same time and on whether, in the event the patient has
ready access to the combination of resources appropriate to it (Wallmann, 1996).
Shortage of non money resources, time, confidence, someone to mind the
children may be more decisive than the presence of cash. The viability of any
system is limited by the scarcest element. Studies on household costs in Burkina
Faso showed that time costs are often substantial and make up one third of the
costs (Sauerborn and Nougtara, 1996).
3.4. Utilisation in Guinea
3.4.1 Curative care
Household surveys conducted in Guinea and Benin in 1990/91 after the
introduction of an essential drug policy and fees for services with community
financing showed that a large part of the population does not use health centers.
In Guinea the poorer population has few alternative sources of care and often opts
not to seek care. The average number of illness episodes was 1,0 episodes per
person/ year. Between ¼ to 1/3 of the episodes lead to seeking care at public
health centers. The other 1/3 to ½ illness episodes were treated at home, the rest
looked for other sources of care (Soucat and Gandaho, 1997).
Following the revitalization of health centers and the re-establishment of drug
availability, utilisation of curative care initially in Guinea increased significantly and
stabilized at around 0,3 visits per capita (1997: 0,26).
Page 24
24
3.4.2 Antenatal care and EPI
After the introduction of BI policies utilisation of prenatal care went from less than
5% in 1987 to almost 55% in 1993 (1997; 59%). 60% of pregnant women
benefited from at least one antenatal visit. In 1997 the average number of contacts
per antenatal visit was 2.52.
Tetanus Toxoid treatment of pregnant women increased progressively from 6% in
1986 to 66% in 1995. In 1997 the average number of contacts was 2.52. 18% of
deliveries are done at health center level. Half of all deliveries take place in the
home, unattended by a professional midwife. Traditional birth attendants deliver
43% of the children (Soucat and Gandaho, 1997) , (MSP, 1997).
1995 after seven years of program implementation, EPI coverage (the proportion
of children having received DPT3 before the age of one), reached 73% (from less
than 5% in 1986).
3.5. Utilization related to distance (geographical accessibility)
Health care facilities in third world countries serve a relatively small area
surrounding the facility. As most patients travel on foot and few have access to
motorized transport, distance significantly affects the utilisation of health services.
Patients living some distance away from a health care facility tend to delay using
its services, usually preferring alternate forms of treatment, including self-
treatment with traditional or commercially available medicines (Airey, 1989). The
major users of the services are the population immediately living around health
facilities. The proportion of people from larger distances declines exponentially.
According to the principle of economic rationality consumers will look for care in
the nearest health care facility in order to minimise route distances and so reduce
costs (Egunjobi, 1983).
The distance separating potential patients from the nearest health facility is an
important barrier to its use, particularly in the rural areas. The greater the distance
the less services are used. Long distances can be actual obstacles to reaching
Page 25
25
health facilities and they can be an disincentive even to trying to seek care. In
developing countries, the effect of distance of service use becomes stronger when
combined with the lack of transportation and with poor roads, which contribute
towards indirect costs of visits (Noorali and Luby, 1999).
In a study by Bichmann (1985) in Benin it was found that the majority of users
(60 %) come from the vicinity (less than 5 km) although only (28%) of the total
district population lived within the distance. On the other hand 20% take the
trouble to walk long distances to attend outpatient clinics. The study by Egunjobi,
examining factors influencing the choice of hospitals in Nigeria showed that
although proximity was the leading factor in consumers choice (32%), other factors
like quality of service (25%), relative living in hospital town (15%), Finance (12%),
Ease of transport (11%), and others (5%) play a major role. The rate of distance
decay in utilisation levels varies according to the type of facility, socio-
demographic variables and illness. Travel by younger children depends on the
availability of an adult or elder sibling to escort or carry the child. Elderly patients
are liable to find long distances a significant if not an absolute barrier to obtaining
treatment ( Stock, 1983)
In the case of underutilisation of a clinic when the quality of the services provided
is perceived as low, it is not the inability to travel that prevents the use of health
clinics but the expectation of receiving unsatisfactory service. By increasing the
quality of the services people will “go the extra mile” and travel more for the
services and utilisation rate will increase.
3.6. Utilization related to costs (financial accessibility)
3.6.1. Household expenditure for health Guinea
With an average fee level of, on average 1,1 US$ (Other BI projects Africa;
average 1,5) or US$ 0,19 – 0,56 p/y the household expenditure was on average 4
US$ per year. This would permit 3,6 full treatments per year. The out of pocket
Page 26
26
expenditure on health care is probably much higher because of travel costs and
under the table payments (Knippenberg, 1997).
Hotchkiss (1998) found in Nepal that in rural areas 74% of modern treatment
expenditures is spent on pharmaceuticals, (86%) It is higher in urban areas. The
vast majority of household expenditures are used for drug purchase on the private
market. Health care accounted for 5,5% of total household expenditure. The share
of total expenditure devoted to health increases with the level of household
income. Rural households spend more on health care than urban households after
controlling for income status.
3.6.2 Costs incurred of overcoming distance
Another variable which has been extensively examined in the literature is financial
cost for receiving care. Cost and distance often go hand in hand as longer
distances entail higher transportation costs. The demand for a health institution
should decline as costs of reaching it increases. The cost to obtain medical care
can be divided into two categories and make up the overall costs
(Airey, 1989):
Direct costs which include facility fees, the cost of medications and other
supplies.
Indirect costs which include transport , other costs for food and lodging and
the cost of time otherwise spent productively (opportunity costs).
Additional costs incur when drugs are unavailable at the health facility and these
have to be purchased at private pharmacies “Under-the-table- payments” are
frequent and add to the costs. In the case of inpatient treatment, indirect costs
such as transport expenses, including the return ticket, daily expenses for food
and to lodging of family members in the town of the hospital may be significantly
higher than the direct costs. Prospective patients, especially women, do not travel
alone to a health facility. They are accompagnied by other adults and by children
who cannot be left at home alone, because caretakers are not available. All the
additional people swell the cost for transport (Thaddeus and Maine, 1994).
Page 27
27
The opportunity costs of the time used to seek health services, the time spent
getting to, waiting for and receiving health services is lost from other, more
productive activities such as farming, fetching water or wood for fuel, herding,
trading, cooking and so on. Especially women carry out a large majority of these
tasks, the value of their time and the competing demands on it are major
determinants (Thaddeus and Maine, 1994). The cost of time otherwise spent
productively, e.g. the loss of agricultural working time, is particularily critical during
planting and harvest time. Especially people of lower socioeconomic classes can
often not afford any loss in income.
The effect of cost on service utilisation is commonly assessed through interviews
and surveys of users and non-users in which respondents are asked to give
reasons for their choice of actions when they are ill. A proportion of these
respondents give financial constraints as a major reason for not seeking care, or
for seeking one form of care rather than another.
But many studies indicate that the financial costs for receiving care is not a major
determinant of the decision to seek care (Thaddeus and Maine, 1994).
A survey conducted in Nigeria revealed five factors that influenced people‘s
decision to seek traditional or western medical care: Respondents ranked cost and
distance fourth and fifth (Thaddeus and Maine, 1994). Kloos (1990) reported that
in Ethiopia, cost of services was often a less important consideration in utilisation
than were quality of services and perceived efficacy of treatment. The study by
Egunjobi (1983) examining factors influencing the choice of hospitals in Nigeria
showed that although proximity was the leading factor in consumers choice (32%),
finance was at fourth place with 15%, among other factors like quality of service
(25%), relative living in hospital town (15%) and ease of transport (11%). Cost in
contrast is most likely to affect compliance with prescribed treatment as the cost of
medications is often very high (Thaddeus and Maine, 1994).
The cost/ benefit ration of using medical services may be viewed differently in
emergency cases.
Page 28
28
3.7 Utilization related to Quality of care
Examining the literature of quality of care in the primary health care setting often
poor quality of consultation is observed with a high proportion of prescription
problems. Consultations largely center around the prescription of a drug and
ignore the wider requirements of good medical practice. The process of diagnosis
is often too poor to ensure correct diagnosis ( limited training and skills of most
staff). Often although some aspects of the history of the complaint are reviewed,
little examination is undertaken. A high proportion of the diagnoses are identified
as symptoms or other diagnoses.
A study conducted in rural Burkina Faso by Krause et al (1998) showed that in
only 20% of the diagnoses the nurses took a sufficient history and in only 40%
they conducted a sufficient clinical examination. Only 12 % of the diagnosis was
based on sufficient history taking and adequate clinical examination. The results
were low diagnostic quality, dissatisfaction of the population with the health care
services offered and low utilisation rates.
3.7.1 Perceived quality and reliability
Quality and reliability are critical for the attractiveness of services. If patients have
serious doubts that they will obtain the services they desire or they do not trust
that a hospital provides an adequate backup to their local primary care provider
(Stock, 1983), it is unlikely they see advantages for themselves and their family.
People hesitate to travel several kilometres to a dispensary for treatment when
they know that it is often closed and the supply of injectable drugs or other drugs
are often exhausted (Stock, 1983). The experience with the introduction of user
fees showed that higher costs of the provision of health services associated with
quality improvements did effectifely increase utilisation of the services.
Haddad and Fournier (1998) studied the criteria for quality of health services as
determinants for utilisation , perceived by the population in Guinea, using focus
group discussions. Open discussions of what the participants considered as
Page 29
29
important aspects of good medical care were conducted and the identified criterias
listed hierarcically under five categories, based on how frequently they were
mentioned.
1.Technical competence of the health care personnel, 2. Interpersonal relations
between the patients and health care provider, 3. Availability and adequacy of
resources and services, 4. Accessibility, 5. Effectivness of care.
3.7.2 Donabedian’s classification of quality of care
The most widely used classification of quality of care was developed by
Donabedian (1980, 1988) who developed three major categories:
1. Structure denotes the attributes of the setting in which care occurs; Material
resources (facilities, equipment, money), human resources (number and
qualification of personnel), organisational structure (medical staff organisation,
peer review, methods of reimbursement).
2. Process denotes what is actually done in giving and receiving care. It includes
the patient activities in seeking care and carrying it out as well as the
practitioner’s activities in making a diagnosis and recommending or
implementing treatment.
3. Outcome denotes the effects of care on the health status of patients and
populations, health status (Improvements in the patient’s knowledge and
salutary changes in the patient’s behaviour, degree of patient’s satisfaction with
care)
Donabedians category of process was devided into the components technical
competence and interpersonal relations. Eight criteria were greately valued. Most
of them relate to the structure (availability of drugs and accessibility of the health
facility) or the process of care (reception of patients, overall care, good clinical
examination, dispensing drugs and/ or good drugs). Recovery of health is the most
Page 30
30
frequently encountered, suggesting that quality of health services is first and
foremost judged in terms of outcomes. Little emphasis is placed on preventive
services.
A previous study by Haddad and Fournier (1995) regarding quality, cost and
utilisation of health services conducted in Zaire had simular results and showed
that the top qualitiy mentioned in semi-structured intervieuws of villagers were the
nurse’s interpersonal qualities (respect, patience, courtesy, attentiveness,
friendlyness and straightforwardness). Technical qualities (good treatment, good
work, good diagnosis and punctuality) came only second to interpersonal qualities.
The availability of drugs was as well given high priority. Technical competence is
often overestimated in relation to good behaviour.
Increasingly it has been acknowledged that the technocratic perspective of what is
good quality of care defined by health professionals does not necessarily
correspond with the perceived quality of care by the population. Any study on
quality of care has to investigate user satisfaction, which is the patient’s judgement
on the quality and goodness of care (Donabedian, 1980). This is associated with
improved compliance and as well health improvement
Users appear very sensitive to aspects of interpersonal relations they have with
professionals as well as the technical quality of the care provided (Winefield H,
1995).
3.8. Mutual health insurance schemes
As stated in the introduction health insurance schemes are considered to be a one
of the solutions for the problem of financial exclusion. Formal health insurance
systems as concluded in an overview of 23 sub-saharan countries by Vogel, have
not promoted greater equity in access to health services by the poor. Locally
developed and district based schemes targeting poor rural self-employed
Page 31
31
populations are considered to be more successful. The problem is that there is
still very little analytical information available about such schemes (Criel, 1997).
Bennet (1997) who examined 80 insurance schemes, concluded that although
some risk sharing schemes are clearly successful, a number of common failings is
evident. Schemes in low-income countries have generally only achieved limited
population coverage. With a few exceptions, cost recovery rates are low, the
schemes examined by her had very little ability to protect the poorest parts of the
society. She observed that many schemes suffer of a poor design. She noted that
very few schemes appear to have used graduated premiums according to income
or household location or exemption mechanisms. She recommended to make
greater use of these mechanisms.
It was mentioned in the study by Yi (1998), that villagers of remote rural
communities in Ghana perceived the difficulty of transport as the greatest
obstacle of establishing health insurance schemes in those areas.
Noterman (1995) observed in Zaire that the payment of the same subscription fee
for households living far away from the hospital penalized them. In fact these
households actually subsidize the hospital care of households living close to the
hospital, who use it more frequently.
The further the people live from the hospital the higher are the indirect costs and
the higher the opportunity costs of an admission. Noterman proposed the
implementation of gradients in fees (sliding scales according to distance) . This
was tried out in Bwamanda in Zaire by setting different co-payment level according
to distance from the health center to the hospital. It was stopped, however, after
one year because there was no positive impact on the hospital admission rates of
the more remote insured population and because of the more complex
managemet and control procedures required (Criel and Kegels, 1997).
Atim, (1998) in a synthesis of research in nine west African countries gave the
following recommendations concerning design features that enhance the schemes
success: A mandatory reference or gatekeeper system, a requirement for
compulsary participation or at least mandatory family membership, a waiting
Page 32
32
period for new members, the use of efficient provider mechanisms, the inclusion of
essential and generic drug policies in their agreements with providers as well as
preventive and promotive services in the benefits packages.
For promoters the major recommendation were to reinforce the institutional,
managerial, and administrative capacities of the health insurance schemes in such
areas as setting up adequate HIS systems, setting premiums, managing funds,
pricing and assessing the quality of care. The organizers of the scheme need
information on health spending and utilization and risk patterns to be able to set
premiums at levels that would be self financing (Atim, 1997).
Kutzin (1997) stated that a scheme should be designed in a way that keeps
premiums low. He thinks that it is important that the services covered by the
insurance scheme initially focus on relatively high cost low frequency events. Atim
(1997) stressed also the fact that the schemes have more success if they are set
up around a health provider with a good reputation for good quality in terms of
waiting time, staff attitude towards patients and drug availability. This is confirmed
by other authors. Clinic management, staff quality and morale, drug supply and
relations with the community as a whole are probably more important influences
on utilisation than a payment scheme (Soucat et 1997).
Criel (1999) stated that the major determinants for high subscription rates are the
general performance of a health facility as well as the quality of its interaction with
the community.
Gilson (1997) summed up her finding after examining the literature about
experiences after the introduction of user fees. Risk sharing schemes should not
be seen as a source of finance but rather as ways of organizing health service
financing and delivery. Establishing a health insurance scheme may enable to
introduce organizational changes such as tighter referral control, contracting
arrangements between purchasers, accreditation and service quality improvement,
and performance related pay. Insurance should be seen as a supporting strategy,
not an exclusive ”financing alternative” .
Page 33
33
3.8.1 Data on utilisation rates of “Maliando” insurance members in previous
studies
In an assessment by Diallo (2000) of records of attendances of “Maliando” health
insurance members a total number of 1796 patient contacts within the period June
1999 to Mai 2000 by 1014 “Maliando” members were reviewed. The patients came
from 21 villages from the UNEM Mata (subunit of the health insurance scheme)
and from 11 and 5 villages of the UNEM Touffoudou and Yende. The calculated
total Utilisation rate for “Maliando” members was 2,08. In the document villages in
the 5 km radius were differentiated from villages beyond the 5 km range.
Utilisation rates were 2,33 and 1,54.
Hohmann mentioned in the evaluation of PRIMA (1999) that on the basis of an
asessment of records from August – November 1998 the average utilisation rate
for “Maliando” members were 1,68 contacts per year. Range 0-5 km: 2,0
5-15 km: 1,24. In the study by von Roenne (2000) the global utilisation rate for the
total population is 0.56 with large differences between the insured group (1.9) and
the uninsured group (0.47). Members of the insurance scheme use the HC 4 times
as much as non-members but represent only 7% of all patient contacts.
3.9 Methodological aspects
3.9.1 Quantitative techniques used in the study
Spatial Analysis
The impact of spatial factors on health care behavior has been major themes in
medical geographical research. It has been general practice to explain consumer
behaviour patterns in terms of spatial accessibility (Stock, 1983)
Page 34
34
For utilization studies a variety of distance measures can be employed. Distance
can be measured as a straight-line joining the point of origin and the destination,
or the actual measured length of journey. Travel costs or travel time may be used
as frictional effects of distance (Stock, 1983).
The functional distance to a facility may be greater than the physical distance to
the closest facility. The simple Euclidean distance (physical distance between the
patients homes and the health center) is a sub-optimal measure of distance as it
ignores physical barriers such as rivers, swamps, hills, the road and traffic system
and socio-cultural factors (Mueller and Smith, 1998). Differences in physical
terrain, availability of public or private transport, and patients access to alternative
forms of transport (motor vehicles, bicycle and foot) all affect the functional
distance. Moreover such functional distances may change frequently particularly
seasonally like in the rainy season (Ayeni, 1987).
3.9.1.2 Per capita utilisation rate
To measure utilisation by stating or comparing total numbers of attending patients
is not adequate as it does not take the population distribution into account. Per
capita utilisation rates are a much more sensitive indicator of the level of use of a
health facility. It relates the number of patients from each sublocation of a health
facility to its general population (the denominator).
Stock related the per capita utilisation rate to the factor distance in a study of rural
health facilities in Nigeria by calculating the per capita utilization rates for
concentric distance bands around each health facility. These bands had a width of
2 km up to 10 km from the center and 5 km thereafter (Stock, 1983).
In a further step the decline of the per capita utilisation rate (% decline/ km) can be
calculated. The same method was used in a study by Osibogun (1998), who
investigated the outpatient register of a rural health center in Nigeria.
Page 35
35
He compared the percentage of users coming from the vicinity (<5km) with the
percentage of the population which lives within the 5 km range (Utilisation
differentials).
In a study by Criel (1999) the impact of the Bwamanda hospital insurance scheme
on the hospital utilization pattern was studied. Health center areas in relation to the
referral hospital were chosen. In the study the total admission rates per health
center area were determined, stratified by insurance status. Utilisation differentials
between the insured and non-insured population were established.
5.9.1.3 Concept of distance decay
A dominant theme in studies of health service utilization is the concept of distance
decay. The rate of interaction (utilization) with the health services tends to vary
inversely with distance. Distance decay describes the relationship between
distance and the rate of utilisation as a negative exponential function.
The rate of distance decay in utilisation levels will vary according to the type of
facility, socio-demographic variables and illness and can be established for these
variableS (age, sex, diagnosis, symptoms,diagnosis, services used be established
(Stock, 1983)
5.9.1.4 Population Census
To calculate the per capita utilisation rate it is necessary to have a good estimate
or better exact figures about the population (denominator data).
For this purpose an estimated population is usually derived from updating older
census data by extrapolation, adding the annual population growth rate (Airey,
1989).
In practice this procedure poses problems as the data from large national surveys
are often outdated and of dubious quality. They are only rarely updated at sub-
district level and population density figures are mostly unavailable at small
administrative units. It is therefore difficult to produce reasonable accurate
Page 36
36
estimates. It would often require a basic census to provide denominator data to
calculate coverage (Satia, 1994)
The planning of a household census is similar to the planning and implementation
of household survey. In the training modules for Household Surveys by WHO
(1988) the aspects good selection of surveyors, adequate training time and pre-
testing of the forms is considered essential. For high quality data good field
supervision is a crucial aspect. Quality control measures have to be implemented
at all stages of the survey.
5.9.1.5 Global Positioning System (GPS)
During the study a GPS system was used to help to establish physical distances.
The GPS system uses earth orbiting satellites, which regularly transmit precisely
timed signals. These are received by a special electronic device which provides
direct measurement of the position on the earth’s surface. When the GPS device
is activated the site is located and defined in longitudinal and latitudinal
coordinates. The GPS system can then calculate distances between recorded
locations (UNFPA, 1996).
5.9.1.6 Catchment area population
A catchment area should ideally comprise a population of 10.000 per primary
health care unit. It is the geographical area from which the majority of its patients
of a health unit is drawn (Kloos, 1990).
This can be the basis for the planning for the establishment of health centers.
Pangu, (1988) documented this when examined the planification strategy in Zaire.
The limits of the planned health centers were traced starting from the peripherie of
the reference hospital, aiming basically to allocate a population of 10 000 to a
health center area taking into account infrastructure conditions.
The “Monitorage” guidelines for health centers of the Ministry of Health Guinea
states that curative care and preventive services are only accessible for the
Page 37
37
population living within a radius of 10 km. For health posts the same distance
applies for curative care applies but for preventive services like EPI, ANC and
Family Planning it is only the population living in the 5km zone that have access to
care. The catchment area as defined by the Ministry of Health concerning
preventive services as well includes the population of villages where effective
outreach activities (strategie efficace) had been conducted . Effective outreach is
that at least 3 ANC and EPI sessions were organized in the 6 month monitoring
period (MSP, 1994). This is reflected in the denominator, the population served,
for accessibility calculations.
Other indicators used in the monitorage are availability, effective coverage
(examining quality of the services rendered) and management indicators (finance ,
utilisation of drugs ).
To determine the catchment area of a health facility with a rapid assessment
method, Nordberg (1993) asked health staff to fill out questionnaires, to identify
on a map the areas from where the majority of its patients are drawn and to
estimate the total number of people living in the area.
Criel (1996) developed a data collection tool which aimed to define the catchment
areas. The exercise consisted of monitoring the utilization of the curative services
of all existing health facilities during a period long enough to control for seasonal
variations. Curative care was considered a good proxy for the overall utilisation
pattern of a health facility. Detailed maps were drawn and excact population
figures established. After compiling and mapping, each of the villages was
assigned to a given facility. In this methodology the catchment area is the actual
used area, determined by the patients preferences
5.9.1.7 Coverage/ utilisation data
In the following the methodology is explained how coverage and utilisation rates
are calculated during the monitoring (“monitorage”) sessions (MSP, 1994), (MSP,
1999) The figures for the monitoring of December 1999 for the six month period
Page 38
38
from June to November are given in order to be later compared in the results
section with the data from the current study (Diallo, 1999).
The monitoring team based its calculation on the following population figures: The
target population of the whole sub-district Yende is 17.319. The sub-district
population figures are yearly updated from older census data by multiplying the
previous year population with the factor 1,027. For the 6 monthly monitoring the
total target population was divided by the factor 2.
5.9.1.8 Calculating accessibility
The accessible population for the health center Yende combined with two
associated heath posts was considered to be 12751. This figure is the population
which is located in the proximity of 10 km of the HC and the two HP. Relating the
figure 12.751 to the total sub-district population of 17.319 gave the figure of 74 %
for accessibily. Taking this into consideration the monitoring team based its
calculation on the figure of 15.660 (accessibility: 90%).
5.9.1.9 Calculating Utilisation
Curative Care (CPC)
The health center during that period had 4.135 attendances and the health posts
combined 892 which gave the total figure of 5.027. Related to the whole
population of the sub-district Yende( 17.319 / 2 = 8.6595) gave the figure of 58%
(= 0,58 utilisation rate)
Antenatal care (ANC)
The target population of pregnant women (expected deliveries) is calculated
through the formula target population/ 2 (sub-district Yende) x the factor 0,045
which resulted into the figure of 390. 381 pregnant women had received antenatal
services during the reporting period (Coverage of 99%).
Page 39
39
Adequate Coverage is the percentage of women with at least 3 ANC contacts and
2 Tetanus Toxoid vaccinations. The figure indicated in the report is 96%.
The effective coverage was 83 %.
Expanded program of immunisation (EPI)
The target population is the same as in antenatal care. Utilisation calculations are
based on the formula Target population / 2 x 0,040 (346). The total number of
children vaccinated in the reporting period were 344 (utilisation 99%). Adequate
coverage is the percentage of children who have received all the required
vaccinations (4 sessions) before reaching the age of one (96%). The effective
Coverage was 92 %.
Family Planning (FP)
The target population for Family Planning services are women in reproductive age
derived from the formula Target popultion x 0,060 which resulted in the figure of
1039. The ANC department had 120 clients (Utilisation rate 12 %).
Adequate and effective coverage were both 6 %.
3.9.2 Qualitative methods
3.9.2.1 Combination of qualitative and quantitative approaches
Quantitative methods like the ones applied in the current study can be
complemented with qualitative research (Focus Group Discussions, In depth -
Interviews, case studies, observational fieldwork) in order to find explanations for
unusual unexpected results. The most commonly advanced reasons for combining
the 2 methods are for different stages of a project, to compensate for each other’s
shortcomings and for the purposes of triangulation. Qualitative methods may be
used to develop sophisticated quantitative research tools. Qualitative work might
also follow a large scale survey to explore further the mechanisms by which
variables are connected.
Page 40
40
The logic of this combination of different methods (Triangulation) is based on the
premise “that no single method adequately solves the problem of rival causal
factors. Because each method reveals different aspects of empirical reality,
multiple methods for observations must be employed”. (Patton, 1989) “The
findings from alternative sources enable researchers to make more subtle and
sophisticated analysis: any marked differences can be highlighted, investigated
and explained “ (Dowell, 1995)
It has been recognized that qualitative approaches can enhance quantitative
studies in four ways; by providing insight into the process of data construction, by
helping to identify the relevant variables for a study, by furnishing explanations for
unexpected and anomalous findings and by generating hypotheses or research
questions for further investigation (Barbour, 1999)
Page 41
41
4 METHODOLOGY
4.1 Study area description and study population
The catchment area population (15 km radius) was estimated to comprise 3000
households living in 53 sectors/ villages with the total of 17.250 inhabitants
(Roenne, 2000). A private health center is situated at 2 km distance from Yende in
the district Gueckedou, other health centers are in more than 20 km distance.
4.2 Time frame
The study was carried out from the May 3rd to July 1st 2000. The first 4 weeks
were used for preparation and further adaptation of the research tools. Data
collection was carried out in the whole of June. Part of the data analysis was
done in the last week of June, but the major part after returning to the home
country.
4.3 Study type and design
This is a retrospective study using quantitative methods. Information from personal
communication and quantitative data collected by other researchers was used to
complement and enrich the study. It was organised in different steps using 3
different methods of data collection.
1. Census of the population living in 15 km proximity of the HC Yende
2 Assessment of the outpatient register for first contact Curative Care
Consultation
3. Assessment of patient cards for preventive services (ANC, EPI, and Family
Planning) kept at health center level
4.4 Finance and logistics.
For the population census one 4 wheel drive vehicle was provided by PRIMA, 2
motorcycles by PLAN Guinee, a NGO operating in Yende subdistrict as well as
from the EPI department of the health center Yende. The budget was provided by
Page 42
42
GTZ/ PSR. Two computers of PRIMA were used for data entry for the review of
the outpatient register.
4.5 Data collection
4.5.1 Data collecting tools
4.5.1.1 Population Census
The total number 20.002 persons were counted in 2844 households in the area of
15 km surrounding Yende. For data collection 3 census forms were based on
prepared material and adapted in Excel format. The first form was printed on
green A4 hardcopy, the color of the health insurance scheme with the heading
“Maliando” Household Card”. After completion of the card by the censors it was
left with the household. The censors transcribed the information on a separate
form, the “household summary” form which did not contain names but a summary
of the information. It contained the number of household members in different age
groups, the sexes and resident status. In addition it contained information about
the number of wives of the household head, his profession and a listing of the
main economic activities in hierarchical order. A separate form in the census
condensed general information about the village, like distances to health
structures, road conditions, information about sanitation, schools and other.
A GPS (Global Positioning System) handhold electronic device, was available to
determine physical distances of the villages to the HC Yende. During supervision
visits and supply and collection of material to the censor groups with the vehicle,
the GPS system was activated in different villages by the main supervisor. In
addition previously collected data about distances existed in Excel Files.
4.5.1.2 Assessment of outpatient register
The outpatient records of the Yende health center was asessed and information of
the total number of visits to the center between the 12-month period covering
Page 43
43
June 1st 1999 and Mai 31st 2000 extracted. The following variables were noted:
the village or sector of town of origin, the age, the sex, the profession, the
insurance status, the documented complaints, diagnoses and the drugs prescribed
as treatment for the presenting complaint.
Data entry sheet were developed in the statistical software package “File Maker”
converting the column headings of the patient register into categories.
To reduce typing time, the entry sheet was prepared in such a way that during the
time of data entry most data could be selected from preexisting lists and were
made available in form of rolling bands on the screen. The correct values were
entered by choosing from a menu moving with tabulation and the keyboard curser.
Only numbers had to be typed or in exceptional cases new expressions.
4.5.1.3 Asessment of patient cards (ANC, EPI, FP)
Records on these activities existed in form of hardcopies for ANC, EPI and FP,
which were kept in the offices of the respective programs at the HC. The forms
had annotations distinguishing between fixed strategy contacts (the patients came
by themselves to the health center) and contacts which resulted as part of an
outreach strategy. A form was designed, in which a complete list of approximately
180 villages where listed in the order of administrative units.
It contained columns for the three programs which each had the following 3
subdivisions; patients/clients registered, number of contacts, Outreach strategy
Yes or No. The collected data was then entered into a prepared FM file.
4.5.2 Selection, staff training and Pretest
4.5.2.1 Population census
Three census teams were formed, each with one supervisor, 3 censors and a
guide. The supervisors were three PRIMA staff. The three censors of each team
were made up of a regional head of the health insurance “Maliando”, a health
worker of the district and an independent censor who was involved previously in
Page 44
44
qualitative research activities with PRIMA. A one day training session of the
censors was conducted. The responsibilities of censors and supervisors were
defined, the different forms explained and pre-tested, which resulted into
corrections of the forms.
4.5.2.2 Assessment of outpatient register
Two teams were formed, each consisting of a trained secretary experienced in
entering data in computer software, with notions of medical expressions from
previous typing work. They were assisted by a secondary school student, whose
task was to dictate from the register. The training of the teams, explaining data
entry and about technical terms took 2 days. The list of medical expressions for
the different categories was printed out and relationships between complaints,
diagnosis and treatment explained in order to facilitate understanding and
recognition of written data.
4.5.3 Data collection procedure/ Computer Data entry
4.5.3.1 Population Census
A one day mapping session was conducted, which gathered members of
“Maliando” and health workers involved in outreach activities. Three large
handdrawn maps which had been copied beforehand from existing maps and the
existing villages and roads were discussed, verified or corrected by the
participants. Additional small villages or separate groupings of houses (“hameau”)
belonging to villages were added. The distances between the villages were noted
and combining all information a detailed movement of the three census teams
were planned.
Local government and health authorities of the two Prefectures concerned had to
be informed and a written permission with the signature and stamps of all
administrative levels obtained (“ordre de mission”). Two meetings of community
leaders of the area were used to inform the population about the census activities.
Page 45
45
As well the information was given through churches and mosques. The local
authorities actively supported the program.
The census lasted 14 days. The censors moved door to door and usually asked
the household heads for the information about household members. Movements of
the censors were mainly done on foot but one pickup and 2 motorbikes were as
well in the field. For the town of Yende, as well as other larger villages detailed
maps with the streets and numbered buildings were prepared beforehand by
“Maliando” members. This method was especially useful to identify households,
which did not provide the information at the first visit and where the censors had to
return.
Data of distances to Yende as operational distance and walking time, were
collected by questioning a group of villagers of the different towns and noted on a
separate form. The selected villagers were asked to give values for perceived or
known distances and walking time to Yende. In addition we obtained data for 20
villages on physical distances using the GPS system.
4.5.3.2 Assessment of Outpatient register
The data from 4996 patient contacts from June 1999 to January 2000 were
extracted from five separate patient register hard copy books and entered into the
computer. Each team did approximately 150 data entries each day.
4.5.3.3 Assessment of patient cards (ANC, EPI,FP)
For ANC and FP the midwife in charge of the program sorted out all forms which
included contacts of the period between June 1999 to Mai 2000.
One by one the forms were read and each patient for whom a card existed was
noted once as registered client of the program with her village of origin. In addition
the number of ANC or FP contacts during the one year time period were noted
down as well as whether the contact was part of the fixed or outreach strategy.
For EPI the health agent responsible sorted out all hardcopy forms and the
coordinator of the study noted down the information. In a second stage all
Page 46
46
information was entered in a prepared „File Maker“ entry sheet and later exported
as an Excel file for further data analysis.
4.6. Data analysis
4.6.1 Population Census
A team of 4 persons with a supervisor transferred the data from the household
forms on another sector/village synthesis sheet. Each line represented the
information of one household. At completion of one sector/village sheet in the
bottom line all figures were added up, giving the total figures. The information on
profession and economic activities was synthesized on a separate sheet.
The data was transferred in a “File Maker” file, and after completion exported into
Exel format for further analysis. The patient entries of contacts with no location
stated (237) and of not concerned Sub-districts 31 were eliminated from the
sample, which made a total of 7122 contacts used for data analysis.
4.6.2 Assessment of Outpatient register
To obtain the final results several steps were necessary. The data from the
population census provided the denominator data. The review of the patient
register provided the exact number of patients having attended the health center
from the different towns (Nominator).
As a result it was possible to calculate per capita utilisation rates for the different
services offered at the HC for each village. The villages surrounding the HC (15
km) were grouped into concentric distance bands every two km (1-2, 3-4, 5-6, 7-8,
9-10, 11-12, 13-14, 15-16 km). As a result the per capita utilisation rate for each
of these concentric distance bands were calculated.
4.7 Validation
To ensure a correct census the supervisors were constantly in the field with the
censors and checked the data sheets after completion of a working day.
Page 47
47
During data entry of data from the outpatient register on most days a health
professional (medical doctor or nurse) was in proximity to answer comprehension
questions. The teams were instructed to mark any expression they did not
understand for easy recognition. The data on provenience, age, sex, profession,
insurance status and diagnosis should be considered as reliable, as no problems
were observed during the time of data entry. In contrast the data on complaints
and treatment need again to be examined thoroughly if these data should be used
to evaluate the quality of establishing diagnosis and the corresponding treatment..
During data compilation of the census data each synthesis form was double
checked by a supervisor to avoid calculation or transcription errors.
4.8 Limitations
A document review can only be as accurate and complete as the primary data
source is. 269 patient entries (3,6 %) of the data entered into the computer had to
be eliminated during analysis because the data on the origin of the patient was not
stated or could not be read.
4.9 Ethical considerations
Consent for the study was obtained from the district administrative and health
authorities. Participation in the study was voluntary. Permission was sought from
the chief of the clinic Yende and the head of the different departments
To examine the data of their services.
Page 48
48
5. RESULTS
5.1 Introduction
This study was aimed to describe the attendance pattern of the health center
Yende related to the factor distance. To determine utilisation rates a precise
denominator for the calculation was needed. For this reason it was deemed
necessary to conduct a population census. In the following the results of the
census and the assessment of the outpatient register and documentation of the
preventive services (ANC, EPI, FP)
5.2 Population census
5.2.1 Global figures
During the census 86 town sectors of Yende villages and separate units of villages
(“hameau”) up to 19 km distance from Yende were visited. The great majority of
the villages were in the 15 km range (93%). The total counted population was
19.961 living in 2834 households (19.224 in the 15 km range). Three additional
villages should as well be within the 15 km range where the census did not reach.
These were not included in the study. The proportion of Yende’s share of the
population living within the 15 km radius is 57,4 % (11.032 people). The rest
42,6% (8192 persons ) live in the neighboring sub-districts Boloudou,
Guendembou and Temessadou.
5.2.2 General population description
48 % of the counted population were male, 52 % female. The percentages of the
age groups were the following: 0-11 months 3.6%, 1-4 years 13.3%, 5-14 years
30.2%, 15-49 years 42%, 50+ years 11%. The percentage of women in the
reproductive age (15-49 years) was 23.5%.
Page 49
49
5.2.3 Relating distance to the population figures
For the purpose of analysis the area surrounding Yende was divided into
concentric distance bands of 2 and 5 km. The following figures represent the
number of persons living within the concentric distance bands. Half of the persons
counted was living in a radius from 5 km; 9870 (51,3 %). In the area from 6 - 10
km 5781 lived (30 %) and 3573 (18,6 %) people were living in the remaining 11 –
15 km. In summary 81,3 % of the population were living in the 10 km cycle.
5.3.1 Assessment of Outpatient register (CPC) and of preventive Services
(ANC, EPI, FP)
5.3.1.1 Origin of attendees
There were a total of 7.390 reported visits to the during the 12 months of the
period under study which were entered into the computer. From these 269
entries were eliminated as the village of origin was not stated and entries from
towns and villages which were definitely from outside the region (7221). In the
further data process to allocate the villages into concentric distance bands 47
patient contacts were lost and 7074 were used as basis for the calculations.
The curative care users which were included in the analysis (7121) came from
six sub-districts. From Kissidougou district: Yende 6140 (86,3 %) and from
Gueckedou district, Bolodou 607 (8,5 %), Guendembou 233 (3,3 %),
Temessadou 109 (1,5 %) and Kondiadou 23 (0,32 %). In total 13,7 % of the
users came from the neighboring districts.
From the total population living in the 15 km range 11254 live in Yende sub-
district. 5.987 patients from Yende sub-district consulted the HC from within the
15 km range and 153 from beyond. 5363 came from the areas 1-5 km 5363 (89
%), 6-10 km 413 ( 6,9 %), 11-15 km 211 (3,5 %).
Page 50
50
Related to the total figure of patients which consulted at the HC Yende from the
whole of the Yende sub-district (7121) the figures were the following:
1-5 km 5368 (75,3 %), 6-10 km 413 (6,9 %), 11-15 km 211 (3 %), 16-20 km 17
(0,2 %), 21-25 km 56 (0,7 %), 26 – 30 km 40 (0,6), 31 – 35 km 24 (0,3 %), 36-
40 km 16 (0,2).
5.3.1.2 Age groups
The percentages of the age groups were the following: 0-11 months 3.6%, 1-4
years 13.3%, 5-14 years 30.2%, 15-49 years 42%, 50+ years 11%. The
percentage of women at reproductive age (15-49 years) was 23.5%.
5.3.1.3 General features attendance curative care (CPC)
More females than males attend the health center. 58 % of all patients were
female, 41,4 % male which makes a female / male ratio of 1,4. The Division of
patients between the different age groups was the following;
0 – 11 months 13,7 %, 1-4 years 18,9 %, 5 – 14 years 14 %, 15 – 49 years 44,2
%, more than 50 years 9,2 %.
Attendance varied by age and was highest In Infants (<I years). As well children
from less than five years old attended more than can be expected from their
proportion of the general population (13,3 %). The age group 5-14 years
attended half the number of times expected by their proportion of the population
(30,2). The other age groups attendance rate corresponded with their
proportion.
Table 2: Outpatient attendance compared with census data
0–11 Months 1 – 4 Y 5 – 14 Y 15 – 49 Y 50+ Y
Gen. CPC 13,7 % 18,9 % 14 % 44,2 % 9,2 %
Census 3,6 % 13,3 % 30,2 % 42% 11%
Guinea 97 4,0 % 14 % 26 % 56 %
Page 51
51
4.3.1.4 General Distance
The great majority of the patients attending the health center for curative care
came from the area immediately surrounding the health center. From within the
distance of 5 km attended 5629 patients (79 %), from the distance of 6-10 km
859 (12,1 %) , from 11- 15 km 339 (4,8 %) and the rest from more than 15 up
to 40 km distance 276 patients (3,9 %). In summary from within 10 km came 91
% and from within 15 km 95,9 % of all patients.
4.3.1.5 Sexe-Distance Interaction
The female/ male ratio is 1,4 from 1 – 5 km, from 6-10 km 1,3 , between 11 and
15 km 1,6 and for the area going beyond 15 km it stays at the same 1,6.
Comparing with the average female/ male ratio of 1,4 for CPC the general trend
of females attending significantly more than males is even more pronounced
with distance.
4.3.1.6 Age groups-Distance Interaction
Infants are generally brought more often to the health center 1-5 km 15,8 %,
their attendance declines drastically with distance. 6-10 km 9,6 %,
11-15 km 0 %. Practically no infants are brought to the health center the
parents living more than 10 km away.
The inverse situation is the case with the age group of less than 5 years. Their
attendance rate is proportionate at 1-5km 14 % , but then increases steeply to
33 % 6-10 km and increases even further at 11-15 km to 41 %. The age group
5-14 years attends only half as much compared with their proportion in the
population. In the other age groups differences are less marked.
Page 52
52
Table 3: Attendance Outpatient care Age groups in relation to distance
Distance 0–11 Months 1 – 4 Y 5 – 14 Y 15 – 49 Y 50+ Y
1 – 5 km 890 (15,8 %) 790 (14 ) 795(14,2 ) 2603 (46,2 ) 551(9,8 )
6 – 10 km 83 (9,6) 291(33,9) 103 (12 ) 334 (38,9) 48 (5,6 )
11–15 km 0 140 (41,3) 60 (17,7 ) 103 (30,4 ) 36(10,6 )
15 + km 0 120 (43 ) 40 (14,5 ) 97 (35,1 ) 19 (6,9 )
4.3.1.7 Distance Decay
There is a steady drop in utilisation rate of patients coming from locations
further away from the health center (distance decay). The area surrounding the
health center was divided into concentric distance band of 2 km. The following
values represent the number of CPC contacts with attendance rates; 1-2 km:
5183 (0,69), 3-4 km: 309 (0,20), 5-6 km: 277: (0,16), 7-8 km: 438 (0,16), 9-10
km: 233 (0,11), 11-12 km: 163: (0,12), 12-14 km: 52 (0,04), 15 km: 120 (0,12).
The medium utilisation rate in the 15 km radius is 0,35.
raph 1: Utilisation rate of Outpatient care related to distance
Page 53
53
Utilisation rate for Yende Sub-district
The utilisation rate from within the 15 km zone is 5987 / 11254 = 0,53
(Utilisation 53 %). The remaining population of the Yende sub-district outside
the 15 km radius should be 17.319 minus 11.254 which equals 6.065. The
utilisation rate of this population is 0,025.
35 of the population of the Yende sub-district population lives beyond the 15
km radius. The utilisation rate for the population coming from within the sub-
district Yende is 5.987 / 17.319 which equals 0,35.
4.3.1.8 Utilisation of Preventive Services
The results for preventive services can not be compared as for CPC contacts.
The health center Yende does not usually provide ANV, EPI and Family
Planning services to people coming from other sub-districts than the sub-district
Yende. For coverage calculations only the area of the 15 km radius surrounding
Yende belonging to Yende district containing 11.032 people (57,4 %) can be
taken into consideration. Therefore Coverage for concentric distance bands can
not be calculated as usual for preventive services.
Antenatal care and Family Planning
The ANC section of the health center Yende had during the year under study,
according to the Antenatal forms reviewed, (“fiche consultation prenatale”) kept
at the office 885 clients with the total number of 2609 contacts ( average 2,8
ANC contacts per patient). Of these 208 clients were part of the outreach
strategy (24 %). As described before the coverage calculation has to be
modified by multiplying by the factor 0,574. Considering this factor the overall
coverage rate for the share of the sub-district Yende of the 15 km range was
0,41. From the total of 885 ANC clients 622 were from within the 15 km range
(70,3 %).
In addition the ANC department saw 158 Family Planning clients with an
average of 2,16 contacts per client.
Page 54
54
Expanded Program of Immunization
The EPI section vaccinated 764 children with in average 2,9 contacts per child.
389 (51 %) children were vaccinated as part of an outreach strategy. The
coverage rate for the children within the 15 km range in the sub-district Yende
was 0,39. If the health center would not have engaged into EPI outreach
activities the coverage rate would have been 0,26 (only fixed strategy). 471
children (61,6 %) of the children vaccinated were living in the 15 km range.
4.3.1.9 Distance and coverage rates
The coverage rate for ANC and EPI compared with utilisation for CPC is as
described in the following table;
Table 4: Comparison of utilisation rates Outpatient-, Antenatal- care, Expanded Program
of Immunization and Family Planning (5 km Distance Bands)
Dist.Bands Total
Pop.
CPC Util.Rate ANC Couv. EPI Couv. FP
1-5 km 9870 5592 0,57 464 1,04 317 0,98 134
5-10 km 5781 848 0,15 78 0,30 79 0,30 8
11-15 km 3573 335 0,09 80 0,50 75 0,66 5
Total 19224 6775 0,35 622 0,72 471 0,67 147
Page 55
55
Table 5: Comparison of utilisation rates Outpatient-, Antenatal- care, Expanded Program
of Immunization and Family Planning (2 km Distance Bands)
Dist.Bands Total
Pop.
CPC Util.Rate ANC Couv. EPI Couv. FP
1-2 km 7486 5183 0,69 397 1,18 270 1,02 131
2-4 km 1578 309 0,20 61 0,86 39 0,98 3
5-6 km 1710 277 0,16 21 0,27 19 0,31 2
7-8 km 2812 438 0,16 37 0,29 35 0,37 3
8-10 km 2065 233 0,11 26 0,28 33 0,26 3
11-12 km 1374 163 0,12 55 0,89 45 1,07 1
12-14 km 1187 52 0,04 10 0,19 12 0,36 2
15 km 1012 120 0,12 15 0,33 18 0,46 2
Total 19224 6775 0,35 622 0,72 471 0,67 147
There is a marked decrease in utilisation for all services with greater distance.
There is the exception of the PEV utilisation rate at 9-10 km with 0,62 which
may be due to active outreach activities.
4.3.1.10 Attendance “Maliando” Members
The number of patient contacts of the 1114 “Maliando” insurance members was
1999 which represents 27,8 % of all attendance’s. Considering the max 9 %
coverage rate of the insurance scheme this is a high number. This observation
corresponds with the fact that “Maliando” insurance members use about 4 times
as much the health facility compared with non-members. In this study
“Maliando” members use the HC 5 times as much.
695 members (69 %) live within the 5 km , 260 (26 %) in the 6-10 km and
5, 8 % in the11- 15 km range. The health insurance scheme attracts very few
members from distant locations.
Page 56
56
As a further step utilisation rates of insurance and Non-Insurance members for
concentric distance bands of 5 km width were established with increasing
distance for both groups. The following table represents the decrease in
utilisation rate of Members and Non-members of the health insurance
Table 6: Utilisation rates of Non-Members and Members
Non-Members Members
1 – 5 km 0,52 1,96
6 – 10 km 0,12 1,44
11 – 15 km 0,09 1,02
Compared with the group living within the 5 km radius the insured group in the
15 km radius used half as much the services of the health center for curative
care. The decrease in utilisation according to distance is much more marked in
the Non-Member group. In the 10 – 15 km distance band their utilisation rate
drops by the factor 5,8 compared with the insured group (2). The relationship
Member to non-Member is 3,8 within the 5 km range, 12 within the 5-10 km
range and stays about equal with 11,3 in the 11-15 km range. Beyond 10 km
distance affects both groups equally. The utilisation rate of Non-Members
declines in the range 5-10 km by the factor 4,3 whereas it drops in the Member
group only by the factor 1,4. The average utilisation rate of members is 1,77 in
the current study (contacts; 1793, members; 1014).
Graph 2: Utilisation Rates according to insurance status
5.4.1 Summary of the Findings
The major users are the population immediately living around the health center.
79 % of the patients came from within the distance of 5 km and 91 % live up to
10 km. The proportion of people coming from larger distances declines
exponentially.
Page 57
57
There is an important difference between sexes in attendance rates, with
woman attending 41 % more often than the males. The study shows age as a
factor in the utilisation of health services. Those under the age of 5 years made
more use of the health center services than any other group, the highest
attendance has the infant group which shows 3,8 times more attendance than
expected by their proportion of the population.
There were marked differences in the distribution of age groups of the
population attending from different concentric distance bands.
Infants attendance rate decrease significantly with distance, Under five
attendance rate increases with distance.
Surprisingly much less people from the neighboring sub-districts attend the
health center compared with their proportion of the population. Yende
comprises only 57 % of the population in the 15 km radius, but provides 86,3 %
of all the patients of the health center.
“Maliando” members living in a distance of 11 –15 km use the services half as
much as members living within the 5 km range.
Decrease in utilisation according to distance is much more marked in the Non-
Member than the Member group. Between 5 and 10 km the utilisation rate of
Non-Members declines by the factor 4,3 whereas it drops in the Member group
only by the factor 1,4.
The health insurance scheme attracts very few members from distant locations.
Page 58
58
6. DISCUSSION
6.1 Introduction
As described in the literature review the utilization of a health center is
influenced by several factors one important is the factor distance. In this part
the results of the study are compared with results of a „monitorage“ and a
previous study on utilization of health insurance members by the PRIMA team.
The results are put into context with the current situation of the health insurance
scheme „Maliando“ .
6.2 Institutional versus administrative catchment area
population/Health
The institutional catchment area population is estimated to englobe the
inhabitants of villages and towns living within an 15 km radius of the health
center. The older estimation by von Roenne was that it comprises 17.250
inhabitants (Roenne, 2000) The results of the household survey increased the
estimated figure by 10 % to 19.224. The catchment area does not respect the
administrative repartition and contains areas of 5 sub-districts Yende,
Boloudou, Guendembou and Temessadou. This constitutes a difference with
the administrative catchment area, which is the whole sub-district of Yende.
Previous analysis of the patient register had shown that the population living
more than 15 km up to 40 km away in the North of the health center very little
use its services. Often this population attendance is only during the market days
when they come for other purposes and combine a visit to the health center.
A considerable part of the HC attendees were thought to come from the
neighboring Sub-districts. Even if these populations may without any problem
use the services of the HC belonging to another administrative unit they do not
have access to preventive services (ANC, EPI, FP). The supply of vaccines and
other material corresponds to the administrative population, so their supply is
given to the health center corresponding to their residence. For utilisation or
coverage calculations the administrative catchment area serves as the
denominator, although in reality many of this population realistically do not have
Page 59
59
access. This was one of the assumptions of the study. To approach the subject
the findings of the study are compared with the results of the “monitorage” of
the HC Yende from December 1999.
6.3 Results from the “Monitorage” December 1999
Table: Results “ Monitorage” HC Yende Dec. 1999
Service Denominator
(Target Pop.)
Nominator Util.Rate/
Coverage
Accessibility CPC 17.319 12.751 74%
Prev. Services 17.319 15.011 99%
Utilization/
Coverage
Calculations
CPC 17.319 5027 58%
ANC 390 381 99%
PEV 346 344 99%
FP 1039 120 12%
6.3.1 Accessibility
The figure for the population living in the 10 km range all neighboring sub-
districts is 15.651. From this population only 57,4 % live in Yende sub-district.
According to this calculation 8.984 persons live in the 10 km range of the HC
within the sub-district Yende which represents 52 % of the total sub-district
population. At the time of analysis no figures about the population living within
the 10 km range of the 2 health posts of Walto and Firadou were available.
Combining the findings of this study with the figures as basis for the calculations
for the “monitorage” this population should be 5.676. In this case accessibility
would be 90 % as stated in the “monitorage” report.
Page 60
60
6.3.2 Curative Care Utilisation from within the Yende sub-district
As elaborated in the results chapter the “monitorage” gave a figure of 58% for
the whole sub-district. The average figure in Guinea for CPC utilisation in 1997
was 0,26, with the region of Kissidougou having 0,39 (MSP, 1997).
The figure of 0,35 established by the current study can not be directly
compared as the attendance’s of the health posts are not known and would
raise the utilization rate. Taken the number of attendants of the 2 health posts
from June – November 99 (892) multiplied by 2 an estimation can be made.
Adding the number of attendees of the HC from Yende Sub-district and the
extrapolated attendees of the HP’s the total number is 7924 (Coverage rate 45
%). The difference to the Utilization rate from the “monitorage” in December can
be well explained by the fact that 1250 data entries were not considered. 278
contacts could not be localized and 972 (13,7 %) attendees were not
considered in this calculation as belonging to other sub-districts. Adding this
number would make a coverage rate of 53 %, which is very close to the 58 %
result from the “monitorage”. Normally it is not possible to exactly determine the
origin of patient that’s why there is no other way than to use the mode of
calculation as in the “monitorage”. Patients from the sub-district will as well
attend health centers in other sub-districts which will probably equalize the
situation.
Looking at the small proportion (17 %) of attenders from other sub-districts in
comparison with their population share 42,3 % of the 15 km radius the
difference between institutional and administrative catchment area is less
marked than expected. Yende comprises only 57 % of the population in the 15
km radius, but provides 86,3 % of all the patients of the health center.
It is even surprising how few people use the facility who live relatively close but
are from another sub-district. Several reasons may play a role, one being that a
private health center is established in 2 km distance in southern direction in
Mano, but this does not provide preventive services. An explanation may be
that people would not like to attend where they will not receive comprehensive
services (including ANC, EPI, FP) but from practical observations this seems
Page 61
61
unlikely. Does the percentage of 13 % of attendees from other sub-districts
require special administrative or logistical arrangements? The results of the
study would argue against it.
6.4 Member and Non-member utilisation rate
The utilisation rates obtained in this study for insurance members (1-5 km 1.96,
6-10 km 1,44, 11-15 km 1,02) corresponds with the findings of Hohmann
(1999) and are less than in the documentation by Diallo. The average utilisation
rate of members is 1,77 in the current study (contacts; 1793, members; 1014)
are as well comparable with the other studies. The health insurance scheme
attracts very few members from distant locations. Only 26 % of the members
come from the 6-10 km and 6 % from the11- 15 km range.
Members living more distant to the health center use the health center less and
produce less costs to the insurance scheme. The study shows that insurance
member households living within 5 km of the health center use the services
twice as much than households living beyond 10 km distance.
Households, who live far away and pay the same subscription fee actually
subsidize the households living close to the health facility who use the services
more frequently as stated by Noterman (1995). Remote households can not
see a benefit in joining as they rarely attend the clinic, even if they are sick.
The biggest obstacle to join the health insurance scheme is the premium which
is perceived as being to high for a complete family. Rural households find it
difficult to accept the payment of larger amounts of money in advance for
possible illness. The aspect of being protected against major health expenses
for a whole year and the feeling of security associated with it is not easily
appreciated. In the case of catastrophic illness solidarity amongst community
members usually dictates to help each other. In contrast it would be difficult to
obtain a loan for preventing major expenses for illness.
To improve acceptability of the health insurance in the population graduated
premiums (sliding scales) according to household location could be introduced
Page 62
62
while at the same time the premiums would be still self-financing. This was as
well proposed by Noterman (1995) and Bennet (1997).
Another option would be to suit service packages to the different location of
households for instance by insuring only low frequency high cost events
(catastrophic illness) for lower premiums. This is the basis for the relatively
successful Bwamanda health insurance scheme in Zaire. To put such more
complicated mechanisms into practice organizers of a insurance scheme need
information on utilization and risk patterns to be able to set appropriate
premiums. The current study hopes to provide information on this subject.
6.5 “Maliando”/ PRIMA
Targeting the poor rural mainly self employed populations, the health insurance
scheme “Maliando” faces the same problems as many other district based
schemes. Many schemes have limited population coverage and do not function
cost-efficient. (Bennet, 1997). Despite the low membership figures there are
obvious achievements of the scheme which go beyond the aspect of pure
health financing. One of the major achievements of PRIMA with the MUCA
model (“Maliando”) is that it initiated a dialogue between users and providers of
health services. Despite official proclamations about community participation
and control this is rarely put into reality. Criel stated that a very “utilitarian”
content that is given to community participation still very much prevails among
health care managers. The view of users are too often neglected. The MUCAS
model gives some level of power to the users and defers specific rights and a
forum to formulate claims and complaints. This clearly showed in public
discussions in which “Maliando” members expressed their dissatisfaction with
the quality of services. It is a novum that doctors and nurses are confronted
openly with critiques. It led to a situation that the system that had not changed
much its recommendations and guidelines since the early years of PHC had to
adapt to a new situation. Criel stated (1999) “that there still remains a very
strong grip of the central level on peripheral decision-making and that there
Page 63
63
(still) is a poorly developed culture among district managers to, themselves,
plan and take decisions”.
With the occurrence of the health insurance scheme health managers have to
start to question themselves. The Ministry of Health who is interested in further
research in this issue accepted that in Yende new forms of service delivery
which could be more acceptable to the population may be experimented. One
of the problems are the very strict guidelines which do leave very little room for
negotiations with the clients. If the health agents do not prescribe according to
the guidelines their rating at the “monitorage” will decrease. The perceived
quality of care seems to be a major reason for misunderstanding between the
health services and the users. There an important difference between what the
general population perceives as good quality health care and what the health
services are technically able to provide, following the principle of good standard
practice. It was observed that people do not value the comprehensive services
offered at the health center, equalizing the medical treatment with dispensing
drugs. The amount of drugs received represents the good or bad treatment.
Injections are considered high quality treatment. (“Strong medicine“). People
perceive the drug variety as too restrictive and consider Aspirin, Chloroquine
and Cotrimoxazol as low quality treatment because it is so often prescribed. A
lot of misconceptions exist, technically the treatments offered by the health
services are cost-efficient and usually adequate but not well accepted by the
population. Considerable effort by the promotors of the health insurance
scheme have to be made to overcome these misconceptions in order that
providers and users speak a common language. The problem of questionable
quality of care by the health services shall not be overlooked. Partly through the
influence of “Maliando” the health services are becoming more interested in the
issue and have initiated the participation of Yende HC in the first phase of a
quality improvement project.
Another problem which has come on the open is the frequent „Under- the-
counter- payment“ of health staff in Guinea, which is now discussed openly.
Von Roenne (2000) documented this very detailed for the health center Yende.
Page 64
64
Another problem observed is that although the health center generates enough
money it is not allowed to buy drugs in private pharmacies in case of rupture,
which can always occur. The purchase of drugs have to go through the official
drug procurement channels, which makes a quick solution to a minor problem
difficult. Generally the experience of Maliando proves that establishing a health
insurance scheme may enable to introduce organisational changes and service
quality improvement as mentioned by Gilson (1997).
Page 65
65
7. CONCLUSIONS AND RECOMMENDATIONS
Based on the findings of the study the perceived difference between
administrative and institutional catchment area is much less marked than
expected. Only 13 % of attendees come from neighboring sub-districts.
No major administrative or logistical arrangements are necessary to
accommodate this low percentage of users from other districts. For the
calculation of drugs and vaccines for the health center remote communities are
included which will never attend as the distance is too far. For these populations
the distance constitutes a significant barrier to access and they will not use their
share of the logistics. It is doubtful that in the current situation of budget
restrictions it is feasible to add more health posts to cover isolated populations.
It is more important to improve the quality and therefore the attractiveness of
the existing health facilities.
It is a strategic decision whether health centers should only be allowed to
provide preventive care to the population of their administrative district. The low
quantity of patients does not justify major changes.
The acceptability of the community based health insurance scheme „Maliando“
is low. To increase the attractiveness Maliando members in positions of
responsability have to be better informed about basic medical standarts, to
reduce misconceptions about quality of care, to better appreciate standart
medical treatment. Only if members themselves dispose of sufficient and
convincing arguments about the advantages of the scheme they will have
success in convincing others. Considerable effort by the promotors of the health
insurance scheme have to be made to overcome the mentioned
misconceptions in order that providers and users speak a common language.
The lower co-payment levels for Maliando members living in a distance is not
satisfactory to provide a response to the problem of distance. More powerful
incentives, especially lowering the most important barrier, the annual premium
Page 66
66
may be more adequate. The main recommendation would be to include
information on health spending, utilization and risk pattern in calculating
premiums at levels acceptable to the population. A system of graduated
premiums according to household location (sliding scales) may have a positive
effect. At the same time the health services have to continue to improve the
quality of their services by fully engaging in quality cycle activities
8. ANNEXE