Rationale
There is a lack of information on socio-economic and other differences in the distribution of the benefits
of free public health services and of the financing burdens associated with different financing
mechanisms in health, particularly in terms of financial protection for the poorest and other vulnerable
groups. This information is required by policymakers and programme managers so that they can develop
and implement financial risk protection strategies that aim to reduce inequity in financial access to, and
utilisation of, healthcare services by the poor and ensure that the poor and vulnerable groups are not
impoverished by healthcare spending.
Objectives
The aim of the study was to generate new knowledge about the burden of different mechanisms for
financing the use of health services by households, and to assess the distribution of the benefits from
government expenditure on a set of priority public health interventions. This research was undertaken
with the aim of using the findings to improve the equity of financing and provision of healthcare services
in Nigeria. The specific objectives of the study were: to estimate the relative contributions of the major
healthcare financing mechanisms at household level to the health expenditure in Enugu and Anambra
states, Nigeria; to assess the financing incidence (based on socio-economic groups and rural-urban
location) of the different financing mechanisms; to determine the level of catastrophic healthcare
payments and their distribution across socio-economic groups; and to evaluate the benefit incidence
(based on socio-economic groups, gender, rural-urban location) of a subset of publicly-funded priority
health services.
Methods
Study area
The study took place in 2 selected Local Government Areas (LGA); 1 rural and 1 urban in each of Enugu
and Anambra states respectively (total of 4 LGAs). Study sites were selected to try to capture the variety
of major financing mechanisms operating at the household level.
Data collection
The study involved household surveys and document reviews. Interviewer-administered pre-tested
structured questionnaires were used in household surveys. A one-month expenditure recall period was
used in the household survey.
Data analysis
STATA and SPSS software packages were used to analyse the data. Financing incidence analysis (FIA) and
catastrophic costs were assessed at the household level (n=4873). Our data showed that direct out-ofpocket
spending (OOPS) dominates household expenditure on health. On the basis of this it was chosen
as the main focus of the FIA. To examine the Benefit Incidence Analysis (BIA) the data from all
individuals in the households were used (n=22,169). The frequency distributions of the variables by
socio-economic status (SES), rural-urban location and gender (depending on the research objective)
were calculated and the chi squared (Chi2
) test for trend analysis for statistical differences across the
states was applied.
8
Principal components analysis (PCA) was used to create a SES index using information on the
households’ ownership of some assets, together with the weekly household cost of food. The index was
used to divide the households and individuals into SES quintiles respectively. Concentration indices were
calculated for all SES differences.
Analysis of the level and distribution of healthcare payments
The average expenditure by households on out-patient visits and in-patient stays were calculated. The
comparisons of healthcare expenditure levels were disaggregated by SES, gender, and location
(urban/rural). In addition, data were collected and analysed on: levels of enrolment; use of various
financing mechanisms; SES; and geographic differences in the use of different health financing
mechanisms.
Financing incidence of out-of-pocket spending
The absolute amounts of expenditure through the different financing mechanisms were calculated.
However, because of its dominance as a financing mechanism, FIA was limited to OOPS. To assess the
equity of the distribution of out-of-pocket payments, the concentration index was used. Concentration
curves of OOPS were plotted, using the Lorenz curve, of total household expenditure to show the
distribution of the burden of OOPS by SES compared with total household expenditure. The Kakwani
index was calculated to examine the overall progression or regression of OOPS.
Incidence of catastrophic health expenditure
A number of different indicators and threshold levels were explored to determine the incidence of
catastrophic health expenditure. A threshold of 40% of non-food expenditure was used in order for the
results to be comparable to the international literature for the main interpretation of catastrophic
spending.
Benefit incidence of priority publicly provided health services
The BIA focused on a set of priority public health services that are supposed to be provided free of
charge in the public sector. The key steps in calculating benefit incidence guided the analysis. These
were: determine the level of utilisation of the goods or services; group users by socio-economic and
rural-urban categories; determine the unit cost for the service; and multiply the net unit cost by group
service utilisation to determine group-specific benefits. Information on unit costs of services was
obtained from the Ministries of Health at the Federal and state levels. Subsidies for different services
were then aggregated. In addition to individual level analysis, analysis was also conducted by household.
The net benefits were calculated by subtracting payments made for services from the value of benefits.
Results
General health service use
? Interviews were conducted with 4,873 households (2,483 urban and 2,390 rural). Data were
collected on 22,169 individuals residing in the households.
? Malaria was the major health condition that required out-patient department (OPD) visits and
in-patient department (IPD) stays. Hypertension was the most common non-communicable
disease that required OPD and IPD treatment.
? Patent medicine dealers (PMDs) were the most commonly used providers of healthcare services
(41.1%). They were followed by private hospitals (19.7%) and pharmacies (16.4%). There were
inequities in use of the different providers. The rural dwellers and poorer SES groups mostly
used low-level and informal providers.
9
? The main reasons that people who needed healthcare services did not seek care were that the
condition was not serious enough or they could not afford the cost of services.
? The three main suggestions that respondents gave for improving the provision, utilisation and
financing of healthcare services were: provision of free services; subsidising healthcare; and
construction of more public hospitals.
Analysis of healthcare payments at household level
? The average household health expenditure per month was 2,354 Naira (US$19.6). The average
monthly household expenditure on out-patient care was 1,809 Naira (US$15.1), whilst it was
610 Naira for IPD services (US$5.08). Higher expenditure was incurred by urbanites, residents of
Anambra state and the better-off SES groups.
? Household health expenditure was mostly paid through OOPS and the average monthly
household OOPS was 2,219 Naira (US$18.5). There was almost complete absence of health
insurance.
? The average household expenditure per month in public hospital facilities was 661 Naira
(US$5.51), whilst in private hospitals it was 980 Naira (US$8.17). The lowest monthly household
average expenditure was incurred by herbalists (3 Naira or US$0.02) and PMDs (35 Naira or
US$0.29).
? Average monthly household expenditure for communicable diseases such as malaria was quite
high 1,401 Naira (US$11.67) for OPD care and 12,442 Naira (US$103.7) for IPD services for
communicable diseases).
? Enrolment rates in health insurance were very low; only 51 (1.0%) of the households had a
member that was enrolled in a health insurance scheme. The primary enrolees were mostly
adults, who were mostly enrolled and covered by the National Health Insurance Scheme (NHIS).
The number of people that were covered by another household member’s health insurance
scheme was also very low, and most of this coverage was with respect to NHIS.
Financing incidence analysis of out-of-pocket spending
? 3,150 (98.8%) payments were made using OOPS, 9 (0.3%) using reimbursement by employers, 1
(0.03%) through private voluntary health insurance (PVHI), 9 (0.3%) using instalment and 14
(0.44%) through other payment mechanisms. No payments were reportedly made using NHIS or
Community-based health insurance (CBHI).
? There were variations in the use of payment mechanisms by different population groups.
? The Kakwani index for financing incidence of OOPS was -0.18 showing that OOPS was regressive.
Level of catastrophic healthcare payments
? The overall incidence of catastrophic expenditures was 27% at the 40% threshold level.
? The poorer SES quintiles and rural dwellers incurred a higher level of catastrophic health
expenditures.
Benefit incidence analysis of priority public health services
? It was found that 3,281 individuals out of the 22,169 in the households consumed wholly free
services.
? There was a greater consumption of free services by urbanites, residents of Anambra state,
females and poorer SES quintiles.
? Immunisation services were the most commonly accessed free services (2,992 individuals). This
was followed by insecticide-treated nets (ITNs) (313 people) and free antimalarial drugs (61
10
people). Only one person accessed free HIV treatment services. The results also show that free
ante-natal care (ANC), childbirth services and TB treatment were accessed by 22, 3 and 7
individuals respectively.
? The average value of benefits of accessing immunisation services per individual was 440 Naira.
Other values were 96 Naira (ITNs), 79.4 Naira (TB treatment), 49 Naira (ANC), 23 Naira
(antimalarial drugs), 7 Naira (childbirth services), 0 Naira (antiretrovirals (ARVs)), and 0 Naira
(family planning (FP)). A value of 0 implies that nobody accessed the service.
? Overall, the better-off SES quintiles benefited more from the immunisation services, ITNs and
ANC services, whilst the poorer SES quintiles benefited more from free antimalaria drugs and
treatment for TB. In the urban area, the distribution of benefits was more equitable for
immunisation services, ITNs and ANC. In the rural area the distribution of benefits was more
equitable for antimalarial drugs, childbirth services and TB treatment. In absolute terms, males
received a greater share of the benefits of immunisation services relative to their population
share, whilst females received more of the benefits for the remaining goods and services.
? High levels of payment were observed for immunisation services, ITNs, antimalarial drugs, ANC,
and childbirth services, all of which are supposed to be provided for free.
? Compared to rural dwellers, more urbanites spent money on most public health services except
for treatment of TB. Females and residents of Enugu state were more likely to spend money on
most public health services compared to males and those from Anambra state respectively.
However, it is not clear whether these were informal payments or that people chose to receive
the services from the private sector where they were charged fees.
? The average expenditure on public health services were 613 Naira, 76 Naira, and 4 Naira for
antimalarial drugs, immunisation services, and ITNs respectively. Other average expenditures
were 486 Naira (childbirth services), 151 Naira (ANC) and 3.9 Naira (TB treatment).
? In contrast to the rural area, more money was spent on all services, except for childbirth
services, in the urban area. There was also greater expenditure in Enugu state and amongst
better-off SES quintiles.
? The net benefits were significantly higher statistically for the rural area, Anambra state and
poorer quintiles, in comparison to the urban area, Enugu state, and better-off quintiles.
Conclusion
The high levels of out-of-pocket expenditure in the study population indicate a lack of financial risk
protection for healthcare in the study areas. Those that experienced the highest burden of health
expenditure were the rural dwellers and those from the poorest socio-economic quintile. OOPS is still
the principal payment mechanism for healthcare and accounts for the very high levels of catastrophic
spending that were found in the study. Overall, the distribution of OOPS was regressive, as measured by
the Kakwani index. In addition, the low and inequitable coverage of priority public health services – that
are supposed to be delivered free of charge – suggests that there may have been illegal payments which
further hindered access to the public health services. People expressed a desire for increased free public
health services, subsidised healthcare services, and the construction of more public hospitals. Hence,
policymakers and programme managers should introduce health reform mechanisms to develop,
implement and scale-up financial risk protection initiatives in the two states. In addition, reforms should
identify constraints which impede the equitable distribution and access of free or subsidised public
health services for the general population especially poor people and rural dwellers. Reforms should
also ensure that priority healthcare services such as immunisation services are increased become more
equitable in terms of coverage.