Public Delivery of Primary Health Care Services 2002
Quantitative Service Delivery Survey (QSDS)
This survey covering 252 primary health facilities and 30 local governments was carried out in the states of Kogi and Lagos in Nigeria in the latter part of 2002. Nigeria is one of the few countries in the developing world to systematically decentralize the delivery of basic health and education services to locally elected governments. Its health policy has also been guided by the Bamako Initiative to encourage and sustain community participation in primary health care services. The survey data provide systematic evidence on how these institutions of decentralization are functioning at the level local—governments and community based organizations—to deliver primary health service.
The evidence shows that locally elected governments indeed do assume responsibility for services provided in primary health care facilities. However, the service delivery environments between the two states are strikingly different. In largely urban Lagos, public delivery by local governments is influenced by the availability of private facilities and proximity to referral centers in the state. In largely rural Kogi, primary health services are predominantly provided in public facilities, but with extensive community participation in the maintenance of service delivery. The survey identified an issue which is highly relevant for decentralization policies—the non-payment of health staff salaries in Kogi—which is suggestive of problems with local accountability when local governments are heavily dependent on fiscal transfers from higher tiers of government.
Kind of Data
Sample survey data [ssd]
Data were collected in 30 local governments, 252 health facilities, and from over 700 health workers, in Lagos and Kogi states.
Producers and sponsors
National Primary Health Care Development Agency (NPHCDA) and World Bank
Local consultants at the University of Ibadan
A multi-stage sampling process was employed where first 15 local governments were randomly selected from each state; second, 100 facilities from Lagos and 152 facilities from Kogi were selected using a combination of random and purposive sampling from the list of all public primary health care facilities in the 30 selected LGAs that was provided by the state governments; third, the field data collectors were instructed to interview all staff present at the health facility at the time of the visit, if the total number of staff in a facility were less than or equal to 10. In cases where the total number of staff were greater than 10, the field staff were instructed to randomly select 10 staff, but making sure that one staff in each of the major ten categories of primary health care workers was included in the sample.
Health facilities were selected through a combination of random and purposive sampling. First, all facilities were randomly selected from the available list for 30 LGAs. This process resulted in no facility being selected from a few LGAs. Between 1-3 facilities were then randomly selected from these LGAs, and an equal number of facilities were randomly dropped from overrepresented LGAs, defined as those where the proportion of selected facility per LGA is higher than the average proportion of selected facilities for all sampled LGAs.
A list of replacement facilities was also randomly selected in the event of closure or non-functioning of any facility in the original sample. An inordinate amount of facilities were replaced in Kogi (27 in total), some due to inaccessibility given remote locations and hostile
terrain, and some due to non-availability of any health staff. The local community volunteered in these cases that the reason there was no staff available was because of non-payment of salaries by the LGA. This characteristic of the functioning of health facilities in Kogi is a striking result that will be discussed in this report.
Dates of Data Collection
Data Collection Mode
Data Collection Notes
The survey was undertaken during June-August 2002, with data collected in 30 local governments in Lagos and Kogi states, 252 health facilities, and from over 700 health workers.
The approach adopted to addressing these issues revolves around extensive and rigorous survey work, at the level of the primary health care facilities and the local governments. Two basic survey instruments of primary data collection were agreed upon, based on collecting information from government officials and public service delivery facilities:
1. Survey of primary health care facilities—including interviews of facility managers and workers, as well as direct collection of data on inputs and outputs from facility records.
2. Survey of local governments (under whose jurisdiction the health facilities reside)—including interviewers of local government treasurers for information on budgeted resources and investment activity, and interviews of primary health care coordinators for roles, responsibilities, and outcomes at the local government level.
Survey instruments at the health facility level
The facility level survey instruments were designed to collect data along the following lines:
1. Basic characteristics of the health facility: who built it; when was it built; what other facilities exist in the neighborhood; access to the facility; hours of service etc.
2. Type of services provided: focusing on ante-natal care; deliveries; outpatient services, with special emphasis on malaria and routine immunization
3. Availability of essential equipment to provide the above services
4. Availability of essential drugs to provide the above services
5. Utilization of the above services, referral practices
6. Tracking and use of epidemiological and public health data
7. Characteristics of health facility staff: professional qualifications; training; salary structure, and whether payments are received in a timely fashion; informal payments received; fringe benefits received; do they have their own private practice; time allocation across different services; residence; place of origin
8. Sources of financing-who finances the building infrastructure and its maintenance; who finances the purchase of basic equipment; who finances the purchase of drugs; what is the user fee policy; revenues from user fees; retention rate of these revenues; financing available from the community
9. Management structure and institutions of accountability: activities of and interaction with the local government and with the community development committees
Survey instrument at the local government level
The local government survey instruments were designed to collect data along the following lines:
1. Basic characteristics: when was the local government created, population, proportion urban and rural, presence of an urban center, presence of NGOs and international donors
2. Number of primary health care facilities by type (types 1 and 2) and ownership (public-local government, state, and federal government; private-for-profit; private-not-for-profit)
3. Supervisory responsibilities over the general functioning of the primary health care centers
4. Health staff: number of staff by type of professional training and civil service cadre; salary;
5. Monitoring the performance of health staff: how is staff performance monitored and by whom; are staff rewarded for good performance or sanctioned for poor performance, and how; instances when local government has received complaints; what disciplinary action was taken
6. Budget and financing: data on actual LGA revenues and expenditure from available budget documents;
7. Management structures: functioning of the Primary Health Care Management Committee (PHCMC), the Primary Health Care Technical Committee (PHCTC), and the community based organizations-the Village Development Committee (VDC) and the District Development Committee (DDC)
8. Health services outputs at the local government level: records of immunization, and environmental health activities
The focus of the study is thus public service delivery outcomes as measured at the level of frontline delivery agencies—the public primary health care facilities. We also originally planned to include interviews of patients present at the health facilities, to get the user’s perspective on public service delivery, but found that difficult to follow-through given local capacity constraints
in implementing a survey of this kind.
The survey instruments were developed through an iterative process of discussions between the World Bank team, NPHCDA, and local consultants at the University of Ibadan, over the months of March-May 2002. During May 2002, four questionnaires were finalized through repeated field-testing—1) Health Facility Questionnaire: to be administered to the health facility manager, and to collect recorded data on inputs and outputs at the facility level;
2) Staff Questionnaire: to be administered to individual health workers; 3) Local Government Treasurer Questionnaire: to collect local government budgetary information; and 4) Primary Health Care Coordinator Questionnaire:
to collect information on local government activities and policies in primary health care service service delivery.
Random Data Checking Procedure
Following the dual data entry of all records by Nigerian consultants and the merging and cleaning of the data files(as outlined below) by World Bank staff, the hard copies of the questionnaires were randomly checked against the entries in the data files (*) for errors by World Bank staff. Five LGAs were selected at random in both the Kogi and Lagos states. In each of these ten LGAs, the hard copy of the PHC Coordinator Questionnaire, the hard copy of the LGA Treasurer Questionnaire, and up to five hard copies of both the Staff Questionnaires and the Health Facility Questionnaires were randomly selected and checked against the entries in the data files. While in several instances parts of the alphanumeric entries were abbreviated or omitted, no substantive differences between the hard copies of the questionnaires and the entries in the data files were found.
Merging and Cleaning of Data Files
In order to facilitate the evaluation of data across the different levels of the Nigerian primary health care system surveyed in this study, the four survey data files were merged. Three types of problems arose: (1) In matching the Health Facility Questionnaire data file to the LGA Questionnaire data file, there were six facilities (out of 252) that shared three unique codes (**). In matching the Health Facility Questionnaire data file to the Staff Questionnaire data file, (2) there were 7 facilities (out of 252) with no corresponding staff in the Staff Questionnaire data file (3) and 13 staff (out of 719) with no corresponding facility in the Health Facility Questionnaire data file.
Problem (1) was resolved for two of the six facilities by checking the hard copies of the questionnaires against the respective data files and correcting discrepancies. Problem (1) was resolved for the remaining four facilities by requesting an updated list of facility codes from the Nigerian consultants and correcting discrepancies.
Problem (2) was resolved for four of the seven facilities by checking hard copies of the questionnaires against the Health Facility Questionnaire data file and correcting discrepancies. Problem (2) was resolved for the remaining facilities by looking at the comments on the hard copies which revealed that conditions did not a give the respective interviewers the opportunity to interview staff at the facilities in question. These facilities were retained in all the data files since they are indicative of the functionality of the respective facilities.
Problem (3) was resolved for three of the thirteen staff by checking hard copies of questionnaires against the Staff Questionnaire data file and correcting discrepancies. The remaining ten staff still cannot be matched to any existing facilities. In one case, the hard copy of the Staff Questionnaire cannot be found. We conjecture that these ten records were misplaced prior to data entry. These ten facilities were removed from the merged data file yet retained for evaluation in the Staff Questionnaire data file.
(*) For example, in one of the LGAs randomly selected in the Kogi state - Ijumu - only three health facilities were surveyed (two original facilities and one replacement facility), and, thus, only three Health Facility Questionnaires were selected and checked against the entries in the data file.
(**) These unique codes were the result of a systematic combination of a unique state code, a unique LGA code and a unique facility code.
Development Research Group
Use of the dataset must be acknowledged using a citation which would include:
- the Identification of the Primary Investigator
- the title of the survey (including acronym and year of implementation)
- the survey reference number
- the source and date of download
National Primary Health Care Development Agency (NPHCDA) and World Bank. Public Delivery of Primary Health Care Services (PDPHCS) 2002. Ref. NGA_2002_PDPHCS_v01_M. Dataset downloaded from http://microdata.worldbank.org on [date]
Disclaimer and copyrights
The user of the data acknowledges that the original collector of the data, the authorized distributor of the data, and the relevant funding agency bear no responsibility for use of the data or for interpretations or inferences based upon such uses.