# Service delivery and coverage in primary healthcare in a community-health project in Ibadan, Nigeria

## TL;DR

This study in Nigeria found that while most health records in community clinics are accurate, some important events are still missed.

## Contribution

The study provides empirical evidence on the accuracy and completeness of primary healthcare records in a Nigerian community.

## Key findings

- Record accuracy was high, with over 80% of selected names traceable across registers.
- Some women and children were missed in the records, indicating incomplete coverage.
- There was a steady improvement in maternal and child health service coverage over four years.

## Abstract

Standard health-service delivery aimed toward improving maternal and child health status remains elusive in Nigeria because of inaccuracies in data documentation leading to a lack of relatively stable evidence.

Through a community-health project, this study tested the accuracy of record keeping in primary healthcare services in nine clinics run in Ibadan, Nigeria.

A validation exercise was performed through a sample of the 10 most recent names extracted from three registers maintained by each clinic.

A review of the register covering a period of four years showed a steady increase in: fully-immunised children, registration for antenatal care during the first trimester of pregnancy, the number of women who attended antenatal care at least three times, the overall number of women who booked for antenatal care and women who delivered in Eniosa Community-Health Project facilities over the four-year period. It was possible to trace 86% of those selected from the antenatal care register, 88.9% of those from the birth register and 81.1% of those from the immunisation register. Four women who should have been included for antenatal care, seven who had delivered (but were not in the register) and 13 who reportedly received immunisation but were not listed were found during the validation exercise.

This study concludes that the names appearing in the register are likely to represent valid events, but that the registers did not capture all such events in the community.