Utility of local health registers in measuring perinatal mortality: a case study in rural Indonesia

Abstract

Background: Perinatal mortality is an important indicator of obstetric and newborn care services. Although the vast majority of global perinatal mortality is estimated to occur in developing countries, there is a critical paucity of reliable data at the local level to inform health policy, plan health care services, and monitor their impact. This paper explores the utility of information from village health registers to measure perinatal mortality at the sub district level in a rural area of Indonesia.

Methods: A retrospective pregnancy cohort for 2007 was constructed by triangulating data from antenatal care, birth, and newborn care registers in a sample of villages in three rural sub districts in Central Java, Indonesia. For each pregnancy, birth outcome and first week survival were traced and recorded from the different registers, as available. Additional local death records were consulted to verify perinatal mortality, or identify deaths not recorded in the health registers. Analyses were performed to assess data quality from registers, and measure perinatal mortality rates. Qualitative research was conducted to explore knowledge and practices of village midwives in register maintenance and reporting of perinatal mortality.

Results: Field activities were conducted in 23 villages, covering a total of 1759 deliveries that occurred in 2007. Perinatal mortality outcomes were 23 stillbirths and 15 early neonatal deaths, resulting in a perinatal mortality rate of 21.6 per 1000 live births in 2007. Stillbirth rates for the study population were about four times the rates reported in the routine Maternal and Child Health program information system. Inadequate awareness and supervision, and alternate workload were cited by local midwives as factors resulting in inconsistent data reporting.

Conclusions: Local maternal and child health registers are a useful source of information on perinatal mortality in rural Indonesia. Suitable training, supervision, and quality control, in conjunction with computerisation to strengthen register maintenance can provide routine local area measures of perinatal mortality for health policy, and monitoring of newborn care interventions. Similar efforts are required to strengthen routine health data in all developing countries, to guide planned progress towards reduction in the local, national and international burden from perinatal mortality.

Figure 1

Distribution of pregnancies identified across local health registers in the study sample of villages in Pekalongan, 2007.