Assessing maternal mortality data, A look into the quality of maternal mortality data registration in Kilimanjaro region, Tanzania

Abstract

Maternal mortality in Tanzania is high. The maternal mortality ratio stands at 578 maternal deaths per 100,000 live births. Initiatives to tackle high maternal mortality are underway in Tanzania and the Health Management Information System (HMIS) provides a routine mechanism to monitor progress made in reducing maternal deaths and other health indicators. The quality of this HMIS data, however, is questionable. The HMIS is weak at all levels and comprehensive registration of maternal deaths is a challenge. International health statistics experts have stated that a focus on HMIS in developing countries is the answer for sound and up-to-date health statistics. The objective of this study, therefore, is to assess the quality of maternal death data registration in Tanzania. The results of this study provide insight into ways that the maternal death data collection and reporting process can be strengthened to improve the accuracy and reliability of maternal mortality statistics in Tanzania. A qualitative, case study in the Kilimanjaro region was conducted during June 2010 to examine data registration for maternal deaths. Four levels within the national health system were investigated-health facility, district, region, and national. In-depth interviews with data management workers and health management staff took place at every level. Data analysis was performed qualitatively to answer the research questions and fulfill the study objectives. It was found that maternal deaths primarily take place within the hospital setting and are registered through the maternal death reviews and additionally in the HMIS. The reviews in the Kilimanjaro region closely follow the nationally-recommended guidelines for maternal death review. Several factors were identified to promote (supervision and feedback, training, worker's perception of data collection, and mechanisms to capture community maternal death data) and other factors were found to limit (time taken to complete the report, data inconsistencies and errors, and shortage of staff) data registration. In all, implementation of the maternal death review has brought new hope that maternal deaths will be investigated thoroughly, recorded accurately, and prevented more effectively.