[Comparison of two methods for the investigation of maternal mortality in a municipality of the Brazilian southeast]

Abstract

Purpose: to compare maternal death data from the National Death Information System (DIS), with a death survey of 10 to 49 year-old women at reproductive age (RAMOS), in order to identify sub-notification and to search for causes of maternal death (MD) from 1999 to 2006.

Methods: population based temporal series taken from death certificate (DC) information from Fundação Sistema Estadual de Análise de Dados (SEADE) database, with the death causes codified by the International Classification of Diseases (ICD), tenth revision, and the number of born alive babies (BA). Death was categorized into declared, presumptive MD and non-maternal. The identification of cases was done from a list with both the birth and death dates in the municipal morgues, and further information was obtained in the epidemiological sector of the Municipal Committee of Surveillance of Maternal Death (MCSMD). Information on MD was raised in the DIS. Sub-notification rates in cases of declared and non-declared MD were identified, maternal official death rates (MDR) and the adjusted factor for the period were calculated and corrected, and MD cases were reviewed and classified.

Results: twelve MD were identified, six of them declared and six non-declared. Sub-notification rate was 50%, giving an adjusting factor equal to 2. The official MDR was 14.7 and the corrected one was 29.4 deaths by 100,000 born alive. In most of the cases, the basic causes of death were mistaken. Direct obstetric causes were more prevalent, among them eclampsia and HELLP syndrome, followed by infections.

Conclusions: political and administrative measures are needed for the effective action of MD survey committees. The prevalence of direct obstetric causes indicates failures in maternal and perinatal care.