RESUMEN
Context: Maternal mortality is considered a key health indicator of Maternal and Child Health. Considering the fact that complications are preventable and most of them are modifiable, the study has been planned to analyse maternal deaths in order to suggest recommendations for preventing it. There are various delays according to the three-delay model at primary and secondary level; therefore, interventions are needed at those levels to prevent maternal deaths. Aims: To determine the various direct and indirect causes of maternal deaths, analyse the association of medical and social factors with maternal deaths and ^to determine the predictors of maternal deaths. Settings and Design: Hospital-based retrospective cross-sectional study of all the maternal deaths occurring in the last 4 years at a tertiary health care facility. Methods and Material: Data were collected from the Facility Based Maternal Death Review forms. Statistical Analysis Used: Data were entered and analysed by IBM SPSS version 25.0 software. Results: For maternal deaths, direct obstetric causes were responsible in 128 (74.4%) and indirect causes in 45 (26.2%) cases followed by unspecified causes in 78 (45.3%) and 1 (0.6%) coincidental cause. Statistically significant associations were observed between maternal death and period of gestation, mode of delivery and outcome of delivery (P = 0.12, P = 0.04 and P < 0.001, respectively). Conclusions: The health professionals of primary and secondary level should be well equipped to diagnose the complications and to manage it as early as possible. Thus, maternal mortality rates can be decreased to significant level.