Determinants of maternal and non-maternal mortality, evidence from hospital data, Nakuru

Abstract

The main objective of this research was to find determinants and levels of maternal mortality as compared with non-maternal female mortality in the reproductive age set and to establish its significance. The study analyzed tenvariables: socio-cultural, socio-economic, demographic, environmental, medical factors and abortion and their operation variable as related to maternal mortality. The research was carried out in hospitals in Nakuru Municipality though the data is not representative of any administrative area. The data collected will be institutionally based. Involved were, one Government assisted hospital (Nakuru General Hospital-NGH) and two private hospitals (Menengai Nursing and Maternity HomeMN&MH and Nakuru War Memorial Hospital-NWMH). Having considered the surveys studies done worldwide, it was found to be of importance to study maternal mortality in an urban set-up which isof its kind inKenya. The period of study that was analyzed was 5years i.e 1981-1985. The sample size was expected to be 150 for maternal deaths and about 1000 for non-maternal deaths. The data used was collected from the above mentioned institutions. Descriptive statistics, chi-squared method, cross-tabulations and regression techniques were used for analysis. The analysis showed varied results. For instance, age pattern showed different sequences depending on the variables. Each variable was summarised into two graphs i.e for maternal related cases and non-maternal related case. The number of sequences depended on the number of categories for each variable. All variables were analyzed using the chi-square. Most of the null hypothesis have been rejected except for age at first birth, marital status and ante-natal clinic attendance. As for the regression analysis, six equations were analyzed. While equation 1 and 2 did not offer any valid result.s,equation 3 a-nd5 were shown to be the best. In the best equation i.e equation 3, only age residence and ethnicity were analyzed and found to be significantly related to maternal mortality. In this equation, the constant of standard error was low and the constant of beta was greater than 0.5. The findings suggest that maternal mortality is an area that needs much attention and thorough suervey