Enhancing survival of mothers and their newborns in Tanzania

Abstract: General aims: The main purpose of the present studies was to examine the problem of maternal and perinatal mortality in an upcountry region of a low-income country. This was done by estimating the magnitude of maternal and perinatal mortality, both in the hospital and in the community, through elucidating the underlying causes of maternal and perinatal mortality, and by initiating lowcost interventions and monitoring mechanisms in order to enhance the survival of mothers and their newborns, in Kigoma, Tanzania. Specific aims: To utilise all available evidence to register the causes, contributory factors and real magnitude of maternal mortality in a regional hospital as well as to estimate the magnitude of maternal mortality in the community. To formulate low-cost interventions to address the identified contributing factors to maternal mortality and to follow these interventions over time. To perform regular audits of the causes of maternal mortality in order to elucidate avoidable causes. To monitor and adjust the interventions during the study period, while assessing the impact of these interventions. To investigate the suspected causes of obstetric risk knowledge among community members, health workers, and traditional birth attendants. To assess the utilisation of the simple "three phases of delay model" in the audit of maternal and perinatal mortality. Methods: A retrospective analysis of mortality in the hospital setting utilising all available evidence was undertaken for three years, 1984-1987. The magnitude, causes and contributory factors to maternal mortality were examined in the hospital setting. This led to the formulation of 22 specific, low-cost interventions, which utilized local resources. These interventions were followedup for a period of 7 years. Monitoring was conducted through monthly audit-oriented meetings. Maternal mortality in the community being served by the hospital was assessed utilizing the ',sisterhood method', followed by an assessment of perceptions of obstetric risk among community members, health workers and peripheral staff in order to evaluate factors contributing to further non-reduction of maternal mortality in the hospital. Finally, an assessment utilising the three phases of delay methodology was conducted focussing on the reduction of maternal and perinatal mortality. Results: There was gross underreporting of maternal deaths in the official statistics (849 against 350 per 100,000 live births, respectively). Major causes were haemorrhage, obstructed labour, infections and rupture of the uterus. Several other associated factors comprised lack of equipment, drugs/blood and issues concerning staff and community distrust of the obstetric unit. The application of the 22 specific interventions saw a progressive reduction in the maternal mortality ratio (from 849 to 275 per 100,000 live births) after the 7-year period (p<0.001). This was despite an increase in the number of admissions to the unit (3,000 to 4,296 respectively). Also the case fatality rate for the major causes of death was reduced from 9.2 to 3.1%. However, morbidity increased, which indicated that more sick patients were admitted to the unit. The community assessment undertaken in 2001 revealed the actual MMR at that time to be 447 (urban) and 757 (rural) per 100,000. The result of the assessment in perceptions of obstetric risk revealed low knowledge among the community, staff and traditional birth attendants and that there was distrust in the health system. A final audit using the "three phases of delay methodology" revealed that the major causes of perinatal and maternal deaths occurred in the health system. Conclusions: Maternal and perinatal mortality can be reduced through low-cost interventions available in most low-resource settings. Regular audits of maternal and perinatal deaths can be undertaken in these settings. Low-cost methodologies exist to assess the magnitude and causes of maternal deaths, such as the Sisterhood methodology. To be of value audits must be sustained and used as monitoring mechanisms for service delivery improvements and as managerial tools to reduce maternal and perinatal deaths the "three phases of delay model" is a simple and user-friendly method for the audit of both perinatal and maternal deaths.