[Estimation of need for obstetrical interventions in Morocco. An approach based on the spatial analysis of deficits].

One of the indicators of health system effectiveness with regard to maternal health is the maternal mortality ratio. Measuring this ratio in developing countries is, however, not an easy task since reliable information on mortality is rarely available. An alternative to the maternal mortality ratio measurement, as an indicator of effectiveness, is the assessment of the coverage of obstetrical intervention needs. The authors chose to restrict the notion of "needs" to the obstetrical interventions carried out in order to save a mother's life. Using data from a survey by the Ministry of Health of the Moroccan Kingdom on all the obstetrical interventions carried out in 1989, obstetrical intervention rates for "absolute maternal indications" are analysed according to the mother's origin, by province and urban/rural environment. The spatial analysis of these rates showed large variations in each of the environments (0 to 2.14 % of the expected births in urban areas and 0 to 1.25 % in rural areas) and a significative difference between the rural and urban distributions (median 0.80 % in urban areas versus 0.30 % in rural areas). Applying a reference rate of 1 %, deficits between the expected numbers of needed obstetrical interventions and the observed numbers were calculated for every province in both urban and rural areas. In the whole of Morocco, intervention rates are markedly below what is expected. The spatial analysis of the deficits helps to identify the provinces where the problem is the most prominent in terms of numbers of women whose intervention needs have to be covered. The authors discuss the validity of the reference rate and suggest several strategies to solve the problem. They conclude that the deficits map is a useful tool to decide on priorities for planning and monitoring of strategies to be implemented. The spatial analysis of obstetrical intervention deficits seems to be an instrument both cheaper and more relevant than a maternal mortality estimates survey.