Xi. Mortality Data Collections: a Review of Integrated Multi- Purpose Household Surveys and Multi-round Demographic Surveys·

The need to study levels, changes and differentials in mortality has been increasingly realized by public health administrators and policy makers since the Second World War. This renewed interest in mortality studies has arisen from the realization that effective measures to reduce mortality can best be taken in the light of adequate knowledge about the health risks in existence, from a need to assess the efficacy of intervention programmes and from recent concern about a possible deceleration in the rate of mortality decline in sorne regions. Many aspects of mortality remain unexplored or 'unexplained for the majority of populations in developing countries. There are several reasons for this, from lack of trained manpower to collect, analyse and interpret mortality data to the inadequacies and complexities of methodologies for collecting and analysing such data. On the whole, the inadequacy of data, whether due to unavailability or unreliability, seems to be the major problem in the developing countries. The traditional sources of mortality data have been civil registration systems and population censuses. The combination of a complete civil registration system and periodic reliable censuses provides much of the information needed for basic mortality studies. However, the problems of incompleteness, delays in aggregating the data and biases introduced by the registrars at the local level in reporting events make the civil registration system of little immediate value in many developing countries. Household sample surveys thus represent, in many countries, the most practicable response to the growing need for data on a variety of subjects. Especially during the past 30 years, several developing countries have resorted to household sample surveys for obtaining information on a number of topics, including fertility, mortality and migration. It has been argued·that a household survey is the most flexible data collection method available to national statistical institutions. 1 The methodology of household sample surveys in terms of basic concepts, field procedures and sampling