Microcomputer applications in health population surveys: experience and potential in developing countries.

Abstract

Regrettably, because the content and technology are evolving so rapidly the literature on the application of microcomputer technology to health and development is either out of date, or hard to find (e.g. unpublished reports from consultants working with development agencies or universities). The time needed to review and publish an article or a book dictates that by the time it is published, any statements about current hardware will be obsolete. For example, a recent volume by Ingle et al., quite sound from the theoretical and practical point of view is already out of date with respect to hardware details. For that reason no hardware suggestions are made in this article. On the other hand, detailed reports of experience with first-generation microcomputers are very useful to the individual planning a developing country application. From the literature and our own field experience with respect to microcomputer applications in health/population surveys in developing countries, the following are useful summary points for projects currently going into the field. Careful planning is absolutely necessary if microcomputer technology is to be applied successfully to field data collection problems. Search out organizations with similar experiences. Collaboration among staff at all levels is essential for a system to be installed and utilized effectively. The particular technology combination selected must be adapted to the field situation. Remember survey complexity and size. Training is essential, and should be on-the-job and oriented towards specific tasks. If time consuming and repetitive tasks are given over to the microcomputer the technology will be readily accepted and fully utilized. To the extent that a community of users can be formed, the entire adoption process will be stimulated and results improved. Most maintenance and some repairs should be carried out in the field by local project personnel. For most survey and data acquisition problems some custom programming will be needed; this requires a good working knowledge of a powerful programming language (although this is likely to change in the near future). For descriptive statistics and small data management problems off-the-shelf software is appropriate but can be expensive. Compare options. Standardization and compatibility must be considered in all applications. This is important!(ABSTRACT TRUNCATED AT 400 WORDS)