(1994; 54 pages)
2.2.2 Structured household interviews
The quantitative data was collected through a users’ survey carried out on the basis of a pre-coded questionnaire (see Appendix 1). The questionnaires were filled in on the basis of personal interviews conducted by the research assistants. The questionnaires were administered twice to 360 households in each region. Two weeks after the initial interview, households were visited again and the prevalence of injection use in the two week recall period was recorded. Data collected based on the two week recall period served as an accurate measure of the extent of injection use and also served to reduce memory bias in respondents since the interviewer could refer to a specific event: the last time the household was visited. There were no families lost to follow-up. Apart from establishing the prevalence of injections based on the two weeks recall, the questionnaire also covered questions about when was the last time that anyone in the household received an injection.
Hypothetical questions concerning five tracer conditions were posed5. Two were used in all the countries in the comparative study: a case of cough and common cold, and a case of acute diarrhoea of less than five stools a day in child under five. Three other country specific tracer conditions were used in the Ugandan study: intestinal worms, vomiting and fever (defined as perceived rise in body temperature). During the follow-up visit questions about illness episodes and treatment strategies were also posed. Symptoms were recorded and later classified using categories of tracer conditions and ‘others’. This enabled us to compare the answers to hypothetical questions about illness with actual illness episodes experienced. When severalhousehold members had been ill and treated each episode was recorded, specifying the symptoms. The limitations and methodological difficulties of this method will be discussed in 2.2.3.
5 All tracer conditions selected did not refer to complicated and/or serious conditions which may warrant an injection. These were conditions considered by doctors to be self-limiting or treatable with oral therapy.
The survey also aimed at establishing the various sources of injections, the possession of injection equipment and injectables in the home, and the magnitude and nature of injection complications experienced within the household.