2. The individual questionnaire for evaluation

The questionnaire for the measurement of four prevention indicators [1] is divided into eight sections: identification; background characteristics; marriage and regular partnerships; non-regular and commercial sex; condoms; STDs and health issues; knowledge of AIDS; and risk perception, behaviour change and attitudes toward persons with HIV/AIDS.

The main purpose of the sections on marriage and sexual partners was to measure sexual risk behaviour for each individual respondent. The sequence of questions was carefully designed to progress from consideration of marriage to other regular partnerships and finally to more transient and potentially high risk sexual relationships. Clearly, these distinctions are multi-dimensional and complex. Any comprehensive classification would have to take into account not only the persistence of the relationship over time but also the associated expectations and obligations.

A blunt and somewhat oversimplified solution was reached: a regular partner was defined as a person with whom the respondent has a relationship involving sexual intercourse for a period of 12 months or more; anyone else is a non-regular partner. Among non-regular partnerships, a commercial sex encounter was defined as a relationship where sex was exchanged for money and where partners did not know each other [4].

The underlying principle was that the concept of non-regular partnerships can be conveyed clearly by interviewers only after careful definition and measurement of regular partnerships. Sexual intercourse within such regular relationships carries a low risk of HIV infection, provided that the partnership is mutually exclusive (and of course that neither party is already infected).

This questionnaire design has proved to be successful in identifying who was at increased risk of HIV, with what type of behaviour and where. These surveys have been very informative for the evaluation and development of prevention strategies. However, programme designers and researchers may want to go beyond those objectives and be interested in better explaining why, within the same “risk behaviour” groups, certain people are more likely to be HIV infected than others. Insights into that question will require information on how different individuals are connected via risk networks.