Summary Report: Behavioural Surveillance Survey in Kerala, India: Part 4

Executive summary

This report is a summary of the methodology and findings of the Behavioural Surveillance Survey (BSS). This survey was part of the impact assessment Kerala HIV Risk project of HIV prevention activities in Kerala and was conducted in late 1999 for the Kerala State AIDS Control Society (KSACS). Taylor Nelson Sofres Mode (TNS Mode) had implemented the survey.

The BSS in Kerala was intended to track trends in HIV/AIDS related knowledge, attitudes and behaviours in sub-populations in Kerala, especially those at high-risk of acquiring HIV infection. It was also intended to contribute to the larger monitoring and evaluation efforts of the KSACS. The measurements of indicators obtained in this survey served as a baseline from which future trends could be measured. The core indicators on which data were collected included (a) behaviour indicators and (b) knowledge indicators.

The secondary indicators provided information on, among others, prevalence of STI symptoms, treatment-seeking behaviour, messages recalled on HIV and sources of information on issues concerning HIV and practices related to condom use. These indicators were based on those recommended by WHO/UNAIDS for monitoring and evaluation large scale AIDS programmes.

Sampling

The sub-population groups sampled ranged from those with known high-risk behaviours such as female sex workers (FSW) and male clients of female sex workers (MC) to others with varying estimates of high-risk behaviours such as those having sexual intercourse with a non-regular partner. Representatives from KSACS, State Management Agency (SMA), officials from the state health department and other local experts had collectively prepared a list of possible groups that could be included in the survey. Based on available research findings, anecdotal information and reports from NGOs about the groups included in this list, the risk behaviour of each group was estimated and three sub-populations were finally selected for the study. These included male plantation workers (MPW), male auto rickshaw drivers (MARD) and male university students (MUS).

The BSS sampling universe consisted of areas in the state where major concentrations of the selected groups were located. The seven places where the survey was implemented covered cities and some districts. The sample sizes were decided based on the estimated level of key risk behaviours and the degree of confidence (95%) required to detect a significant change (15%) in behaviour over time. The sampling frame was constructed with the assistance of NGOs and individuals working with the specific sub-populations and through mapping exercises. A two stage systematic random sampling procedure was used to select respondents. Clusters were selected in the first stage and respondents were randomly selected in the second stage. Depending on the nature of the group, the clusters were selected with either probability proportional to size (PPS) or equal probability (EP).

Implementation

Qualitative exploratory research was first conducted to facilitate exposure to the target groups and to gain information that would contribute to development of the questionnaire. The standardised BSS questionnaires were adapted for use in Malayalam following intensive pre-testing. The implementation team received training in data collection methods, HIV/AIDS, sex and sexuality, etc. Many NGO partners from the PSH project assisted TNS Mode to understand the socio dynamics of the sex industry and also in the design and implementation of training of field workers. A supervisory team closely monitored the quality. The data was analysed using SPSS. The questionnaires were administered after informed consent was obtained, either by an interviewer or as in the case of students, through self-administration.