Summary Report: Behavioural Surveillance Survey in Healthy Highway Project, India: Part 7

Exploratory qualitative phase: Since an all-India BSS for the HHP was being conducted for the first time, an exploratory qualitative phase was necessary to (a) collect information that would aid development of a structured questionnaire, (b) identify sensitive issues among various target groups and (c) identify methods that would help access and interview the target groups. A total of 56 in-depth interviews including 18 with truck drivers, 24 with FSWs and 14 with stationary workers were conducted in 12 centres across the country. Lessons learned from the exploratory phase were used for developing the protocol.

Pilot interviews: Draft questionnaires were tested during the pilot phase. The main purpose of this phase was to test whether the questions and their language were uniformly understood by the respondents and the interviewers. The pilot phase had also helped standardize ways to approach each target group and to prepare implementation plans. A total of 620 interviews in the four target segments were conducted during the pilot phase.

Mapping study: The existing database on trucking halt points on major national highways was considered to be incomplete and therefore a decision was taken to update it. More than 350 interviews were conducted with truck drivers from various parts of the country to obtain the list of halt points on each highway. This data was collated and cross-checked by an independent consultant.

Quantitative phase: The halt points to be covered in the survey were selected by probability proportional to size (PPS) sampling method. Sampling frames were prepared for truck drivers and FSWs for each of the six zones and PPS sampling was done within each zone. An all-India sampling frame was used for truck helpers and stationary workers. The cluster size was fixed at 20 for male target groups and 10 for FSWs.

The survey instruments were translated into regional languages after extensive field testing.

Indicators

Behavioural surveillance consists of repeated cross-sectional surveys in selected population groups, which provide measurements on certain indicators. These indicators were based on those recommended by WHO/UNAIDS for large scale/national AIDS programmes. In the baseline wave in HHP BSS, data were collected on two main indicators including (a) behaviour indicators and (b) knowledge indicators. In addition, there were secondary indicators such as incidence of STI symptoms, treatment-seeking behaviour, practices related to condom use. The list of main indicators and the total responses for each are given in the Annex.

The measurements of the core indicators provided information on the proportion of the total respondents who had reported a particular behaviour or knowledge.

Sample size

The targeted sample sizes for the different groups were as given in Table 1. The sample sizes achieved were 4,811 for truck drivers, 1,504 for FSWs, 1,200 for truck helpers and 1,201 for stationary workers.

The key assumption in arriving at the sample size was the proportion engaging in high-risk behaviour and the percentage of change that it would be able to detect over time. In other words, it was the proportion engaging in unprotected sex with a non-regular partner per year in each group. The estimated proportion of high-risk behaviour in the target segments was 25% among truck helpers and stationary workers each, 35% among truck drivers and 100% among FSWs. The sample size was adequate to detect a 15% change of behaviour among these segments.

Issues in behavioural data collection �Validity of self reported data on sexual behaviour

Many questions had been raised about the validity of self-reported data on sexual behaviour. Growing experience in collecting data on sexual behaviour had indicated that the extent to which people answered questions openly and truthfully depended on factors such as the setting of the question, privacy and confidentiality, attitudes and profile of the interviewer, etc. While it is not possible to validate data on sexual practices by direct observation, it is possible to triangulate them with data from other sources to see if the picture presented is consistent and credible. Still, some misreporting of risk behaviour could occur and true levels of risk might well be under or over reported.

Table 1: Sample size for various target groups