RECRUITING RARE & HARD TO REACH POPULATIONS: A SAMPLING STRATEGY FOR

Recruitment of a probability sample of people living with HIV/AIDS presents a formidable challenge for survey researchers. This is a disease that has flourished at the margins of society, beyond the reach of conventional sampling techniques. Many individuals living with HIV or AIDS have unstable housing situations, are active substance users, or are disenfranchised or alienated from mainstream society. As a result, standard procedures for screening households or assembling lists of eligible respondents are not feasible. Strict rules governing patient/client confidentiality preclude a survey researcher from making direct contact with an HIV infected client through a medical or social service provider. Initial contact with such a sampled client must be made through an intermediary who has some pre-existing relationship with an eligible respondent. One attractive solution is to use a two-stage sample in which a probability sample of HIV service agencies is drawn, followed by a random sample of eligible clients within an agency; this would be followed by an agency liaison attempting to recruit the sampled client. The drawback to the researcher, of course, is that the success of the fieldwork then depends on organizational actors external to the research team. These agency proxies work under varying organizational constraints, and they vary in their interest, motivation, and autonomous capacity regarding the research. Here we describe our fieldwork experience assembling a broadly representative longitudinal cohort of individuals in New York City living with HIV/AIDS using just such a strategy. The survey's major objective was to learn about individuals' service needs, utilization, access, and satisfaction. A further objective was to chart the impact of a multi-dimensional service system on an individual's physical, mental and social well-being. The literature relating the experiences of