THE LACK OF SELECTION BIAS IN A SNOWBALL SAMPLED CASE-CONTROL STUDY ON DRUG-ABUSE

Background. Friend controls in matched case-control studies can be a p otential source of selection bias based on the assumption that friends are more likely to share exposure factors. This study evaluates the r ole of selection bias in a case-control study that used the snowball s ampling method based on friendship for the selection of cases and cont rols. Methods. The cases selected for the study were drug abusers loca ted in the community. Exposure was defined by the presence of at least one psychiatric diagnosis. Psychiatric and drug abuse/dependence diag noses were made according to the Diagnostic and Statistical Manual of Mental Disorders (DSM-III-R) criteria. Cases and controls were matched on sex, age and friendship. The measurement of selection bias was mad e through the comparison of the proportion of exposed controls selecte d by exposed cases (pi) with the proportion of exposed controls select ed by unexposed cases (p2). If p1 = p2, then, selection bias should no t occur. Results. The observed distribution of the 185 matched pairs h aving at least one psychiatric disorder showed a pi value of 0.52 and a p2 value of 0.51, indicating no selection bias in this study. Conclu sion. Our findings support the idea that the use of friend controls ca n produce a valid basis for a case-control study.