Attrition bias in longitudinal research involving adolescent psychiatric outpatients.

This study investigated whether the degree of difficulty contacting participants at follow-up in a longitudinal study of adolescent psychiatric outpatients is associated with baseline and/or follow-up Axis I and II psychopathology and sociodemographic variables. At baseline, 101 participants 15 to 18 years old were assessed using standardized diagnostic instruments, and 97 were reinterviewed, face-to-face, at 2-year follow-up. A hierarchical tracking strategy and meticulous follow-up contact log were used. More than one quarter of the sample required multiple tracking efforts to be located and interviewed. The presence and number of Axis I and II disorders at 2-year follow-up was significantly associated with follow-up contact difficulty. Baseline psychopathology and sociodemographic variables were not associated with follow-up contact difficulty. The findings indicate that longitudinal studies of adolescent outpatients that compare dropouts with completers utilizing baseline characteristics are likely to underestimate the extent of psychopathology at follow-up.