Situational factors and behavioral self-predictions in the identification of clients at high risk to drop out of psychotherapy.

Two hundred and eight adults who requested outpatient treatment at a community mental health center were asked to complete brief questionnaires immediately prior to and after their first therapy session. Questionnaires contained items that pertained to self-predicted length of stay in psychotherapy, situational barriers to regular attendance, and past behavior in similar situations (e.g., appointment-keeping with other health care professionals). Therapists also were asked to make predictions with regard to client length of stay. Questionnaire data were analyzed via a series of double cross-validated multiple regression and discriminant function equations, and results indicated that client and therapist predictions of length of stay in therapy were the most stable and consistent correlates of attendance behavior.