Predicting Couple Therapy Dropouts in Veteran Administration Medical Centers

The present study examined predictors of couple therapy dropout in the VA medical centers using six different dropout criteria. The most accurate dropout definitions included using a statistical modeling procedure to determine whether the client's rate of change at the final session was greater than average of change for all clients; clients who were still demonstrating gains greater than the average rate of change at the final session were considered to have terminated prematurely. A total of 177 couples (354 individuals) who sought therapy in the VA medical centers in Charleston, SC and San Diego, CA were examined. With a few exceptions, demographic variables generally did not predict dropout. A couple's relationship adjustment and response to conflict were significant predictors of dropout. The content of therapy sessions predicted dropout only when dropout was defined, at least in part, by client's rate of change at the final session, suggesting that such methods of defining premature termination are the most sensitive to the therapy process. Therapists' characteristics, including gender and level of experience, did not predict dropout across all six definitions of dropout.