Skip to main content

Prediction of therapy dropouts using a model of therapeutic outcome

16 April 2004


Researchers and clinicians alike have struggled to understand why nearly 50% of all psychology clients drop out of therapy (Wierzbicki & Pekarik, 1993). Research in the area of therapeutic outcome has identified several factors that are consistently linked to improved client outcome. In the present study, factors identified as leading to greater therapeutic outcome are examined in relation to retention, specifically, level of improvement that is expected from therapy and extratherapeutic support such as the presence of a primary relationship. Client's demographic variables, stage of change, and level of distress have been found to be related to premature termination. While considering these variables, this researcher also attempts to link extra therapeutic support, and expectations regarding therapy to the area of premature termination. Data from 182 clients (120 females, 62 males) seen at a training clinic were included in the sample with 17.6% identifying as ethnic minorities. Only complete data sets were included in the logistic regression analysis, decreasing the sample size to 97. It was hypothesized that clients with higher levels of distress, lower stage of change, less accurate expectations about the nature of therapy, and less extratherapeutic support would leave therapy at or before the fourth session. In addition, based on the prior literature on premature termination, clients with lower education and from ethnic minority groups were expected to leave therapy sooner. Logistic regression and Chi square analyses identified a dropout rate of 25% in the current sample. However, no predictive pattern of which clients continue or discontinue in therapy by the fourth session was found. Implications of these results are discussed.


Files are restricted to Pacific University. Sign in to view.