How do patients' attendance profiles influence psychotherapy terminations? A preliminary exploration

Objectives: In this study, we intended to examine the differences in patients' attendance profile during psychotherapy process among three patient groups who ultimately completed treatment, prematurely terminate treatment with or without informing therapist in advance. We also attempted to find an effective predicting model to discriminate those three groups. Methods: We used the data of the process notes of 982 time-limited individual therapy sessions over five years, composing of 97 adult patients who were treated by a therapist in a Taiwan psychiatric center. We analyzed 10 attendance variables retrospectively in coded session by session. Results: Three patient groups existed with notable differences in their punctuality, cancellation, no-show, treatment-stage attendance, and overall attendance rates. But no clear difference was observed in the other six variables (time change, early or delayed leave, between-session contact, and evaluation-stage attendance rates). Cancellation and no show were effective variables to distinguish three patient groups. The overall correct predicting rate of the regression model was 67%, and the individual correct predicting rate for treatment completers, informers, and noninformers was 75%, 52.8%, and 75.9%, respectively. No show as a variable clearly discriminated completers from noninformers and informers from noninformers, whereas cancellation variable clearly distinguished completers from informers. Conclusion: Patients with and without informing their therapists before premature termination showed different characteristics. Cancellation and no show in each therapy session were pre-dropout warning signs for later premature termination. Those two behaviors had substantial implications and strong predicting power on the patients' termination conditions