Abstract
Objective. This study investigated the effectiveness of a newly implemented routine outcomemonitoring and feedback system, the Greifswald Psychotherapy Navigation System (GPNS).Methods. 261 patients (M_age = 33 years, 79% female) completed individual cognitive-behavioral therapy (CBT) from 61 psychotherapists. Using Bayesian instrumental variableestimation, we estimated the local average treatment effect (LATE) of using the GPNS onpost-treatment symptom severity and personality impairment. Sensitivity analyses wereperformed to estimate the robustness of our estimates.Results. Using the GPNS significantly reduced post-treatment symptom severity (LATE = -0.11, 95% CrI: -0.18; -0.03) and personality impairment (LATE = -2.65, 95% CrI: -3.72; -1.58). Sensitivity analyses showed that the effects were robust to plausible residualconfounding.Conclusion. The GPNS shows promise for improving outcomes of CBT. Future prospectivetrials should be conducted to obtain a more generalizable estimate of its effect.