Outpatient Psychotherapy Improves Symptoms and Reduces Health Care Costs in Regularly and Prematurely Terminated Therapies

Background: In view of a shortage of health care costs, monetary aspects of psychotherapy become increasingly relevant. The present study examined the pre-post reduction of impairment and direct health care costs depending on therapy termination (regularly terminated, dropout with an unproblematic reason, and dropout with a quality-relevant reason) and the association of symptom and cost reduction. Methods: In a naturalistic longitudinal study, we examined a disorder heterogeneous sample of N = 584 outpatients who were either treated with cognitive-behavioral, psychodynamic, or psychoanalytic therapy. Depression, anxiety, stress, and somatization were assessed with the Patient Health Questionnaire (PHQ). Annual amounts of inpatient costs, outpatient costs, medication costs, days of hospitalization, work disability days, utilization of psychotherapy, and utilization of pharmacotherapy 1 year before therapy and 1 year after therapy were provided by health care insurances. Symptom and cost reduction were analyzed using t-tests. Associations between symptom and cost reduction were examined using partial correlations and hierarchical linear models. Results: Patients who terminated therapy regularly showed the largest symptom reduction (d = 0.981–1.22). Patients who dropped out due to an unproblematic reason and patients who terminated early due to a quality-relevant reason showed significant but small effects of symptom reductions (e.g., depression: d = 0.429 vs. d = 0.366). For patients with a regular end and those dropping out due to a quality-relevant reason, we observed a significant reduction of work disability (diff in % of pre-test value = 56.3 vs. 42.9%) and hospitalization days (52.8 vs. 35.0%). Annual inpatient costs decreased in the group with a regular therapy end (31.5%). Furthermore, reduction of symptoms on the one side and reduction of work disability days and psychotherapy utilization on the other side were significant correlated (r = 0.091–0.135). Conclusion: Health care costs and symptoms were reduced in each of the three groups. The average symptom and cost reduction of patients with a quality-relevant dropout suggested that not each dropout might be seen as therapy failure.