Cost-Outcome Descriptive Analysis of an Internet Based Treatment for Depression, Social Anxiety, and Panic Disorder: A Post-Implementation Open Naturalistic Design Study

OBJECTIVES: Economic evaluations of implementation strategies have increased, yet few assess post-implementation evidence. This study performs a cost-outcome analysis of an internet-based treatment for mental health disorders, ‘eTreatment’, after its national implementation in Norway, using four health enterprises as cases. Each enterprise includes several clinics and has organized eTreatment differently. Two enterprises use a centralized approach with one clinic handling all referrals and therapy. The other two use a decentralized approach with all clinics collaborating on referrals and therapy. Additionally, decentralized enterprises accept only GP referrals, while centralized ones also allow direct patient contact.

METHODS: Patients reported on clinical and economic outcomes at baseline and post-treatment (September 2021 to June 2024). Program costs, including maintenance, training, and therapist time, were estimated using administrative records supplemented with expert opinion from clinic leaders and therapists. Regression models will examine the association between location-level differences and clinical and economic outcomes.

RESULTS: These results are preliminary and based on a pilot sample from one location. At ISPOR, we will present results from the complete dataset that includes all the service locations.

Pilot data from one service location (n=194) indicate that eTreatment improved patients’ clinical and economic outcomes. Post-treatment, patients showed improvement in primary clinical indicators for depression and anxiety ((∆ score, s.d.) ∆PHQ-9: -2.91 (5.8), ∆GAD-7: -2.18 (4.5)). Percentage of patients using any medicine decreased from 35% to 29%, whereas sickness absence reduced by 4.7 days (s.d. 9.6). Work and social adjustment scale decreased by 3.7 points (s.d. 6.5), representing a reduction in impairment. Health related quality of life (EQ5D) increased by 0.05 points (s.d. 0.19). Program cost per patient for this location was €903.14 (s.d. 55.82).

CONCLUSIONS: Post-implementation analyses can assess observed differences in outcomes between locations in the light of organizational differences. This is of value for policy considerations seeking to improve efficiency and resource allocation.