Evidence-Based Mental Health at Scale: Benchmarking Retrospective Cohort Study of a Digital Employee Benefits Program for Depression and Anxiety

Online Journal of Public Health Informatics · Published 2025-09-23 · DOI 10.2196/72999

Free full text

Authors being retrieved — see the publisher record. https://doi.org/10.2196/72999

Abstract

BackgroundDepression and anxiety affect millions worldwide; yet, many people face barriers to timely and effective mental health care, underscoring the need for scalable, high-quality interventions. ObjectiveThis study aimed to evaluate the clinical effectiveness and quality of a centralized, employer-sponsored mental health program in treating depression and anxiety during a period of rapid growth in access to mental health care. MethodsThis retrospective cohort study included participants using a digital mental health benefit (Spring Health), sponsored by 589 US employers from 2021 to 2024. Participants had access to therapists, psychiatrists, and care navigators. Primary measures were clinical effectiveness (treatment duration, Patient Health Questionnaire 9-item depression scale, Generalized Anxiety Disorder 7-item Scale) and clinical outcomes (reliable change, recovery, and remission). Outcomes were benchmarked to meta-analytic results of evidence-based therapy. ResultsA total of 52,929 adult participants started therapy from among 6868 providers during the study period, scored positive for depression or anxiety, and had at least one mental health assessment before and during treatment. Depression symptoms decreased with each log-day in treatment, resulting in a total reduction of 6.91 (95% CI –6.84 to –6.98) points at one-week post treatment, corresponding to a large effect size (d=1.61; 95% CI 1.60-1.63), significantly greater than the meta-analytic pre-post benchmark for psychotherapy (effect size difference=0.13, z=15.6, P<.001). Anxiety symptoms also decreased, resulting in a total reduction of 6.01 points (95% CI –5.95 to –6.08), corresponding to a large effect size (d=1.82; 95% CI 1.80 to 1.84), significantly greater than the meta-analytic benchmarks (effect size difference=0.64, z=61.9, P<.001). White participants and participants of color had similar outcomes. Logistic regression showed 92.3% (95% CI 92.0% to 92.5%) of participants’ symptoms (in depression or anxiety) reliably improved, and 61.7% (95% CI 61.1% to 62.4%) achieved remission by 1-week posttreatment. ConclusionsAmong a large and diverse sample, using a digital mental health benefit with a centralized system of care produces clinical outcomes in depression and anxiety significantly greater than what is typically observed through meta-analyses of psychotherapy. By using data to monitor, incentivize, and improve quality of care, the clinical outcomes outperform or equal benchmarks as a growing number of individuals across race, gender, and age access mental health care.

Abstract from DOAJ. Public domain (CC0 1.0).

Read the article at the publisher →

Publication details

Year
2025

Related articles