JMIR Publications

ABSTRACT

Background:

Digital therapeutics (DTx) are increasingly used to address gaps in access to evidence-based treatments for depression, yet rigorous trials isolating content-specific effects remain limited.

Objective:

This study evaluated a prescribable, web-based DTx (edupression.com®) integrating cognitive-behavioral and psychoeducational content, measurement-based care elements, and structured mood monitoring in adults with mild-to-moderate depression.

Methods:

We conducted a single-center, rater-blinded, parallel-group randomized controlled trial delivered entirely remotely from a coordinating center in Vienna, Austria, with participants residing in Austria and Germany recruited online and through clinical referral. Adults aged 18-65 years with mild-to-moderate unipolar depression were allocated 1:1 to the intervention group (IG) or an active sham control (ASC), both in addition to unrestricted usual care. Both arms were stand-alone, unguided self-help applications. The ASC matched the intervention in interface and engagement intensity but excluded therapeutic content and feedback. The primary endpoint was change in self-administered PHQ-9 from baseline to week 12, analyzed in an intention-to-treat framework using baseline-adjusted linear regression with multiple imputation (MICE); reference-based multiple imputation (RBMI) was prespecified as a conservative sensitivity analysis. Safety was assessed in all participants.

Results:

Of 907 individuals who expressed interest, 250 were randomized (IG=125; ASC=125). Observed week-12 PHQ-9 data were available for 40/125 (32%) IG and 56/125 (44.8%) ASC participants. Non-adherence did not differ significantly between arms (44.0% IG vs 39.2% ASC). In the primary intention-to-treat analysis, the intervention showed a greater reduction in PHQ-9 than the ASC (adjusted mean difference −4.0 points, 95% CI −5.8 to −2.1; Cohen d=0.84; P=.004). The RBMI sensitivity analysis yielded a smaller but statistically significant difference (adjusted mean difference −1.1 points, 95% CI −1.7 to −0.5; Cohen d=0.20; P=.001), and the complete-case estimate was −4.8 points (Cohen d=0.99). Response was more frequent in the IG than the ASC (52.8% vs 15.4%; number needed to treat 2.7), as was remission (24.3% vs 7.0%). Depression literacy improved more in the IG (between-group difference 1.4 points, P<.001) and remained significant under the conservative RBMI analysis. No serious adverse events, suicide attempts, or deaths were reported. Clinically relevant deterioration was less frequent in the IG (4.0% vs 11.2%). Emergent suicidal ideation on the daily mood chart did not differ between arms (6.9% vs 10.9%; risk difference −4.0%, 95% CI −15.7% to 7.2%).

Conclusions:

In this outpatient sample, the DTx was associated with greater short-term reductions in depressive symptoms and improvements in depression literacy than a structurally matched active sham. The design enabled a conservative, content-specific efficacy test with no evidence of intervention-related harm. Limitations include the selective outpatient sample, the high proportion of unobserved primary-endpoint data, and a 12-week observation period. These findings support the use of structured, unguided DTx as an accessible adjunct within routine outpatient care. Clinical Trial: ClinicalTrials.gov NCT04839822; German Clinical Trials Register DRKS00032121.