A prescription app added to an antidepressant gave modestly higher odds of meaningful improvement

Design

Post hoc analysis of a phase 3 randomised sham-controlled trial (Mirai)

Population

Adults with major depressive disorder on antidepressant monotherapy

Primary outcome

Meaningful within-patient change on MADRS (8- and 10-point thresholds)

Effect

24–47% higher odds of meaningful improvement with Rejoyn vs sham app

Rejoyn is an app-based digital therapeutic authorised by the FDA as an adjunct to antidepressant medication in adults with major depressive disorder, on the basis of the phase 3 Mirai trial against a sham app. This post hoc analysis of Mirai asked what size of change on the Montgomery–Åsberg Depression Rating Scale (MADRS) a patient would recognise as meaningful, then re-analysed the trial using that threshold.

Anchoring MADRS change to clinician-rated global severity and to the patient-rated PHQ-9 gave thresholds of 8 and 10 points. Using them, the Rejoyn group had 24% to 47% higher odds of meaningful improvement than the sham group, and intention-to-treat analyses of response, PHQ-9 and CGI-S also favoured Rejoyn.

The finding supports the original trial rather than extending it. The analysis was post hoc, and the abstract does not give absolute response rates, so the number who benefit cannot be estimated. Rejoyn is not available in India. The broader point for practice is that this trial's 8–10 point MADRS threshold is one reasonable yardstick for 'better' when you read depression trials.

• Treat a MADRS fall of about 8–10 points, the threshold derived in this trial, as a change a patient is likely to notice.
• When reading an antidepressant or adjunct trial, ask how many patients crossed a meaningful threshold, not just the mean difference.
• Rejoyn is authorised by the FDA; there is no Indian regulatory position.
• Keep app-based tools as adjuncts to medication and review, not replacements for them.

Why it matters

Regulators are authorising software as treatment, and clinicians need a way to judge whether the benefit is one patients would feel.

Don't overread it

This is a post hoc re-analysis without absolute response rates; it does not strengthen the original trial's evidence.

The statistics, in plain English

'24–47% higher odds' is a relative figure. If 30% of the sham group improved meaningfully, higher odds of that size would mean roughly 35–39% in the Rejoyn group — a difference of perhaps one extra responder in every 12 to 20 treated. The paper does not give the actual rates, so this illustrates the scale rather than reporting it.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.