What it means when the FDA 'clears' a prescription app to treat a disease

When a company says its app is "FDA cleared" to treat a condition, it means the app went through the same authorisation pathway as a physical medical device and produced clinical evidence that convinced the agency — a genuine bar, not marketing shorthand [s2]. What clearance does not guarantee is that the app works dramatically, that a doctor will prescribe it, or that anyone will pay for it. The category of prescription digital therapeutics has struggled less with the regulator than with the last of those [s3].

What "cleared" actually certifies

A handful of software products have been authorised by the FDA to treat disease. reSET, a 12-week programme for substance use disorder, was granted marketing authorisation through the De Novo pathway in September 2017; its sibling reSET-O for opioid use disorder followed as a 510(k) clearance in December 2018; Somryst, a digital cognitive behavioural therapy for chronic insomnia, was cleared in March 2020; and EndeavorRx, a video-game-based treatment for paediatric ADHD, received De Novo authorisation in June 2020 [s2].

Those pathways matter. The De Novo route is how the FDA authorises a novel, low-to-moderate-risk device that has no existing equivalent, and it establishes a new device type; a 510(k) clears a device by showing it is substantially equivalent to one already on the market [s2]. Both require the device to be safe and effective for its stated use. This is a different and higher bar than the one that governs a "general wellness" app, which makes no disease claim and needs no review — a distinction that has become the regulated line for wearables too. The word to watch is the claim: an app that says it treats insomnia is regulated; one that says it helps you "sleep better" generally is not.

What the trial data actually showed

Clearance rests on trials, and the trials are worth reading rather than trusting. The pivotal study behind EndeavorRx, STARS-ADHD, was a randomised, double-blind, controlled trial in 348 children aged 8 to 12 with confirmed ADHD, who played either the therapeutic game or a control game for 25 minutes a day, five days a week, for four weeks [s1]. The primary outcome was the change on a computerised attention measure, the Test of Variables of Attention Attention Performance Index [s1]. The treatment group improved by a median of 0.88 points (95% CI 0.24 to 1.49; p=0.0060), against a mean change of 0.93 [s1].

That is a statistically significant result on a computerised attention task, and it is also a narrow one: it is a change on a laboratory index, in medication-free children, over four weeks, and the trial was not designed to show that the game improves day-to-day functioning at school or home [s1]. This is the recurring texture of the category — the effects are real, specific and modest, and they are measured on the outcome the software was built to move. A companion point recurs across digital therapeutics: the effect sizes are not comparable across products, and the ones that involve a clinician tend to outperform the fully self-guided versions.

The wall the category keeps hitting

If regulation were the binding constraint, cleared products would be thriving. They have not been, and a 2026 analysis locates the reason in policy and payment rather than approval. Reviewing the landscape for prescription digital therapeutics, its authors describe a field where "despite rising interest and encouraging evidence of clinical benefits, the policy landscape remains fragmented," with jurisdictions varying widely in regulatory approach, reimbursement pathway and route to clinical integration — creating "uncertainties for developers, payers and healthcare providers" [s3]. Their proposed fix is a maturity framework for assessing how ready a health system is to adopt these products at all, built by expert consensus set at 70% agreement [s3].

That is the quiet story of prescription digital therapeutics: a product can clear the FDA, publish a positive trial, and still fail because no reliable billing code, formulary slot or prescribing habit exists to carry it to patients. The bottleneck moved downstream of the regulator.

What it means for a reader

"FDA cleared" on a health app is meaningful and specific — it signals a disease claim backed by trial data and agency review, which most apps in a store do not have [s2]. It is not a promise of a large effect; the pivotal evidence tends to be a modest, targeted improvement on the outcome the software was designed to change [s1]. And clearance says nothing about whether a clinician will prescribe the product or an insurer will cover it, which is where the category has actually stalled [s3].

Sources

• [s1] A novel digital intervention for actively reducing severity of paediatric ADHD (STARS-ADHD): a randomised controlled trial — The Lancet Digital Health (2020)
• [s2] FDA device clearance records for reSET, reSET-O, Somryst and EndeavorRx — US Food and Drug Administration (2017–2020)
• [s3] Developing a policy maturity model for prescription digital therapeutics based on expert consensus — BMJ Open (2026)