Therapy apps help people cut drinking, but they never beat a human therapist

Cognitive behavioural approaches are among the most widely used treatments for alcohol and other drug problems, but access to a trained therapist is a real bottleneck for many people who want to drink less. Delivering the same techniques through apps and online programmes promises the benefit without the waiting list. Earlier reviews suggested these digital versions do reduce substance use, but a big question remained open: do they also improve the rest of life, things like day-to-day functioning and quality of life, or do they only move the consumption numbers? This analysis set out to answer that across the full range of outcomes.

The researchers searched the literature through October 2024 for clinical trials of digitally delivered cognitive behavioural interventions for adult alcohol and other drug use. They ended up with 65 randomised trials reported across 110 publications, yielding 753 individual effect estimates. Every available study endpoint was eligible, and each was sorted into one of 13 outcome types, such as abstinence, quantity consumed, cognitive outcomes and quality of life, then grouped into two broader classes: consumption and psychosocial functioning. Pooled estimates were then compared by outcome class and by the type of control the digital programme was tested against.

Used as a stand-alone treatment against a minimal control, the digital programmes produced small but statistically significant benefits for both consumption (g=0.27) and psychosocial outcomes (g=0.16). Added on top of usual care, they improved consumption (g=0.23) but not psychosocial functioning. Tested against another digital or in-person intervention, or against cognitive behavioural therapy delivered by a therapist, no advantage was observed. Within the digital treatment groups themselves, improvement from baseline was common, with 60% to 80% of participants improving, though effect sizes and disagreement between studies varied depending on the outcome examined.

This is described as the first comprehensive look at digital cognitive behavioural interventions across functional endpoints rather than substance use alone, so it extends earlier reviews that stopped at consumption. It confirms the established picture that these programmes beat minimal support, and adds the useful nuance that some psychosocial benefits appear too, at least when the programme stands alone. It also sharpens a limit: nothing here suggests the digital format outperforms a human therapist or other active treatments. Open questions remain about why the studies disagreed so much on some outcomes and which specific outcome types respond best.

For anyone thinking about cutting back on drinking or drug use, this is a reason to take well-designed digital programmes seriously as a genuine option rather than a gimmick, particularly where a therapist is out of reach or unaffordable. The honest sizing matters, though: the average benefits were small, and the programmes did not outperform seeing a real therapist or other active treatments. It is worth knowing that the gains were clearest for how much people consumed, while the wider benefits to daily functioning were less consistent, especially when the app was simply bolted onto existing care.