GAPTIQ Logo
Request a Briefing
← All signals
Healthcare & BiotechEmerging

The cheapest fix for health anxiety is the one without the trial

Surfaced by the GAPTIQ engineJune 22, 2026 · 5 min read

Health anxiety lands where it costs most: nearly one in five medical outpatients. Individual CBT is proven; the group model that would actually scale has no published randomized trial behind it. The gap is the trial, not the therapy.

The call, up front. Health anxiety is chronic, common and expensive, and it lands where it costs most: nearly one in five general medical outpatients meets the criteria. Individual CBT (cognitive behavioural therapy) treats it, with a large and durable effect. But individual delivery doesn’t scale. The group model would scale at the lowest cost per patient. And it is the one delivery format with no published randomized trial (RCT). The gap is not a therapy. It is the trial that would let payers fund the cheap one.

19.8%Of general medical outpatients meet health-anxiety criteria (≈25% in neurology)
d = 1.01Combined effect of CBT vs control (the comparison group) across 14 trials: the proof, for the individual model
£1,064Saved per patient over 12 months vs usual care: 'strict economic dominance'

The gap

The market funds what it can prove. Individual CBT has the evidence and none of the throughput (the capacity to treat many patients); group CBT has the throughput and none of the evidence. The decision a health system faces is which model to reimburse, and right now the affordable one is unfundable.

Exhibit 1Every delivery format has a published effect size, except the one that scales cheapest
1.62
1.01
0.89
0.80
No RCT
Internet CBTIn-person CBT (14 RCTs)Group ACTInternet ACTGroup CBT
Between-group effect size on health anxiety. Internet CBT, in-person CBT, group ACT and internet ACT (ACT = acceptance and commitment therapy, a related talk therapy) all have published randomized evidence. Group CBT, the format with the lowest cost per patient, has none. The empty bar is the gap.
So what

The bankable asset isn't a new therapy: it's the trial. Run the rigorous group-CBT RCT and you manufacture the reimbursement case for the lowest-cost-per-patient treatment in the category.

Source: Hedman 2011 (BJPsych); Cooper 2017 (Behav Cogn Psychother); Eilenberg 2016 & Hoffmann 2021 (Psychol Med); group-CBT trial NCT02131883, registry re-checked Jul 2026

Bars are between-group Cohen's d (effect size: how big the improvement is) versus control on validated health-anxiety measures, drawn from separate randomized trials: a rough guide to each format's evidence, not a head-to-head comparison. The Group CBT bar is intentionally empty: the one registered group-only CBT trial (NCT02131883) reached completion in January 2019 and has published no peer-reviewed outcomes as of 2026.

Exhibit 2What's missing is evidence for the scalable model, not a new treatment
Scalable, fundable treatment for health anxiety
Individual CBTproven, costly
Large, durable evidence (d ≈ 1.0). No throughput
Internet CBTscaling
Strong effect (d up to 1.6), but people drop out
Group CBT: the missing RCTTHE OPENING
Lowest cost per patient, zero published evidence

Source: GAPTIQ engine: challenge decomposition

Exhibit 3Twenty years of evidence: almost all of it for the model that doesn't scale
  1. 2007Cochrane review: CBT helps health anxiety, but on six small trials
  2. 2014CHAMP RCT (444 patients) proves individual CBT works and lasts, to 5 and 8 years
  3. 2019The one registered group-CBT RCT (NCT02131883) reaches completion
  4. Jul 2026Registry check: still no results posted and no linked publication, seven years after completion. The gap is intact

Source: Cochrane 2007; Tyrer et al. (CHAMP) 2014; ClinicalTrials.gov NCT02131883, re-checked Jul 2026

The move

Whoever runs the rigorous group-CBT trial doesn’t publish a paper: they manufacture the reimbursement case for the lowest-cost-per-patient model in a category that already drains health systems. The clinical work is done; the evidence is the product. The trial that has sat unpublished since 2019 is the whole opportunity.

Source: Outcome of CBT for Patients With Severe Health Anxiety Treated in Group Only: A RCT (NCT02131883), ClinicalTrials.gov, registry re-checked Jul 2026 (status: completed Jan 2019, no results posted). Surfaced by the GAPTIQ engine.

More signals

All signals →
Validated
Professional Services

The fake-citation fix is a checking gate, not a better model

Three model generations in, 2026 deep-research agents still fabricate 3 to 13% of the URLs they cite, and fabricated references in published papers grew 12x in three years. The same 2026 research shows a URL-checking step cuts bad citations to under 1%. The fix is published; it just isn't standard.

Read signal
Validated
Energy

Green steel just cleared a €1.4bn proof point

Steel is 7–9% of global CO2 and the hardest sector to abate. Stegra raising €1.4bn for hydrogen-based production is the signal that the capital wall, not the chemistry, is starting to crack.

Read signal

The desk behind the read

Marius Suteu

Founder & principal analyst · Dankaro Solutions

Runs GAPTIQ coverage end to end: reads the sources, sets the bar a signal must clear, and writes the call.

The GAPTIQ engine

Innovation intelligence

Scans 100+ live sources, ranks the emerging challenges, and stages the evidence behind every read.

The signal feed

Get the next signal in your inbox.

A short, no-noise dispatch when GAPTIQ surfaces a new emerging challenge: the data and the gap, before the market prices it in. No sales, unsubscribe anytime.

Prefer to talk? Request a briefing →