More prescriptions, clinicians, and appointments are visible signs of healthcare activity. Q3 2026 showed why they are not interchangeable with recovery. England published prescribing and treatment results, US survey estimates were recalibrated, Switzerland reported psychiatric quality measures, and new trials complicated familiar breakthrough claims. The practical question was what each additional unit of care changed for patients.
Key findings
95.5 million and 9.1 million: antidepressant prescription items and identified patients in England in 2025/26, released July 23. Neither is a count of people recovered. [1]
50.1%: England’s NHS Talking Therapies recovery measure in 2025/26, versus 50.5% the preceding year, among eligible completed referrals. [4]
81%: reported growth in Abu Dhabi’s mental-health workforce since 2022, announced in August. This is capacity, not a recovery change. [5]
160 locations and 89,012 adult treatment cases: the scope of Switzerland’s 2025 psychiatric measurements released in September. Cases are not necessarily unique patients. [6]
England: prescribing grew faster than identified patients
Calculated from the published totals. Identification coverage changes affect patient comparisons. The chart does not establish dose, treatment duration, prevalence, or quality. [1]
One patient can receive repeated items. Antidepressants also have more than one indication. A dispensing record cannot supply the diagnosis or show what happened after treatment.
For families purchasing private care, the stronger question is how progress is reviewed. Population trends are not instructions to change an individual’s medication.
USA: compare the revised history, not old headlines
The July NSDUH release introduced a new model for mental-illness estimates and recalculated earlier years. It reported 21.0% of adults with any mental illness in 2025, versus 21.5% in the recalculated 2024 series. The 2021 to 2025 trend was classified as no change. [2]
| 2024 adult measure | Earlier model | Updated model |
|---|---|---|
| Any mental illness | 23.4% | 21.5% |
| Serious mental illness | 5.6% | 7.8% |
The categories overlap. Their different revisions do not represent people becoming healthier or sicker between two observations. Source: SAMHSA’s method report. [3]
The recalculated young-adult any-mental-illness estimate fell from 33.2% in 2021 to 28.6% in 2025. That positive direction deserves reporting without implying that the same individuals were followed into recovery. [2]
England and Switzerland measured different clinical outcomes
NHS Talking Therapies reported 677,595 completed courses, 50.1% recovery, and 47.9% reliable recovery in 2025/26. Recovery applied to eligible completed referrals meeting the starting threshold, not everyone referred. It was a discharge measure, not twelve-month recovery. [4]
Switzerland’s ANQ reported average symptom improvement between admission and discharge, with indicators broadly stable against the previous year. It explicitly warns against treating its measures as a simple hospital ranking. [6]
| Source | Population and outcome | Limit |
|---|---|---|
| NHS Talking Therapies | Eligible completed referrals crossing a symptom threshold | Not every referral or long-term follow-up |
| ANQ psychiatry | Inpatient cases with admission-to-discharge measures | Different setting, population, and indicators |
| Private-provider claim | Depends on the provider’s definitions | Requires comparable patients, measures, and timing |
These are valuable reporting systems, not interchangeable performance scores. Private services should be equally clear about their denominators and follow-up.
Abu Dhabi expanded capacity and its measurement framework
The August announcement identified 640 mental-health professionals across 143 facilities, including thirteen specialized centers. It also described nineteen patient-outcome indicators and training for more than 1,930 family physicians. [5]
These are Abu Dhabi figures, not UAE-wide totals. Adopting an indicator framework is not the same as publishing improved values for each indicator.
The useful next report would connect access, actual treatment, patient functioning, and longer-term results. Announced plans to double capacity should remain plans until implementation is established.
The newer oral alcohol trial missed its primary craving endpoint
The July oral-semaglutide trial randomized fifty adults with moderate or severe alcohol use disorder. Its primary laboratory-craving comparison was nonsignificant, while selected secondary drinking outcomes favored active treatment. [7]
The primary result and secondary findings should remain in the same story. The short trial does not establish durable recovery or current authorization for alcohol treatment. [7]
Dietary evidence had two different clocks
A July ketogenic-diet study randomized 58 participants for one month. The later four-month extension was available to both groups, and 25 completed it. Its later psychiatric findings did not retain the original concurrent randomized comparison. [8]
A four-month outcome does not make the entire study a four-month controlled trial. The publication supports further investigation, not medication-replacement instructions or a universal psychiatric diet claim.
Personalization needed more than a technical label
A September trial randomized 51 patients to specified personalized accelerated stimulation approaches or sham; 43 completed treatment. It did not establish an active-versus-sham advantage on the primary one-week MADRS result. [9]
This is not a verdict against all TMS or imaging. It shows why an individualized procedure needs evidence that using it improves the intended outcome.
A separate European platform description identified approximately thirty sites across six countries, 59 candidate drugs considered, six shortlisted, and two starting oral-drug arms. Those are research-planning measures, not positive efficacy results. [10]
One framework keeps the new claims readable
| Claim | Necessary distinction |
|---|---|
| Positive trial | Primary versus secondary outcome |
| Durable benefit | Controlled phase versus uncontrolled extension |
| Personalized care | Individual information versus demonstrated clinical utility |
| New treatment program | Protocol, recruitment, and results as separate stages |
Research progress can be real even where the outcome is inconclusive. A negative or uncertain result is not a reason to replace measured evidence with a stronger marketing adjective.
What the quarter means for affluent and international families
Resources can widen access without answering which intervention fits. Ask how the evidence resembles the person’s diagnosis, treatment history, and co-occurring needs.
The selected developments connect directly to England, the USA, Switzerland, Abu Dhabi, and European trial networks. They do not create national recovery estimates for every country, including Saudi Arabia and the other GCC markets.
For care abroad, assess the continuation plan as carefully as the intensive phase. A handover should identify accepted responsibility, not merely a list of contacts. No original HNW survey or THE BALANCE outcome rate is reported here.
The bottom line
Q3 made the distinction between activity and benefit harder to ignore. New prescriptions and capacity matter, but their evaluation needs defined patients, relevant outcomes, and follow-up. Families should ask what a service can demonstrate, not simply how much it delivers.
For journalists
Reporting window: July 1 through September 30, 2026. Prepared retrospectively in October 2026, without backdating.
Key contrast: larger healthcare activity and new clinical approaches did not remove the need to examine denominators, primary endpoints, and later outcomes.
Suggested attribution: THE BALANCE retrospective Q3 evidence briefing.
Methodology and sources
This selective narrative briefing retains observation periods, revision status, and trial distinctions. It does not combine incompatible national measures into a ranking. Statistics are source findings; practical client implications are analysis.
- NHSBSA. Medicines Used in Mental Health, 2016/17 to 2025/26. July 23, 2026.
- SAMHSA. 2025 NSDUH annual report. July 2026.
- SAMHSA. 2025 NSDUH methodology.
- NHS England. Talking Therapies annual report, 2025/26. September 24, 2026.
- Department of Health, Abu Dhabi. Capacity and quality framework. August 3, 2026.
- ANQ. Psychiatric measurements 2025. September 17, 2026.
- Schacht and colleagues. Oral Semaglutide for Alcohol Use Disorder. July 29, 2026.
- Abram and colleagues. Ketogenic-diet study. July 8, 2026.
- Appelbaum and colleagues. Personalized stimulation trial. September 16, 2026.
- Gold and colleagues. European depression platform trial. September 15, 2026.


