England’s combined hypnotic and anxiolytic prescribing increased in 2025/26 after a longer decline. But the total hides different movements across medications. For professionals able to buy premium sleep products or private care, the relevant question is not simply whether treatment increased. It is whether people slept better and functioned better the next day.
Key findings
13,576,709: prescription items in the combined category in England in 2025/26.
1,867,857: identified patients in that category.
About 1%: annual growth in items, compared with 1.2% growth in identified patients.
15,771,599: items in 2016/17, showing that the recent rebound remains below the earlier total.
200,351: items in the newer other-hypnotics paragraph, which comprised daridorexant, in 2025/26.
Source: NHSBSA’s July 2026 release. These are covered NHS prescriptions subsequently dispensed in the community, not measurements of insomnia prevalence or recovery. [1]
A small rebound within a longer decline
A recent increase and a longer-term decline can both be true. The chart retains a common zero baseline rather than magnifying the small annual movement. Source: NHSBSA. [1]
The total hides a changing medication mix
The combined category is not a synonym for benzodiazepines. It contains medications with different purposes and clinical profiles. In the latest year, anxiolytic items continued to fall while hypnotic and other-hypnotic items increased. Calling the entire category sleeping pills obscures those differences.
| Category | 2023/24 | 2024/25 | 2025/26 |
|---|---|---|---|
| Anxiolytics | 5,607,023 | 5,382,329 | 5,176,191 |
| Hypnotics | 7,910,723 | 8,023,747 | 8,200,155 |
| Other hypnotics and anxiolytics | 1,554 | 40,358 | 200,351 |
| Barbiturates | 539 | 163 | 12 |
Source: NHSBSA’s BNF paragraph breakdown. The other-hypnotics paragraph in these years comprised daridorexant. Items are not doses or unique patients. [1]
The distinction changes the story. Growth in an aggregate can accompany a decline in one component and expansion in another. The figures alone cannot tell us whether a newer medication replaced existing treatment, added treatment, or improved outcomes.
A small starting value also matters. A large percentage increase from a low base can sound dramatic even when the absolute total remains a small part of the wider category. Both the counts and the period belong beside a growth headline.
Newer does not mean right for everyone
NICE’s daridorexant appraisal describes a defined pathway for adults with long-term insomnia and substantial daytime effects. It considers circumstances in which cognitive behavioral therapy for insomnia has not worked, is unavailable, or is unsuitable. Response and continuing need require review. [2]
That is a treatment decision, not a consumer upgrade. The prescribing increase does not show whether each patient met the relevant criteria or benefited. Nor does the appraisal make a recommendation for every person having difficulty sleeping.
The guidance also separates trial observations from assumptions used to model longer-term effects. A study can establish improvement on specific measures over a defined period without demonstrating every hoped-for long-term benefit. An economic projection is not the same as observing those outcomes in patients.
The newer category in absolute numbers
The growing category describes treatment activity, not effectiveness. Source: NHSBSA. [1]
Better sleep should include the following day
Sleep outcomes can include how long it takes to fall asleep, time awake overnight, total sleep time, perceived quality, and daytime functioning. These measures are related, but they answer different questions. Improvement in one should not be described as complete recovery.
A published secondary analysis of a randomized daridorexant trial compared adults aged 65 and older with younger adults over three months. It examined sleep measures and daytime symptoms. Its useful lesson is the specificity of the question: which patients, which comparison, which outcomes, and how long? [3]
For someone with significant professional responsibilities, functioning may be particularly important. But treatment should not become a promise of enhanced productivity. Working sustainably may be a more appropriate goal than working longer, and a wearable score is not interchangeable with a clinical assessment.
Why this matters for affluent and mobile families
A comfortable bedroom, privacy, and a stable routine may be attractive features of a care setting. They are not outcome measures. A program should explain what changed for the people receiving care and whether the improvement lasted after they returned to their usual environment.
That distinction matters when several interventions are combined. Improvement after a package does not identify which component helped, whether a simpler approach would have produced similar results, or what concurrent treatment contributed. The most marketable feature should not automatically receive the credit.
For an internationally mobile patient, continuity is another question to settle before treatment starts. Who reviews progress after travel? Who has the clinical history? Who should be contacted if symptoms or adverse effects change? These are planning questions, not evidence that a particular proportion of wealthy travelers experiences poor care.
The same questions can guide a discussion in the UK, United States, Germany, Switzerland, Netherlands, or GCC without assuming identical rules. Medication availability, prescribing arrangements, and travel requirements need jurisdiction-specific advice. An English prescribing table cannot resolve them.
What an accountable comparison would include
| Claim | Question | Evidence |
|---|---|---|
| Better sleep | Which aspect improved? | Measure, baseline, and follow-up |
| Restored functioning | Was daytime functioning assessed? | Appropriate instrument and personal goals |
| Lasting benefit | How long did follow-up last? | Timing, response rate, and concurrent care |
| Better private care | Compared with what? | Similar patients and consistent measures |
These questions make a comparison more useful than a claim that people slept well while staying somewhere restful. They also leave room for patient preferences. An outcome can be meaningful without implying that every symptom disappeared or every person needs the same treatment.
Separate a comfortable stay from lasting improvement
A useful sleep-care report should specify the difficulty that prompted treatment. Trouble falling asleep, repeated waking, and impaired daytime functioning are not interchangeable outcomes. A broad statement that sleep improved may conceal which measure changed and which did not. The description should make the patient group, starting point, and observation period easy to identify.
For a person whose usual schedule includes travel or high professional demands, the return home is part of the outcome question. A positive result in a structured setting does not establish how things will look after ordinary responsibilities resume. That does not make the initial result unimportant. It makes the timing of the assessment essential to interpreting it.
The follow-up report should record other care received in the meantime. Continuing medication, psychotherapy, or changes in daily circumstances may affect how a later result is understood. A program cannot confidently assign all subsequent improvement to one component merely because that component was prominent in its marketing. A package-level result and a component-level effect are different claims.
Families should also ask how unsuccessful treatment is handled. Is there a planned reassessment when the first approach does not help enough? Are adverse effects recorded alongside benefits? Who explains the next step? These questions make the proposed service more transparent without assuming that any intervention will work for everyone.
Outcome measurement should not turn into a competition over who returns to work fastest. The person’s own goals, health, and sustainable functioning matter. A demanding schedule can be a practical context for care, but it should not dictate that treatment must produce longer working hours. The purpose of reporting is to describe meaningful change, not to repackage productivity as a clinical endpoint.
For journalists, the strongest follow-up would connect the prescribing changes to evidence about assessment, review, and patient outcomes. The current administrative series cannot make that connection by itself. It provides a clear starting question: as the mix of treatment changes, which measures show whether the people receiving it are doing better?
What the numbers cannot tell us
The prescribing series does not measure insomnia severity, diagnoses, sleep quality, daytime functioning, individual wealth, or the complete private market. It cannot show whether executives became more likely to use medication or whether treatment worked better for them.
Lower prescribing would not automatically establish better care, just as higher prescribing does not establish inappropriate care. Explaining the trend requires information about clinical need, access, practice, and outcomes. Any individual medication change should be discussed with the prescriber rather than prompted by a population headline.
The bottom line
The rebound is real within the defined English series. The stronger story is the changing medication mix and the distance between counting treatment and measuring improvement. For people with many options, the useful advantage is careful assessment and reliable follow-up, not simply more interventions.
For journalists
Key comparison: 13.58 million combined-category items in 2025/26, still below 15.77 million in 2016/17.
Important caveat: this is prescribing activity, not an insomnia-prevalence estimate or a trial of private care.
Suggested attribution: THE BALANCE analysis of NHSBSA prescribing data and selected sleep-treatment evidence.
Methodology and sources
The analysis retains the July 23, 2026 NHSBSA release and its medication categories. Counts are exact in the table and rounded in chart labels. Clinical guidance and trial evidence are discussed separately. No causal effect, original patient result, or private-market estimate is calculated.
