4 Nov 2026 · Evidence review · in progress

Private medical insurance: what the evidence supports

The note this review supportsPrivate medical insurance is sold on absence. Nobody has tested that.

How this was checked

The question was split into tiers: who has group private medical insurance and how many use it, what it does to waiting, whether sooner treatment changes outcomes, sickness absence and return to work, staying and joining, cost and return, who is left out, and what the government’s Keep Britain Working review says about it.

Each tier was searched in reviews, trials, register studies, official statistics and government surveys, then graded with GRADE logic. Randomised trials start high and lose grades for risk of bias, inconsistency, indirectness and imprecision. Observational studies start low. Material from insurers, brokers, trade bodies and consultancies was recorded and labelled, and never allowed to set a grade.

The search and first grading were done on 3 November 2026 by AI agents over 215 sources. The tiers on sickness absence and presenteeism were still running when this review was written, so those sections were built here from the sources already opened. The seeded 30 per cent re-check that the brief asked for had not run when this was written. Every passage in the ledger was reopened and copied from its source on 3 November 2026 and checked by machine against the source text. Where only an abstract could be read, the ledger says so. I sell this benefit for a living. Every grade here is provisional.

The headline

What is said	What the independent evidence says
Private medical insurance adds £2.5 billion a year to the economy through less absence and more productivity C37	That is a model commissioned by an insurer, with no comparison group, and the paper calls the benefits under-researched C37 C38. No independent study anywhere has set group PMI against absence, output or the premium.
Health and protection insurance delivered £6.1 billion of benefits in 2021 C35	A trade body model that borrows effect sizes from studies of other programmes and halves them C36.
PMI reduces sickness absence and gets people back faster C40	The one study of employer-paid insurance and absence, in Denmark, found no significant effect in its main analysis and a mixed result in its matching analysis C27 C28. The one UK trial of faster access measured time off work and found nothing to report C20.
PMI gets people seen and treated sooner	Probably, but nobody has measured it for insured employees in the UK. NHS-funded patients at private hospitals waited about half as long C15 C16, and insured people waited less in Ireland C17.
Keep Britain Working backs PMI	The review mentions it once, as a coverage statistic C44, lists no return for it C45, and names group income protection when it talks about pooled insurance for small firms C47.

Who has it

The last representative government figures are from 2009 and 2010. In the 2010 employer survey, 20 per cent of employers in Great Britain had provided subsidised private medical insurance to any employees, 18 per cent of small, 53 per cent of medium and 61 per cent of large employers, and 21 per cent in the private sector against 5 per cent in the public sector C1. In the 2009 employee survey, 23 per cent of employees said their organisation provided it, from 11 per cent in workplaces of 50 or fewer to 31 per cent at 250 to 499, and 30 per cent in the private sector against 10 per cent in the public C2. Twelve per cent of all employees were members of the scheme, and 58 per cent of those offered it and eligible C3.

Nothing representative has been collected since. The later government employee surveys have no item on it C2, HMRC counted 2.38 million people with employer-paid medical or dental benefit in 2016 to 2017 and said payrolled benefit was a large uncounted gap C6, and then stopped counting anything but company cars C7.

What exists now is the industry’s own figures and a survey of HR professionals. The insurers’ trade body reports 4.8 million people on workplace policies in 2024 C8. A market analyst puts two thirds of private medical cover as employer-funded and 8.06 million people able to use cover once dependants are counted C10. The Office for National Statistics puts all voluntary health insurance at 2.8 per cent of UK health spending, £9.5 billion in 2025 C13. The CIPD’s 2025 survey has 25 per cent of responding organisations offering PMI to all employees and another 25 per cent by seniority, with 40 per cent of large private sector organisations and 20 per cent of private sector SMEs offering it to everyone C4; organisations of 1,000 or more were 39 per cent of its weighted sample C5.

Grade: low that group PMI reaches a minority of UK workers, concentrated in larger private sector employers, from representative surveys that are 16 and 17 years old. No current independent figure exists for the share of employers offering it or the share of employees covered.

How many use it

No independent UK source gives the share of covered employees who claim in a year. The nearest is the private healthcare information body’s count of insured hospital admissions: 624,800 in 2023, 664,815 in 2024 and 670,300 in 2025, with insurance paying for 70 per cent of private admissions and chemotherapy the largest insured procedure group C11. Insured admissions fell 2.6 per cent in the first quarter of 2026, the first first-quarter fall in five years C12. The trade body reports 1.8 million people claiming on individual or workplace policies in 2024 and £2.6 billion paid on workplace schemes C9. Dividing its claimants by its covered total gives about 28 per cent, a calculation made here from figures that mix individual and workplace cover.

Grade: low, direction unknown. Use of insured private care rose from 2021 to 2025 on both independent and insurer counts and may have levelled off. The claim rate, the average claim and what people claim for rest on insurer figures only.

The queue

The NHS side is measured monthly. At the end of August 2026, 7.4 million pathways in England were waiting to start treatment, the median wait was 12.2 weeks, and 65.2 per cent had waited up to 18 weeks against a standard of 92 per cent C14.

The insured side is not measured at all. No UK study was found that compares how long people with private medical insurance wait for a consultant, a scan or an operation with NHS patients who have the same condition, and no insurer publishes waiting data. The nearest UK evidence is about NHS-funded patients treated in private hospitals, who are not insured. For hip replacement in 2018 to 2019 they waited 59.1 days on average against 120.5 at NHS providers, a gap the authors attribute mostly to case-mix and shorter private waits C15. Across hip and knee replacements from 1997 to 2019, private-provider waits were half the NHS’s, and the poorest fifth waited longer than the richest C16. In Ireland, more people without insurance than with it waited over three months for hospital services C17. In Australia, more insurance cover in an area barely moved public waits C18.

PhysioDirect is the caution. Telephone assessment cut the median wait for physiotherapy from 34 days to 7, but the wait for a face-to-face appointment did not change C19. Faster first contact is not the same as faster treatment.

Grade: low, improves. The direction is consistent across three countries and the gaps are large, which offsets the indirectness. The size of the advantage for a covered UK employee is unknown, and it moves with NHS waits, which differ by specialty and by month.

Does sooner mean better

For short waits, the trials say little changes. PhysioDirect found equivalent physical health at six months and nothing to report on time off work C20. Immediate physiotherapy for arm pain gave full recovery in 35.8 per cent at 26 weeks against 38.6 per cent after the usual six to eight week wait, with an interval that does not rule out a modest benefit C21. Early physiotherapy for new low back pain in the United States improved disability at three months by less than the smallest change that matters and by nothing detectable at a year C22.

For long waits and serious conditions, the observational studies point the other way. Each four-week delay to cancer surgery was associated with a 6 to 8 per cent higher risk of death across 34 studies C25. Each extra week waiting for hip or knee replacement cost 0.1 per cent of the outcome range, small for one person and large across a year’s patients C26.

Private treatment against NHS treatment is confounded by who gets it. NHS-funded patients in independent treatment centres were healthier, less severely affected and more affluent, and after adjustment the NHS scored slightly worse C23. When a larger study accounted for unobserved differences between patients, most outcome differences disappeared C24. No study compares clinical outcomes for insured patients with NHS patients.

Grade: low, mixed. Whether sooner is better depends on the condition and on how long the alternative wait is. None of these studies tests private medical insurance.

Sickness absence and return to work

One study of employer-paid health insurance and absence was found, in Denmark. Its regression and mixture-model analyses found no significant absence-reducing effect of the insurance C27. Its matching analysis found no effect in one dataset and some evidence of one in the other, where the matching was less complete, and the paper states that nationally and internationally there was no evidence on the question C28. It is observational and cross-sectional, in a country where the cover was then tax free. No UK study exists.

The waiting literature does not settle it either. In Norwegian register data, a quarter of employees off sick were also waiting for specialist care, and the length of the wait made almost no difference to the length of the absence C29. A Norwegian quasi-experiment on orthopaedic queues found the opposite for one group: long waits increased medium to long-term absence and permanent disability benefit among people already off sick at referral C30. A randomised trial of assessment within four weeks against ten to fourteen is recruiting in Norway now C31. In the UK, the PhysioDirect trialists said their results challenge the assumption that faster access reduces time off work C20.

Grade: very low, no clear effect, for absence and for return to work. One observational study abroad with a mixed result, no UK study, and a waiting literature that cuts both ways. That is close to no evidence either way, and it is not evidence that the benefit does nothing. The one plausible mechanism, shortening a long wait for someone already off sick C30, has never been tested with insurance.

Staying and joining

No study measures whether UK employees with group PMI leave less often, apply more or accept more offers. The turnover literature is American, where employer cover is most people’s only health insurance. The job-lock estimate that employer insurance cuts voluntary turnover from 16 to 12 per cent a year comes from married men in 1987 and rests on the cost of losing cover for a pre-existing condition C33, a mechanism that does not exist alongside the NHS. The only UK data are intentions: in the 2009 government survey the factors linked to fewer people thinking about leaving were culture, work-life balance, flexible working, engagement and age, and private medical insurance was not among them C34. Insurer surveys in which employees rank PMI highly are stated preferences, labelled and not graded here.

Grade: very low, direction unknown, for turnover. None found for recruitment.

Cost and return

No independent study anywhere sets the cost of group PMI against absence, output or turnover with a comparison group. That is no evidence of effect, not evidence of no effect.

The figures in circulation are models. The trade body’s £6.1 billion for health and protection insurance in 2021 C35 takes return-to-work effects of 19 to 54 per cent from studies of other workplace programmes and halves them C36. An insurer’s £2.5 billion a year for PMI C37 and 22 days and £14,700 per small business C38 come from a paper that says the benefits are under-researched and publishes no inputs. A second insurer’s £750 million is the trade body model adapted C39. A provider’s £2.87 per £1, which it calls proven, is a calculation from assumed absence savings C40. None has a comparison group and none has been checked independently.

Against the premium sit three taxes. The cover is a taxable benefit for the employee with reporting obligations for the employer C41, the employer pays Class 1A National Insurance at 15 per cent on it C42, and the premium carries insurance premium tax at 12 per cent C43.

Grade: none found for return on investment.

Who gets it and what is left out

Cover follows pay, sector and size. In the 2009 survey it was 30 per cent in the private sector against 10 per cent in the public, and 11 per cent in small workplaces against 31 per cent in large C2. Income, age and sex were key determinants of holding employer-paid cover in English panel data C53. Half of the organisations offering it in the CIPD’s 2025 survey offered it by seniority C4. Policies do not normally cover pre-existing conditions or the ongoing management of chronic conditions such as asthma and diabetes, though larger schemes are more likely to cover pre-existing conditions C52. In Denmark, employer-paid cover was associated with 10 per cent less use of public hospitals C32; whether PMI-funded care changes NHS waits for people without it has not been measured in the UK.

Grade: very low, mixed. The pay and seniority skew is well described but old; the system effects are unmeasured.

What Keep Britain Working says

The final report mentions private medical insurance once, to repeat the CIPD’s coverage figures without a footnote C44. Its list of return-on-investment evidence has employee assistance programmes, mental health and musculoskeletal interventions and employer cases, and nothing for PMI C45. Among possible incentives it lists tax relief on benefits in kind for certified employers, tied to evidence of impact, without naming medical insurance or insurance premium tax C46. The June 2026 update commits to understanding risk pooling for small businesses C49. The September 2026 update suggests pooled insurance schemes for small employers that build on approaches such as group income protection, and does not name PMI C47. The same update says that today a provider sells a service based on claims of efficacy C48.

The trade press reads it the same way. In an analysis of group PMI’s role in Keeping Britain Working published on 2 November 2026, one broker said the sector would need to prove that the benefit reduces absence C50, and a consultancy contributor said that where reliable data are unavailable, business cases are harder to make C51. Both are consistent with this review.

Claims that do not hold up

“PMI provides a £2.5 billion boost to the UK’s economic output every year” C37 is an insurer-commissioned model whose own authors call the area under-researched. “22 fewer sickness absence days” per small business C38 is what that model assumes, not what anyone measured. “£6.1bn” C35 borrows its effects from other programmes and halves them C36. “£750m” C39 is the same model adapted. “We’ve proven that it pays off” C40 is arithmetic on an assumed absence saving. Citing Keep Britain Working as support for PMI misreads a report that mentions it once as a statistic C44 and names a different product when it discusses pooled insurance C47.

In the other direction, “PMI doesn’t work” is not supported either. Most tiers have no evidence of effect, which is not evidence of no effect.

What this changes in the grid of 13 October

The published grid carried private medical insurance at very low, no clear effect, for absence and for long absence and return; very low, mixed, for turnover; low, mixed, for take-up; and none found elsewhere. The absence cells stand C27 C28. Turnover moves from mixed to direction unknown: the mixed reading described the US job-lock literature, whose mechanism does not apply in the UK C33, and no UK study of leaving exists C34. Take-up moves from mixed to direction unknown: the UK figures are descriptive C3 C11 and the mixed reading rested on a Danish question about total care use C32. The grid’s note that PhysioDirect cut the wait from 34 days to 7 now carries the qualification that the wait for a face-to-face appointment did not change C19.

Limits

The two tiers that matter most to how this benefit is sold, absence and presenteeism, were still being searched by the agents when this was written, so their sections rest on the sources opened for the other tiers. Several key studies were read at abstract level, including the two Norwegian waiting studies C29 C30, so their funders could not be checked. The insurers’ trade body and the market analyst publish press releases, not data, and their figures could not be checked. No insurer publishes waiting times for its members, so the tier on access rests on NHS-funded patients at private hospitals and on Ireland and Australia. Three think tanks blocked automated access and nothing from them was read.

What would change these grades

Insurer claims data linked to NHS records, comparing referral-to-treatment times for insured and uninsured employees with the same condition in the same region. That would turn the access grade from low to something.

A staggered introduction of group PMI across UK employers, or a change in who is eligible, with payroll absence records before and after. Or a randomised trial of fast-track diagnosis and treatment for employees already off sick, the group the Norwegian queue study says is harmed by waiting C30. The Norwegian trial now recruiting is the nearest thing C31.

An independent economic evaluation with a comparison group that sets absence and turnover savings against the premium, the benefit in kind tax, the employer’s National Insurance and insurance premium tax. Publishing the inputs behind the £2.5 billion and £6.1 billion models for independent review.

A government survey question, asked of all employees, on whether they have private medical insurance through work, whether they used it in the last year, and for what.

Claims ledger

Each claim is worded exactly as it was checked. Assessed claims show the source, passage, limits and date checked. If a source could not be checked, the entry records what was attempted and why. Claims still being checked are marked in review.

C1Supported
In the Department for Work and Pensions employer survey with fieldwork in 2010, 20 per cent of employers in Great Britain had provided subsidised private medical insurance to any employees in the past 12 months, 18 per cent of small, 53 per cent of medium and 61 per cent of large employers, and 21 per cent in the private sector against 5 per cent in the public sector.
Source
Young V, Bhaumik C (2011). Health and well-being at work: a survey of employers. Department for Work and Pensions Research Report No 750. Open source (full text read)
Passage
The notable exception to this rule was ‘Subsidised private medical insurance’: this was most likely to be reported as an initiative among respondents in the private sector than those in the public sector (21 per cent versus five per cent respectively).
Where
Section 3.1; the size figures are the row "Subsidised private medical insurance" in Table 3.1, which reads 20 total, 18 small, 53 medium, 61 large, 5 public, 21 private; Table 3.2 gives 20 weighted by employer and 48 weighted by employment
Limits
Telephone survey of 2,250 employers with two or more staff, fieldwork February to May 2010, weighted to employers so dominated by small firms. The question asked whether the benefit was provided to any employees, not how many were covered. Sixteen years old, and no later government employer survey has asked the question with a private medical insurance option.
Checked
3 Nov 2026
Back to text↩ 1
C2Supported
In the Department for Work and Pensions survey of employees with fieldwork in 2009, 23 per cent said their organisation provided private medical insurance, 11 per cent in organisations of 1 to 50 staff and 31 per cent at 250 to 499, and 30 per cent in the private sector against 10 per cent in the public sector.
Source
Young V, Bhaumik C (2011). Health and well-being at work: a survey of employees. Department for Work and Pensions Research Report No 751. Open source (full text read)
Passage
Many benefits and initiatives were more likely to be provided to public sector workers than to those working in the private sector (the exceptions were private medical insurance, subsidised canteen/restaurant, schemes to undertake voluntary work and weight loss advice/programmes).
Where
Section 2.1; the figures are the row "Private medical insurance" in Table 2.1, which reads 23 total, 11 small, 25 medium, 31 large, 28 very large, 30 private, 10 public
Limits
Face-to-face survey of 2,019 employees, fieldwork October to December 2009. Employees reported what they believed their organisation provided. Seventeen years old; the 2014 and 2024 to 2025 employee surveys have no private medical insurance item.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2 ↩ 3
C3Supported
In the same 2009 survey, 12 per cent of all employees had been members of their organisation's private medical insurance scheme in the past 12 months, and 58 per cent of those who were offered it and eligible.
Source
Young V, Bhaumik C (2011). Health and well-being at work: a survey of employees. Department for Work and Pensions Research Report No 751. Open source (full text read)
Passage
In the case of private medical insurance, across the whole sample take up was relatively low at 12 per cent. However, among those who were provided with the benefit and were eligible for it, take up stood at 58 per cent (i.e. on a par with other, more widely available benefits).
Where
Section 2.1, text above Table 2.3; the questionnaire item D5b asks "And have you personally been a member of the private medical insurance scheme in the last 12 months?"
Limits
Take up here means membership of the scheme, not making a claim. Self-reported. No later government survey repeats the question.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2
C4Supported
In the CIPD's 2025 survey, 25 per cent of the 1,101 responding organisations offered private medical insurance to all employees and a further 25 per cent offered it depending on seniority; for all employees it was 40 per cent of large private sector organisations, 20 per cent of private sector SMEs, 7 per cent of public sector and 11 per cent of non-profit organisations.
Source
Sinclair A, Suff R (2025). Health and wellbeing at work 2025. Survey report. Chartered Institute of Personnel and Development, supported by Simplyhealth. Open source (full text read)
Passage
Private medical insurance For all employees 25 7 11 32 20 40 Depends on seniority 25 10 13 29 18 37
Where
Appendix, Table 2, Wellbeing benefits on offer, by sector; columns all respondents, public sector, non-profit sector, private sector, private sector SMEs, private sector large
Limits
Respondents are HR professionals answering for their organisation, with large organisations over-represented. The survey is supported by a health cash plan provider. It counts organisations offering the benefit, not employees covered.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2
C5Supported
In the CIPD's 2025 sample, organisations with 1,000 or more employees were 32 per cent of unweighted respondents and 39 per cent after weighting.
Source
Sinclair A, Suff R (2025). Health and wellbeing at work 2025. Survey report. Chartered Institute of Personnel and Development, supported by Simplyhealth. Open source (full text read)
Passage
1,000+ 351 32 429 39
Where
Appendix, Table 1, Profile of respondents, size of organisation; columns unweighted number, unweighted per cent, weighted number, weighted per cent
Limits
The weighting is to the CIPD's own panel, not to all employers in Great Britain, where organisations of this size are a tiny fraction of the total.
Checked
3 Nov 2026
Back to text↩ 1
C6Supported
In HMRC's last count, for 2016 to 2017, 2.38 million people received private medical or dental insurance as a taxable benefit, about two thirds of everyone with a taxable benefit, and HMRC says payrolled medical benefit is a large uncounted gap.
Source
HM Revenue and Customs (2019). Taxable Benefits in Kind: National Statistics commentary 2019, data for 2016 to 2017. Open source (full text read)
Passage
frequently payrolled is private medical benefit/treatment and it is believed that
Where
Annex C, completeness; the count is in the commentary on Table 4.3, "2.38 million individuals received private medical/dental benefits (with a total taxable value of £1.89 billion)", and the summary gives "Around two thirds (65%) of this number received private medical/dental insurance benefits."
Limits
P11D returns only, so people whose benefit was payrolled are missing. Dental cover is included. Nine years old and the series has since been closed.
Checked
3 Nov 2026
Back to text↩ 1
C7Supported
Since 2020 HMRC's benefit in kind statistics cover company cars only, so there is no official count of people with employer-paid medical insurance after 2016 to 2017.
Source
HM Revenue and Customs (2026). Benefit in kind statistics commentary, July 2026. Open source (full text read)
Passage
In 2020 the scope of HMRC’s taxable benefits in kind statistics was restricted to company cars (and company car fuel) only.
Where
Section 1, about this release
Limits
Says what the statistics cover. Whether HMRC holds the data internally is not stated.
Checked
3 Nov 2026
Back to text↩ 1
C8Supported
The insurers' trade body reports 4.8 million people covered by workplace health insurance policies in 2024, out of 6.5 million covered in all.
Source
Association of British Insurers (2026). Insurers process record £4bn across individual and workplace health schemes. Press release, 21 January 2026. Open source (full text read)
Passage
Figures show 4.8 million of these people were covered by workplace policies
Where
Second paragraph, where the sentence continues that this was a second consecutive record; the preceding sentence gives 6.5 million covered in total in 2024
Limits
An aggregate of member insurers' returns, published by the insurers' trade body with no method note. Workplace policies include dependants and are not the same as employees covered. Not independently checkable.
Checked
3 Nov 2026
Back to text↩ 1
C9Supported
The insurers' trade body reports 1.8 million people claiming on individual or workplace health insurance in 2024, up 10 per cent on 2023, and £2.6 billion of claims paid on workplace schemes, up 16 per cent.
Source
Association of British Insurers (2026). Insurers process record £4bn across individual and workplace health schemes. Press release, 21 January 2026. Open source (full text read)
Passage
Insurers processed a record £2.6 billion in claims for workplace schemes alone in 2024
Where
Fourth paragraph, where the sentence continues that this was a 16 per cent increase on 2023; the third paragraph reads "The total number of people who claimed on either individual or workplace health insurance policies reached 1.8 million in 2024, up 10% from 2023."
Limits
Trade body figures with no method note. The claimant count is not split between workplace and individual policies, so a workplace claim rate cannot be read from it. Dividing 1.8 million by 6.5 million gives about 28 per cent, a calculation made here, not a published figure.
Checked
3 Nov 2026
Back to text↩ 1
C10Supported
A commercial market analyst reports that two thirds of private medical cover is funded by employers as an employee benefit, and that 8.06 million people could use private medical cover in 2023 once dependants are counted.
Source
LaingBuisson (2025). The UK health cover market surges by £825 million in a year. Press release for Health Cover UK Market Report, 20th edition, 30 January 2025. Open source (full text read)
Passage
Two-thirds of private medical cover is funded by employers as an employee benefit.
Where
Private medical cover section; the same section reads "once dependents are included, it means that 8.06 million people were able to make use of private medical cover"
Limits
A press release for a paid report built from provider data. The report itself was not read. The count includes self-insured company schemes and dependants, so it is not comparable with the trade body's 4.8 million.
Checked
3 Nov 2026
Back to text↩ 1
C11Supported
Insured private in-patient and day-case admissions in the UK were 624,800 in 2023, 664,815 in 2024 and 670,300 in 2025, and insurance paid for 70 per cent of private admissions.
Source
Private Healthcare Information Network (2026). Private healthcare market update: June 2026, United Kingdom. Open source (full text read)
Passage
The proportion of admissions by payment method remained at 70:30 insured to self-pay.
Where
Payment method paragraph; the table "Insured admissions by nation or English region" has the row "Total 624,800 664,815 670,300" for 2023, 2024 and 2025, and the insurance-funded top ten procedures table has chemotherapy first at 75,350 admissions in 2025
Limits
Admissions, not people; chemotherapy patients usually need several. Employer-paid and individually bought cover are not separated. Outpatient consultations, scans and therapies are not counted.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2
C12Supported
Insured private admissions fell 2.6 per cent in the first quarter of 2026 against the first quarter of 2025, the first fall in first-quarter insured admissions in five years.
Source
Private Healthcare Information Network (2026). Private healthcare market update: September 2026, United Kingdom. Open source (full text read)
Passage
This was the first time there has been a decline in Q1 insured admissions in the past five years.
Where
Insured admissions section; the preceding sentence reads "There was a slight decline (2.6%) in insured admissions compared to Q1 2025."
Limits
One quarter. Admissions, not covered people.
Checked
3 Nov 2026
Back to text↩ 1
C13Supported
Voluntary health insurance was 2.8 per cent of UK spending on healthcare in 2025, or £9.5 billion, on provisional figures.
Source
Office for National Statistics (2026). Healthcare expenditure, UK Health Accounts: 2024 and 2025. Statistical bulletin, 29 April 2026. Open source (full text read)
Passage
Voluntary health insurance, the next largest means of financing, was 2.8% of overall spending on healthcare in 2025, or £9.5 billion.
Where
Main points
Limits
Includes dental and cash plans as well as medical insurance, and individual as well as employer-paid cover. 2025 figures are provisional.
Checked
3 Nov 2026
Back to text↩ 1
C14Supported
At the end of August 2026, 7.4 million NHS referral-to-treatment pathways in England were waiting to start treatment, the median wait was 12.2 weeks, and 65.2 per cent had waited up to 18 weeks against a 92 per cent standard.
Source
NHS England (2026). Statistical press notice: NHS referral to treatment (RTT) waiting times data, August 2026. Published 8 October 2026. Open source (full text read)
Passage
In 65.2% of cases the patient had been waiting up to 18 weeks, thus not meeting the 92% standard.
Where
Main findings, page 1; the same page gives 7.4 million pathways and "the median waiting time was 12.2 weeks"
Limits
NHS waits only. Pathways, not people; some people are on more than one. Four trusts did not submit data for the month.
Checked
3 Nov 2026
Back to text↩ 1
C15Supported
For NHS-funded hip replacements in England in 2018 to 2019, the average inpatient wait was 59.1 days at private providers and 120.5 days at NHS providers, and the authors attribute the difference mostly to case-mix and shorter private waits.
Source
Kasteridis P, Siciliani L, Sivey P (2026). Waiting time prioritisation for public and private providers: evidence from hip replacement surgery. Health Economics. Open source (full text read)
Passage
In 2018/19 (2019/20), inpatient waiting time was on average 59.1 (66.7) days for private providers and 120.5 (120.8) days for public providers
Where
Data section, descriptive waiting times; the abstract reads "Differences in inpatient waiting‐time prioritisation between public and private providers are mostly due to different casemix and shorter average waiting times by private providers."
Limits
These are NHS-funded patients treated in private hospitals, not people with private medical insurance. Private providers treat a healthier, less complex case-mix. One procedure.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2
C16Supported
Across NHS-funded hip and knee replacements in England from 1997 to 2019, waiting times at private providers were half those at the NHS, and the poorest fifth of patients waited longer than the richest fifth.
Source
Kirkwood G, Pollock AM (2025). Outsourcing National Health Service surgery to the private sector: waiting time inequality and the making of a two-tier system for hip and knee replacement in England. International Journal of Social Determinants of Health and Health Services. Open source (full text read)
Passage
Waiting times for private providers were half those for the NHS, and the poorest 20 percent waited longer than the richest 20 percent.
Where
Abstract
Limits
NHS-funded patients, not insured ones. Descriptive; the authors attribute the gap partly to private providers selecting simpler cases.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2
C17Supported
In Irish household survey data from 2007 and 2010, a higher share of people without private health insurance than with it were waiting more than three months for hospital services.
Source
Whyte R, Connolly S, Wren MA (2020). Insurance status and waiting times for hospital-based services in Ireland. Health Policy. Open source (abstract only)
Passage
A higher proportion of those without PHI were waiting more than three months for hospital services relative to those with PHI.
Where
Abstract
Limits
Ireland, where private insurance buys access inside public hospitals in a way it does not in the UK. Survey data, no adjustment for condition reported in the abstract. Read at abstract level.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2
C18Supported
In Australian linked data from 2014 to 2018, each percentage point rise in private insurance cover in an area was associated with about 0.34 days less waiting in public hospitals, which the authors call of limited practical significance.
Source
Yang O, Yong J, Zhang Y (2024). Effects of private health insurance on waiting time in public hospitals. Health Economics. Open source (abstract only)
Passage
We find that one percentage point increase in PHI coverage leads to about 0.34 days (or 0.5%) reduction in waiting times in public hospitals on average.
Where
Abstract
Limits
Australia, one state, instrumental variable design. Measures the effect of insurance cover on public waits for everyone, not the wait of an insured person. Read at abstract level.
Checked
3 Nov 2026
Back to text↩ 1
C19Supported
In the PhysioDirect trial of 2,249 adults, telephone assessment cut the median wait for physiotherapy from 34 days to 7, but the wait to a first face-to-face appointment was no different.
Source
Salisbury C, Foster NE, Hopper C, et al. (2013). A pragmatic randomised controlled trial of the effectiveness and cost-effectiveness of 'PhysioDirect' telephone assessment and advice services for physiotherapy. Health Technology Assessment, 17(2). Open source (full text read)
Passage
There was no evidence that the waiting time to first face-to-face appointment was different between the PhysioDirect and the usual-care arms (Table 13).
Where
Chapter 3, waiting times; Table 13 gives a median of 35 days for usual care and 30 for PhysioDirect, ratio 0.95, 95 per cent confidence interval 0.88 to 1.03. The abstract gives the wait for treatment as "median 7 days vs 34 days"
Limits
The faster route was a telephone call, not a private appointment. Four NHS community services in England, 2009 to 2010. Usual-care waits were short by current standards.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2
C20Supported
In the same trial, physical health at six months was equivalent between the arms, and the trialists say the result challenges the assumption that faster access to care reduces time off work.
Source
Salisbury C, Montgomery AA, Hollinghurst S, et al. (2013). Effectiveness of PhysioDirect telephone assessment and advice services for patients with musculoskeletal problems: pragmatic randomised controlled trial. BMJ, 346, f43. Open source (full text read)
Passage
for example, that faster access to care will reduce patients’ time off work or improve satisfaction with the service.
Where
Discussion, implications; the sentence begins "The findings show that randomised trials of service reorganisation are feasible and are also informative in challenging assumptions". The HTA report's abstract gives the six-month primary outcome as a mean physical component score of 43.50 against 44.18, adjusted difference in means 0.01 lower, 95 per cent confidence interval 0.80 lower to 0.79 higher, and its Table 49 gives mean days off work because of the condition of 2.77 for usual care and 2.30 for PhysioDirect
Limits
Time off work was a secondary, self-reported measure with incomplete data; the days-off figures are means with wide spreads and no test reported. Outcomes at six weeks slightly favoured PhysioDirect. Not a trial of private treatment.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2 ↩ 3
C21Supported
In a UK trial of 538 adults referred for arm pain, full recovery at 26 weeks was 35.8 per cent with immediate physiotherapy and 38.6 per cent after the usual wait of about six to eight weeks, a difference of 2.8 points in favour of waiting with a confidence interval from 11.3 lower to 6.5 higher.
Source
Jones GT, Macfarlane GJ, Walker-Bone K, et al. (2019). Maintained physical activity and physiotherapy in the management of distal arm pain: a randomised controlled trial. RMD Open, 5(1), e000810. Open source (full text read)
Passage
In contrast, there was no significant difference in the probability of full recovery at 26 weeks between participants randomised to immediate (35.8%) as compared with routinely timed physiotherapy (38.6%), a difference of −2.8% (95% CI −11.3% to 6.5%).
Where
Results, primary outcome; Methods states "At the start of the trial, this wait was approximately 6-8 weeks."
Limits
The interval does not rule out a modest benefit or harm from waiting. The wait was a few weeks, not months. One condition, NHS community physiotherapy.
Checked
3 Nov 2026
Back to text↩ 1
C22Supported
In a US trial of 220 adults with recent low back pain, early physiotherapy improved disability at three months by less than the minimum clinically important difference and by nothing detectable at one year.
Source
Fritz JM, Magel JS, McFadden M, et al. (2015). Early physical therapy vs usual care in patients with recent-onset low back pain: a randomized clinical trial. JAMA, 314(14), 1459-1467. Open source (abstract only)
Passage
Among adults with recent-onset LBP, early physical therapy resulted in statistically significant improvement in disability, but the improvement was modest and did not achieve the minimum clinically important difference compared with usual care.
Where
Abstract, conclusions; the results give a one-year between-group difference of 2.0 points with an interval from 5.0 lower to 1.0 higher
Limits
United States, four sessions against none in the first four weeks. Read at abstract level.
Checked
3 Nov 2026
Back to text↩ 1
C23Supported
In an English audit of NHS-funded surgery, patients treated in independent sector treatment centres were healthier, less severely affected and more affluent than those treated by NHS providers, and after adjustment NHS providers had slightly worse Oxford hip and knee scores.
Source
Chard J, Kuczawski M, Black N, van der Meulen J (2011). Outcomes of elective surgery undertaken in independent sector treatment centres and NHS providers in England: audit of patient outcomes in surgery. BMJ, 343, d6404. Open source (full text read)
Passage
Patients in ISTCs were healthier than those in NHS providers, had less severe preoperative symptoms, and were more affluent, though the differences were small.
Where
Abstract, results; the adjusted differences are "−1.7 (95% confidence interval −2.5 to −0.9) on the Oxford hip score and −0.9 (−1.6 to −0.2) on the Oxford knee score"
Limits
NHS-funded patients, not insured ones. The authors warn that adjustment for case-mix may be incomplete. Commissioned by the Department of Health's commercial directorate.
Checked
3 Nov 2026
Back to text↩ 1
C24Supported
In 169,232 NHS-funded hip and knee replacements in private hospitals and 262,659 in NHS hospitals, most outcome differences disappeared when the analysis accounted for unobserved differences between patients, apart from fewer hospital infections and more prolonged stays in private hospitals.
Source
Anderson M, Friebel R, Maynou L, Kyriopoulos I, McGuire A, Mossialos E (2024). Patient outcomes, efficiency, and adverse events for elective hip and knee replacement in private and NHS hospitals: a population-based cohort study in England. The Lancet Regional Health Europe, 41, 100904. Open source (full text read)
Passage
IV analyses produced no significant differences between private and NHS hospitals, except for lower probability in private hospitals of hospital-associated infection (−0.0057, 95% CI −0.0081, −0.0032), and greater probability in private hospitals of prolonged post-operative LOS (0.2653, 95% CI 0.1833, 0.3472).
Where
Abstract, findings
Limits
NHS-funded patients, 2016 to 2019. The lead author declares consultancy fees from the private healthcare information body, unrelated to the study. Says nothing about insured patients.
Checked
3 Nov 2026
Back to text↩ 1
C25Supported
Across 34 observational studies, each four-week delay to cancer surgery was associated with a 6 to 8 per cent higher risk of death.
Source
Hanna TP, King WD, Thibodeau S, et al. (2020). Mortality due to cancer treatment delay: systematic review and meta-analysis. BMJ, 371, m4087. Open source (full text read)
Passage
Surgery findings were consistent, with a mortality risk for each four week delay of 1.06-1.08 (eg, colectomy 1.06, 95% confidence interval 1.01 to 1.12; breast surgery 1.08, 1.03 to 1.13).
Where
Abstract, results
Limits
All retrospective observational studies, mostly North American and European. Says that delay is harmful, not that private medical insurance shortens cancer waits in the UK.
Checked
3 Nov 2026
Back to text↩ 1
C26Supported
In English patient-reported outcome data, each extra week of waiting reduced the health gain from hip and knee replacement by 0.1 per cent of the outcome range, and had no effect on the gain from varicose vein or hernia surgery.
Source
Nikolova S, Harrison M, Sutton M (2016). The impact of waiting time on health gains from surgery: evidence from a national patient-reported outcome dataset. Health Economics, 25(8), 955-968. Open source (abstract only)
Passage
The magnitude of this effect at patient level is small, 0.1% of the outcome measure range for each additional week of waiting.
Where
Abstract
Limits
NHS patients. Small per person; the authors value it as large across a whole year's patients. Read at abstract level.
Checked
3 Nov 2026
Back to text↩ 1
C27Supported
The one study found of employer-paid health insurance and sickness absence, two Danish surveys from 2009 and 2010 analysed with quantile regression and a mixture model, found no significant absence-reducing effect of the insurance.
Source
Pedersen KM (2011). Sickness absence and voluntary employer paid health insurance. Health Economics Papers 2011:1. COHERE, University of Southern Denmark. Open source (full text read)
Passage
No significant results are found on the absence reducing property of health insurance. A two component („short‟ and „long‟ term absence) finite mixture model is also applied with the same result.
Where
Abstract
Limits
Observational, cross-sectional, self-reported absence, Denmark, where the cover was tax free for employees at the time. Not a controlled test. Absence of a significant result in one study is not evidence that there is no effect.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2 ↩ 3
C28Supported
The same paper states that there was no evidence nationally or internationally on whether employer-paid insurance reduces absence, and that its matching analysis found no treatment effect in one dataset but some evidence of one in the other, where the matching was less complete.
Source
Pedersen KM (2011). Sickness absence and voluntary employer paid health insurance. Health Economics Papers 2011:1. COHERE, University of Southern Denmark. Open source (full text read)
Passage
Using the matching estimator approach there are no signs of a treatment effect of health insurance using the presenteeism data set, while there is evidence using the health insurance data set. However, the specification of the propensity score for the latter is not as exhaustive as for presenteeism data set, and in some cases there are statistically significant differences for some control variables after matching.
Where
Abstract; the same abstract reads "However, nationally and internationally there is no evidence about this."
Limits
A working paper, not a peer-reviewed article. The mixed matching result means the study is best read as inconclusive rather than null.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2 ↩ 3
C29Supported
In Norwegian register data on the working population from 2010 to 2012, about one in four employees with a spell of sick leave was also waiting for specialist consultation or treatment, and the length of the wait had no substantial effect on the length of the absence.
Source
Dyrstad K, Halvorsen T, Hem KG, Rohde T (2016). Sick of waiting: does waiting for elective treatment cause sickness absence? Health Policy, 120(12), 1383-1388. Open source (abstract only)
Passage
Yet, the length of the waiting period had no substantial effect on the length of sickness absence (i.e., less than 0.1% reduction).
Where
Abstract
Limits
Norway, public waiting lists, not private insurance. Observational. Read at abstract level; the funder could not be checked.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2
C30Supported
A Norwegian quasi-experimental study of queues for orthopaedic surgery found that long waits increased medium to long-term work absence and permanent disability benefit, with the effect driven by people already on sick leave at referral.
Source
Godøy A, Haaland VF, Huitfeldt I, Votruba ME (2024). Hospital queues, patient health, and labor supply. American Economic Journal: Economic Policy, 16(2), 150-181. Open source (abstract only)
Passage
However, we do find persistent reductions in labor supply: long waits increase medium to long-term work absences and permanent disability receipt. Effects are driven by individuals who are on sick leave at referral.
Where
Abstract
Limits
Norway, public queues, orthopaedics only. It shows that long waits harm people already off sick; it does not test private insurance, and its effect sizes could not be read at abstract level.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2 ↩ 3 ↩ 4
C31Supported
A Norwegian randomised trial, recruiting from January 2023 with a target of 2,500 patients, is comparing assessment within four weeks against assessment within 10 to 14 weeks for people with common mental or musculoskeletal disorders, with work participation as the outcome.
Source
Aars NAP, Bardal I, Terjesen C, et al. (2026). Multicentre randomised controlled trial of the Norwegian health in work service for patients with common mental disorders or musculoskeletal disorders: the Norwegian Sickness Absence Clinic (NSAC) Efficacy Study. BMJ Open, 16, e109083. Open source (abstract only)
Passage
Patients are randomised in equal proportions to either (1) rapid HIA (assessment within 4 weeks), (2) delayed HIA (assessment within 10-14 weeks) or (3) active control, which consists of a monodisciplinary examination at HIA close to diagnosis-specific deadline for examination as suggested by guidelines (8-26 weeks).
Where
Abstract, methods and analyses
Limits
A protocol. No results yet. A public clinic, not insurance.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2
C32Supported
In Danish register data on privately employed full-time workers, employer-paid health insurance was associated with a 10 per cent reduction in use of public hospitals.
Source
Søgaard R, Pedersen MS, Bech M (2013). To what extent does employer-paid health insurance reduce the use of public hospitals? Health Policy, 113(1-2), 50-59. Open source (abstract only)
Passage
The effect of employer-paid health insurance is estimated to correspond to a significant 10% reduction in the total use of public hospitals.
Where
Abstract, results
Limits
Denmark, propensity matching on observed characteristics. Measures hospital use, not waiting times or absence. Read at abstract level.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2
C33Supported
The US job-lock estimate that employer health insurance cuts voluntary turnover from 16 to 12 per cent a year comes from married men in 1987, and rests on the cost of losing one's main health cover because of pre-existing condition exclusions.
Source
Madrian BC (1993). Employment-based health insurance and job mobility: is there evidence of job-lock? NBER Working Paper 4476. Open source (full text read)
Passage
Using data from the 1987 National Medical Expenditure Survey, I estimate that job-lock reduces the voluntary turnover rate of those with employer-provided health insurance by 25 percent, from 16 percent to 12 percent per year.
Where
Abstract; the sample section restricts the data to married men aged 20 to 55
Limits
In the United States employer cover is most people's only health insurance. The mechanism does not exist alongside the NHS. Thirty-nine years old.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2
C34Supported
In the 2009 Department for Work and Pensions employee survey, the factors associated with fewer people thinking about leaving were workplace culture, work-life balance, flexible working, engagement and age, and private medical insurance was not among them.
Source
Young V, Bhaumik C (2011). Health and well-being at work: a survey of employees. Department for Work and Pensions Research Report No 751. Open source (full text read)
Passage
Analysis revealed that a higher than average retention (i.e. few people thinking about leaving) was associated with an organisational culture where the workplace was seen as a fun place to work, generally positive views about work-life balance, and working in organisations where flexible working options were in place. Higher than average retention was also associated with high scores on the employee engagement index.
Where
Appendix E, factors associated with higher than average retention
Limits
An exploratory analysis of intentions, not of leaving. The report says the analysis cannot show causes. Benefits were entered as a group.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2
C35Overstated
Health and protection insurance delivered £6.1 billion of benefits to the UK in 2021.
Source
WPI Economics for the Association of British Insurers (2023). Futureproofing workplace health: scaling up the impact of protection and health insurance. November 2023. Open source (full text read)
Passage
Using these assumptions and the findings from the literature, we estimate that in 2021, the value of benefits from health and protection insurance amounted to £6.1bn.
Where
Chapter 3
Limits
A model commissioned by the insurers' trade body. It applies effect sizes from studies of other interventions to insurer usage data. There is no comparison group and no measurement of what insured employees did. Health insurance is combined with income protection and other products.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2 ↩ 3
C36Supported
The £6.1 billion model takes its return-to-work effects from existing studies of workplace interventions, with a central assumption of a 19 to 54 per cent improvement in time to return, then halves the result as a cautious assumption.
Source
WPI Economics for the Association of British Insurers (2023). Futureproofing workplace health: scaling up the impact of protection and health insurance. November 2023. Open source (full text read)
Passage
Estimates of the benefits below are the mid-point of the central and low estimates, with a further cautious assumption that the overall benefits are half that identified in the existing literature.
Where
Chapter 3; Table 1, Existing literature used in benefit calculations, gives "Central: 19% - 54% improvement in time to return to work depending on condition type"
Limits
Halving an assumed effect is still an assumption. The studies in the table are of vocational rehabilitation and occupational health programmes, not of medical insurance.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2 ↩ 3
C37Overstated
Private medical insurance provides a £2.5 billion boost to UK economic output every year.
Source
WPI Strategy for Bupa (2026). Healthy workforce, healthy economy: a health economics paper. April 2026. Open source (full text read)
Passage
medical insurance and provision remain under-researched.
Where
Executive summary, page 2, in the sentence "Yet the precise economic benefits of better health through initiatives like employer-supported private medical insurance and provision remain under-researched."; the same page reads "Our analysis finds that private medical insurance and provision currently provides a £2.5 billion boost"
Limits
A model commissioned by an insurer, with no methods annex, no comparison group and no published inputs for the time saved per case. The paper itself calls the benefits under-researched. Nobody has independently checked it.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2 ↩ 3 ↩ 4
C38Unsupported
For the average small business, private medical insurance means 22 fewer sickness absence days and £14,700 more output a year.
Source
WPI Strategy for Bupa (2026). Healthy workforce, healthy economy: a health economics paper. April 2026. Open source (full text read)
Passage
through private medical insurance and provision is associated with:
Where
Page 6, "Our modelling finds that for the average small and medium sized business, faster access to treatment through private medical insurance and provision is associated with:", followed by a table reading 22 extra workdays, 24 equivalent days from productivity and £14.7K
Limits
The paper says its modelling finds an association; it did not measure absence in any business. The figure is per business, and the assumed days saved per case are not published.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2 ↩ 3
C39Overstated
Raising take-up of insurance-based workplace health solutions among large businesses by 25 per cent could deliver £750 million of benefits.
Source
WPI Economics for AXA Health (2025). Health and work: improving the resilience of the UK workforce. July 2025. Open source (full text read)
Passage
based solutions amongst large businesses by 25% could lead to benefits of £750m.
Where
Chapter 3, page 12, in the sentence beginning "We adapted the modelling undertaken for the ABI"
Limits
An adaptation of the trade body model, commissioned by an insurer. Same assumptions, no comparison group, no independent check.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2
C40Unsupported
One health insurance provider states that employers get an estimated £2.87 back for every £1 spent on its plans and that it has proven the plans pay off.
Source
Equipsme (2026). ROI 2026-2027: employers recoup £2.87 for every £1 spent on Equipsme plans. Company blog, undated, citing a white paper dated March 2026. Open source (full text read)
Passage
And we’ve proven that it pays off
Where
Quoted statement in the blog post; the opening line reads "Employers investing in health insurance plans from Equipsme stand to reap an estimated £2.87 return on investment (ROI)."
Limits
The return is a calculation from assumed absence savings for a model employee on an average salary, not a measured comparison of absence with and without the plan. The same page says the plans get people back to work twice as fast as the NHS, which restates a waiting-time estimate as a return-to-work result.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2 ↩ 3
C41Supported
Employer-paid medical insurance is a taxable benefit with tax, National Insurance and reporting obligations for the employer.
Source
HM Revenue and Customs. Expenses and benefits: medical or dental treatment and insurance. GOV.UK guidance. Open source (full text read)
Passage
As an employer providing medical or dental treatment or insurance to your employees, you have certain tax, National Insurance and reporting obligations.
Where
Overview
Limits
Guidance, not the legislation. Exemptions exist for some treatment to help a return to work, which this page links to.
Checked
3 Nov 2026
Back to text↩ 1
C42Supported
The employer's Class 1A National Insurance rate on expenses and benefits for 2026 to 2027 is 15 per cent.
Source
HM Revenue and Customs (2026). Rates and thresholds for employers 2026 to 2027. GOV.UK guidance. Open source (full text read)
Passage
The National Insurance Class 1A rate on expenses and benefits for 2026 to 2027 is 15%.
Where
Class 1A National Insurance, expenses and benefits
Limits
Applies to benefits reported on P11D. Payrolled benefits attract the same charge through a different route.
Checked
3 Nov 2026
Back to text↩ 1
C43Supported
The standard rate of insurance premium tax is 12 per cent.
Source
HM Revenue and Customs. Insurance Premium Tax rates. GOV.UK. Open source (full text read)
Passage
Standard rate | 12% |
Where
Rates table, standard rate row
Limits
The page gives the rate, not what it is charged on; medical insurance premiums are standard rated.
Checked
3 Nov 2026
Back to text↩ 1
C44Supported
The Keep Britain Working final report mentions private medical insurance once, to repeat the CIPD's 2025 figures that 25 per cent of surveyed respondents have it for all employees, 40 per cent of large private sector employers, 20 per cent of SMEs and 7 to 11 per cent of public and non-profit organisations.
Source
Mayfield C (2025). Keep Britain Working: final report. Independent review for the Department for Work and Pensions, 5 November 2025. Open source (full text read)
Passage
The CIPD’s Health and wellbeing report at work (2025) found that that only 25% of all surveyed respondents have private medical insurance for all employees
Where
Appendix A, the current workplace health services landscape; the sentence continues with 40 per cent for large private sector employers, 20 per cent for SMEs and 7 to 11 per cent for the public and non-profit sectors. A search of the report for "private medical" finds no other occurrence
Limits
The report repeats a survey of HR professionals supported by a health cash plan provider, without a footnote. The figures describe organisations offering the benefit to all staff, not employees covered, and leave out the further 25 per cent offering it by seniority.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2 ↩ 3
C45Supported
The Keep Britain Working final report's list of return-on-investment evidence covers employee assistance programmes, mental health and musculoskeletal interventions and employer cases, and gives none for private medical insurance.
Source
Mayfield C (2025). Keep Britain Working: final report. Independent review for the Department for Work and Pensions, 5 November 2025. Open source (full text read)
Passage
125. Return on investment (RoI) is a powerful driver of adoption. Evidence already shows that managing health and disability at work delivers strong returns.
Where
Incentives, paragraph 125; the numbered list that follows begins "1. Employee Assistance Programmes (EAPs): An EAP association report (2022) found an average £8 return for every £1 invested"
Limits
Absence from a list is not a judgement by the review; it may simply have found nothing to cite, which is consistent with this review.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2
C46Supported
Among possible incentives, the Keep Britain Working final report lists tax relief on benefits in kind for employers certified as adopters, and says such levers must be tied to evidence of impact.
Source
Mayfield C (2025). Keep Britain Working: final report. Independent review for the Department for Work and Pensions, 5 November 2025. Open source (full text read)
Passage
employers). These must be tied to evidence of impact and designed to endure.
Where
Incentives, paragraph 135, in the sentence "While a positive return on investment is the most powerful incentive, additional levers will be needed to secure widespread adoption (particularly for smaller employers)."; the list under financial incentives includes "Tax relief on benefits in kind"
Limits
An option in a list, not a proposal on medical insurance or on insurance premium tax, neither of which the report names. Budget 2026 contained no such relief.
Checked
3 Nov 2026
Back to text↩ 1
C47Supported
The September 2026 Keep Britain Working update suggests insurance schemes that pool risk across small employers, building on approaches such as group income protection, and does not name private medical insurance.
Source
Mayfield C (2026). Keep Britain Working: a national growth opportunity. Review update, Department for Work and Pensions, 23 September 2026. Open source (full text read)
Passage
This would draw on learning from approaches in some European countries and would build on insurance approaches that also already exist, such as group income protection, where workplace health support is already a prominent benefit.
Where
Chapter four, page 18; the preceding sentence reads "An alternative route for small businesses may be through insurance schemes that pool risks and support across small employers, so that premiums are priced lower across a larger group of employees." A search of the document for "private medical" finds nothing
Limits
Exploratory wording in a progress update. The same passage says such schemes would need to address pricing, adverse risk selection and brokerage.
Checked
3 Nov 2026
Back to text↩ 1 ↩ 2 ↩ 3
C48Supported
The same update states that today a provider sells a service based on claims of efficacy, and proposes that providers compete on performance data published in a common format.
Source
Mayfield C (2026). Keep Britain Working: a national growth opportunity. Review update, Department for Work and Pensions, 23 September 2026. Open source (full text read)
Passage
Today, a provider sells a service based on claims of efficacy.
Where
Chapter five, better intelligence; the next sentence reads "With better intelligence, in the Workplace Health System, providers could compete based on performance data, published in a common format."
Limits
A statement about workplace health provision in general, not about medical insurance specifically.
Checked
3 Nov 2026
Back to text↩ 1
C49Supported
The June 2026 Keep Britain Working update commits to building understanding of risk pooling and insurance approaches to make provision affordable for small businesses.
Source
Keep Britain Working (2026). Keep Britain Working: the story so far, June 2026. Department for Work and Pensions, published 3 July 2026. Open source (full text read)
Passage
Build understanding of risk pooling and insurance approaches to make provision accessible and affordable for small businesses.
Where
Workstream table, workplace health provision row
Limits
A workstream heading. No product is named.
Checked
3 Nov 2026
Back to text↩ 1
C50Supported
In a trade press analysis of group private medical insurance's role in Keeping Britain Working, published on 2 November 2026, a broker contributor said the sector would need to prove that the benefit reduces absence.
Source
Health & Protection (2026). Group PMI's vital role in Keeping Britain Working as demand and costs climb: analysis. 2 November 2026. Open source (full text read)
Passage
The sector will need to keep prices accessible for small businesses while proving that the benefit reduces absence.
Where
Section on preventative health, quoted from a broker; the article quotes eleven people from insurers, brokers and consultancies and names no study or dataset. The only figure given is one insurer's online GP appointment volumes
Limits
The contributor is not named here by choice. The article is a round-up of industry views and does not cite studies; it was not written as evidence. A statement that proof is still needed is consistent with this review, which found no study showing that group PMI reduces absence.
Checked
3 Nov 2026
Back to text↩ 1
C51Supported
In the same article a consultancy contributor said that where reliable data are unavailable, informed decisions and robust business cases are much harder to make.
Source
Health & Protection (2026). Group PMI's vital role in Keeping Britain Working as demand and costs climb: analysis. 2 November 2026. Open source (full text read)
Passage
However, clients have limited budgets and, where reliable data is unavailable, informed decisions and robust business cases are much harder to make.
Where
Section on value-added services
Limits
The contributor is not named here by choice. The statement is about client decisions, and is consistent with the absence of independent outcome data found here.
Checked
3 Nov 2026
Back to text↩ 1
C52Supported
Private medical insurance does not normally cover pre-existing conditions or the ongoing management of chronic conditions such as asthma, diabetes or eczema, though larger employer schemes are more likely to cover pre-existing conditions.
Source
Private Healthcare Information Network (2026). Paying for treatment with private medical insurance. Help page, last updated 8 July 2026. Open source (full text read)
Passage
Any cover you buy from a private medical insurer won’t usually cover the costs of the ongoing management and maintenance of chronic, or incurable, conditions such as asthma, diabetes, or eczema.
Where
Section on what is covered; the same page reads "the larger the number of employees, the likelihood is that pre-existing medical conditions might be covered by the insurance arranged by your employer"
Limits
Consumer guidance from the private healthcare information body, which is funded by the providers it regulates information for. Policies vary; no study measures how exclusions change who in a workforce benefits.
Checked
3 Nov 2026
Back to text↩ 1
C53Supported
In English household panel data from 1997 to 2000, income, age and sex were key determinants of holding employer-paid private medical insurance as well as individually bought cover.
Source
King D, Mossialos E (2005). The determinants of private medical insurance prevalence in England, 1997-2000. Health Services Research, 40(1), 195-212. Open source (abstract only)
Passage
Income, age, sex, and political preference are key determinants of PMI prevalence for both individual and employer paid PMI.
Where
Abstract, principal findings
Limits
Twenty-six years old. Read at abstract level. Shows who holds cover, not what it does for them.
Checked
3 Nov 2026
Back to text↩ 1

Back to the notePrivate medical insurance is sold on absence. Nobody has tested that.