13 Oct 2026 · Evidence review · in progress

Employee benefits against outcomes: what the published evidence supports

The note this review supports320 questions about employee benefits

This review sets out what published studies found when common UK workplace benefits were tested against the outcomes employers buy them for, and it reports each grade as graded, including where nothing was found. It supports the note 320 questions about employee benefits.

How the grades were made

The grid is thirty-two benefits, from group income protection to discount platforms, each against ten outcomes: sickness absence, long-term absence and return to work, productivity at work, staff turnover, recruitment, mental health symptoms, physical and musculoskeletal health, financial wellbeing, engagement, and take-up. That gives 320 pairs.

Each pair was searched in Europe PMC, Crossref and on the open web, then graded with GRADE logic. Randomised trials and reviews of them start high and observational studies start low, and both lose grades for risk of bias, indirectness, inconsistency and imprecision. Industry and insurer surveys were recorded but never allowed to set a grade. Most sources were read at abstract level.

The evidence base was built at Alltoogether on 6 October 2026. A random fifth of the pairs with any evidence (27 of 132) was re-checked against every source by reviewers who had not graded them. One failed, around four in ten needed wording corrected, and all of those corrections are applied. The other four fifths have not had that second check. Every grade here is provisional until they have. Each claim in the ledger was re-read against its source on 6 October 2026.

The headline

None of the 320 pairs reached high certainty. Fourteen reached moderate, 54 low and 64 very low. For 188 there was no usable published study.

Not every gap matters. Around 30 of the 188 are pairs nobody has a plausible reason to study, such as electric car salary sacrifice and back pain. The rest are open questions. A gap in the literature says nothing either way about whether a benefit works.

The whole grid is below, and the sections after it walk through the boxes that matter most.

Benefit	Absence	Long absence and return	Productivity	Turnover	Recruitment	Mental health	Physical health	Financial wellbeing	Engagement	Take-up
Insured cover
Group income protection	Absence: low, mixed	Long absence and return: low, mixed	Productivity: none found	Turnover: none found	Recruitment: none found	Mental health: very low, mixed	Physical health: low, mixed	Financial wellbeing: none found	Engagement: none found	Take-up: none found
Group life assurance	Absence: none found	Long absence and return: none found	Productivity: none found	Turnover: very low, direction unknown	Recruitment: none found	Mental health: none found	Physical health: none found	Financial wellbeing: very low, no clear effect	Engagement: none found	Take-up: none found
Group critical illness	Absence: none found	Long absence and return: none found	Productivity: none found	Turnover: none found	Recruitment: none found	Mental health: very low, improves	Physical health: none found	Financial wellbeing: very low, mixed	Engagement: none found	Take-up: none found
Private medical insurance	Absence: very low, no clear effect	Long absence and return: very low, no clear effect	Productivity: none found	Turnover: very low, mixed	Recruitment: none found	Mental health: none found	Physical health: none found	Financial wellbeing: none found	Engagement: none found	Take-up: low, mixed
Health cash plan	Absence: none found	Long absence and return: none found	Productivity: none found	Turnover: none found	Recruitment: none found	Mental health: none found	Physical health: none found	Financial wellbeing: none found	Engagement: none found	Take-up: none found
Group dental	Absence: very low, improves	Long absence and return: none found	Productivity: very low, improves	Turnover: none found	Recruitment: none found	Mental health: none found	Physical health: very low, mixed	Financial wellbeing: very low, no clear effect	Engagement: none found	Take-up: low, improves
Services
Employee assistance programme	Absence: very low, mixed	Long absence and return: very low, improves	Productivity: low, improves	Turnover: very low, direction unknown	Recruitment: none found	Mental health: low, improves	Physical health: very low, direction unknown	Financial wellbeing: none found	Engagement: very low, mixed	Take-up: low, direction unknown
Workplace counselling	Absence: very low, mixed	Long absence and return: very low, improves	Productivity: low, improves	Turnover: none found	Recruitment: none found	Mental health: low, improves	Physical health: none found	Financial wellbeing: none found	Engagement: very low, mixed	Take-up: low, direction unknown
Digital mental health programmes	Absence: low, no clear effect	Long absence and return: low, mixed	Productivity: low, mixed	Turnover: none found	Recruitment: none found	Mental health: moderate, improves	Physical health: none found	Financial wellbeing: none found	Engagement: low, improves	Take-up: low, direction unknown
Virtual GP	Absence: none found	Long absence and return: none found	Productivity: none found	Turnover: none found	Recruitment: none found	Mental health: low, no clear effect	Physical health: low, no clear effect	Financial wellbeing: none found	Engagement: very low, improves	Take-up: very low, improves
Second medical opinion	Absence: none found	Long absence and return: low, no clear effect	Productivity: none found	Turnover: none found	Recruitment: none found	Mental health: none found	Physical health: very low, direction unknown	Financial wellbeing: none found	Engagement: very low, improves	Take-up: very low, direction unknown
Health checks and assessments	Absence: low, no clear effect	Long absence and return: very low, no clear effect	Productivity: low, mixed	Turnover: low, no clear effect	Recruitment: none found	Mental health: low, mixed	Physical health: moderate, no clear effect	Financial wellbeing: none found	Engagement: low, no clear effect	Take-up: low, direction unknown
Physiotherapy	Absence: low, mixed	Long absence and return: low, improves	Productivity: very low, no clear effect	Turnover: none found	Recruitment: none found	Mental health: none found	Physical health: moderate, no clear effect	Financial wellbeing: none found	Engagement: none found	Take-up: very low, direction unknown
Occupational health	Absence: low, mixed	Long absence and return: low, improves	Productivity: none found	Turnover: very low, direction unknown	Recruitment: none found	Mental health: low, mixed	Physical health: very low, direction unknown	Financial wellbeing: none found	Engagement: none found	Take-up: moderate, direction unknown
Return-to-work support	Absence: moderate, improves	Long absence and return: moderate, improves	Productivity: low, no clear effect	Turnover: very low, direction unknown	Recruitment: none found	Mental health: low, improves	Physical health: low, improves	Financial wellbeing: none found	Engagement: none found	Take-up: low, direction unknown
Money advice and financial education	Absence: none found	Long absence and return: none found	Productivity: none found	Turnover: none found	Recruitment: none found	Mental health: very low, mixed	Physical health: very low, no clear effect	Financial wellbeing: moderate, improves	Engagement: none found	Take-up: low, mixed
Legal advice line	Absence: none found	Long absence and return: none found	Productivity: none found	Turnover: none found	Recruitment: none found	Mental health: very low, mixed	Physical health: very low, direction unknown	Financial wellbeing: very low, improves	Engagement: none found	Take-up: very low, direction unknown
Bereavement counselling	Absence: none found	Long absence and return: none found	Productivity: none found	Turnover: none found	Recruitment: none found	Mental health: low, mixed	Physical health: none found	Financial wellbeing: none found	Engagement: none found	Take-up: none found
Manager assist line	Absence: very low, mixed	Long absence and return: none found	Productivity: none found	Turnover: none found	Recruitment: none found	Mental health: very low, no clear effect	Physical health: none found	Financial wellbeing: none found	Engagement: none found	Take-up: none found
Sleep support	Absence: low, mixed	Long absence and return: none found	Productivity: low, mixed	Turnover: none found	Recruitment: none found	Mental health: moderate, improves	Physical health: none found	Financial wellbeing: none found	Engagement: very low, mixed	Take-up: very low, mixed
Stop smoking support	Absence: very low, improves	Long absence and return: none found	Productivity: very low, improves	Turnover: none found	Recruitment: none found	Mental health: very low, improves	Physical health: low, improves	Financial wellbeing: none found	Engagement: none found	Take-up: low, mixed
Nutrition and fitness programmes	Absence: moderate, no clear effect	Long absence and return: none found	Productivity: low, mixed	Turnover: moderate, no clear effect	Recruitment: none found	Mental health: low, mixed	Physical health: moderate, improves	Financial wellbeing: none found	Engagement: low, mixed	Take-up: moderate, mixed
Other benefits
Pensions and pension salary sacrifice	Absence: none found	Long absence and return: none found	Productivity: none found	Turnover: very low, no clear effect	Recruitment: none found	Mental health: none found	Physical health: none found	Financial wellbeing: low, improves	Engagement: none found	Take-up: moderate, improves
Cycle to Work	Absence: very low, improves	Long absence and return: very low, improves	Productivity: none found	Turnover: very low, direction unknown	Recruitment: very low, direction unknown	Mental health: very low, improves	Physical health: very low, mixed	Financial wellbeing: none found	Engagement: none found	Take-up: low, mixed
Electric car salary sacrifice	Absence: none found	Long absence and return: none found	Productivity: none found	Turnover: none found	Recruitment: none found	Mental health: none found	Physical health: none found	Financial wellbeing: none found	Engagement: none found	Take-up: none found
Flexible benefits	Absence: none found	Long absence and return: none found	Productivity: none found	Turnover: very low, direction unknown	Recruitment: low, improves	Mental health: none found	Physical health: none found	Financial wellbeing: none found	Engagement: very low, improves	Take-up: none found
Wellbeing allowances	Absence: none found	Long absence and return: none found	Productivity: none found	Turnover: none found	Recruitment: none found	Mental health: none found	Physical health: none found	Financial wellbeing: none found	Engagement: very low, improves	Take-up: none found
Recognition schemes	Absence: low, mixed	Long absence and return: none found	Productivity: low, improves	Turnover: very low, improves	Recruitment: none found	Mental health: very low, no clear effect	Physical health: none found	Financial wellbeing: none found	Engagement: very low, improves	Take-up: very low, direction unknown
Discounts and perks	Absence: none found	Long absence and return: none found	Productivity: none found	Turnover: none found	Recruitment: none found	Mental health: none found	Physical health: none found	Financial wellbeing: none found	Engagement: none found	Take-up: very low, direction unknown
Wellbeing apps	Absence: low, no clear effect	Long absence and return: none found	Productivity: low, no clear effect	Turnover: none found	Recruitment: none found	Mental health: moderate, improves	Physical health: very low, no clear effect	Financial wellbeing: none found	Engagement: low, mixed	Take-up: low, mixed
Gym subsidy	Absence: very low, improves	Long absence and return: none found	Productivity: none found	Turnover: none found	Recruitment: none found	Mental health: none found	Physical health: very low, improves	Financial wellbeing: none found	Engagement: none found	Take-up: low, mixed
Mental Health First Aid training	Absence: none found	Long absence and return: none found	Productivity: none found	Turnover: none found	Recruitment: none found	Mental health: very low, no clear effect	Physical health: none found	Financial wellbeing: none found	Engagement: very low, direction unknown	Take-up: low, mixed

The grid, as graded on 6 October 2026. Each box is the certainty of the published evidence for that benefit and outcome: Mod is moderate, Low is low, V low is very low, and an empty box means no usable study was found. The sign is the direction: + improves, = no clear effect, ± mixed, ? unknown. In the take-up column the grade is for the take-up figures themselves, and ? means descriptive figures with no comparison. No box reached high. Every grade is provisional.

Benefits with no outcome studies

Health cash plans have no usable study on any of the ten outcomes. The figure most often repeated, that a cash plan could save an average of 4.1 sick days, sits in a trade article with no sample, method or comparison C1.

Electric car salary sacrifice has none on any of the ten either, though five of its pairs have no plausible pathway to study. Discount and perks platforms have none on nine outcomes. The tenth, take-up, rests on one record: a UK fire and rescue authority found 6.1% of its staff used its discount website in 2010, the scheme’s second year C2. That is graded very low. The benefits with the least outcome evidence include some of the most widely offered.

Evidence that a benefit does not do what it is often bought for

Moderate evidence of no effect is a different finding from no evidence, and several pairs have it.

Workplace nutrition and fitness programmes rest on two large US randomised trials. Across 160 worksites of one retailer, a programme had no significant effect on absenteeism, job tenure or job performance after 18 months C3. In Illinois, the people who joined were already healthier and spent less on medical care in the year before C4, so comparing joiners with everyone else flatters a programme. Grade: moderate, no clear effect, for both sickness absence and turnover.

General health checks have a Cochrane review of 17 trials whose authors concluded they are unlikely to be beneficial C5. Grade: moderate, no clear effect on physical health, downgraded once because the trials were in the general population.

Faster physiotherapy has two UK trials. A telephone-first service gave faster access and equivalent outcomes to waiting-list care C6, and immediate physiotherapy for arm pain gave no added benefit over physiotherapy after a six to eight week wait C7. Grade: moderate, no clear effect on physical health. The pair compares faster physiotherapy with slower physiotherapy, so it measures speed of access. In these trials a faster route bought a shorter wait and the same recovery.

Employer-paid medical insurance was given a tax exemption in Denmark partly on the argument that it would reduce long-term absence C8. An analysis of two Danish surveys found no significant absence-reducing effect C9. Grade: very low, no clear effect, because the data are observational and one matched dataset disagreed.

Where the moderate evidence points

Return-to-work support that involves the workplace has the strongest positive grades in the grid. A Cochrane review of 14 trials found workplace interventions shortened time to first return and reduced total time off sick C10. For depression, combining work-directed and clinical support probably cut sick days C11, by about 25 days a year C12, though after a year or more no more people were at work. Coordination programmes on their own, the kind insurers, healthcare providers and employers run, showed no benefit over usual practice C13. Grade: moderate, improves, for sickness absence and for time to return.

The same pattern appears at lower grades. Physiotherapy with an early workplace dialogue raised the share of patients able to work at one year from 74% to 85% C14. An occupational health consultation for staff at high risk of absence cut mean absence from 30 days to 19 C15, while phone advice for staff at intermediate risk did nothing, mainly because few used it C16. Grades: low, mixed, for physiotherapy and for occupational health on sickness absence. In each of these studies where absence fell, the workplace or its occupational health service was part of the intervention.

Digital mental health programmes produced small reductions in depression, anxiety and stress across 81 trials C17, though most of those trials were at high risk of bias C18. Grade: moderate, improves, and one of the reviewers argued for low. A mindfulness app reduced stress among NHS staff, with small effects C19. Grade: moderate, improves, for wellbeing apps on mental health symptoms. Workplace insomnia programmes reduced insomnia symptoms C22. Grade: moderate, improves, graded on insomnia; on depression and anxiety it would be low.

The same programmes and sickness absence

The NHS app trial found no effect on sickness absence C20, and across 11 trials therapist-guided online CBT did not reduce it C21. Grade: low, no clear effect, for both digital mental health programmes and wellbeing apps. On these data, small gains in symptoms did not turn into fewer days off.

Take-up

Most of these trials report the effect on everyone offered the programme, whether or not they used it, and use is often low. Participation in workplace activity and nutrition programmes had a median of 33% C23. Staff given only an app used it for 45 of 343 recommended minutes C24 and saw no significant benefit, while those who also had classes did C25.

In a 2020 survey of the workforce in Great Britain, half of employees had occupational health access C26, a fifth did not know whether they had it C27, and 70% of those with access through their employer had not used it in the past year C28. Grade: moderate, descriptive. A benefit’s effect in a trial depends on people knowing it exists and using it.

Two common assumptions

A Cochrane review of Mental Health First Aid training could draw no conclusions about its effect on the mental health of people in the communities where it was delivered C29. Grade: very low, no clear effect.

For mental health screening, a review of eight workplace trials found that screening alone does not improve worker mental health C30; screening followed by facilitated treatment showed a small improvement. Grade: low, mixed, applied to health assessments. Finding out who is struggling helped only when it led to treatment.

Limits

The grades are Alltoogether’s, made with GRADE logic and without a guideline panel. Most sources were read at abstract level. Of the 132 graded pairs, 34 rest on UK evidence. Most of the trials tested a service or a programme, often in clinical settings. That is one step removed from the insured or employer-funded benefit itself, and every one of those trials was downgraded for it. Several grades are arguable by one level, and the reviewers said so in the record. An association found in observational studies is not evidence of a cause.

What would change these grades

For health cash plans, a controlled before-and-after study or a staggered roll-out across UK employers using payroll absence records. Discount platforms would need a job-advert experiment that varies whether discounts are mentioned, and a before-and-after study with a validated financial wellbeing scale and real transaction data. For insurer early intervention, a randomised trial against usual care for employees off with common mental disorders. Wellbeing apps and staff turnover would need a stepped roll-out with turnover taken from employer records. Each would move its pair off none found, in whichever direction the data point.

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
The figure that a health cash plan could save an average of 4.1 sick days appears in a trade article that gives no sample, method or comparison group.
Source
Hulme J (2025). Health cash plan myths: the hidden value employers can't afford to ignore. REBA. Open source (full text read)
Passage
an average of 4.1 sick days could be saved if employees had access to a health cash plan
Limits
The article is written by a provider and does not say how the figure was produced. A missing method does not show the figure is wrong, only that it cannot be checked.
Checked
6 Oct 2026
Back to text↩ 1
C2Supported
A UK fire and rescue authority found that 6.1% of its staff used its employee discount website during 2010, the scheme's second year.
Source
Nottinghamshire and City of Nottingham Fire and Rescue Authority (2011). Employee discount scheme: report of the Chief Fire Officer to the Human Resources Committee, 28 January 2011. Open source (full text read)
Passage
6.1% of (67) employees have accessed the site during 2010
Where
paragraph 2.7
Limits
One public employer, one scheme, provider usage figures from 2010. It describes take-up and measures no outcome. Discount platforms have changed a great deal since.
Checked
6 Oct 2026
Back to text↩ 1
C3Supported
In a randomised trial across 160 worksites of one US retailer, a workplace wellness programme had no significant effect on absenteeism, job tenure or job performance after 18 months.
Source
Song Z, Baicker K (2019). Effect of a workplace wellness program on employee health and economic outcomes: a randomized clinical trial. JAMA, 321(15), 1491. Open source (abstract only)
Passage
The program had no significant effects on other prespecified outcomes: 27 self-reported health outcomes and behaviors (including self-reported health, sleep quality, and food choices), 10 clinical markers of health (including cholesterol, blood pressure, and body mass index), 38 medical and pharmaceutical spending and utilization measures, and 3 employment outcomes (absenteeism, job tenure, and job performance).
Limits
One US warehouse retailer, 18 months of follow-up. Staff completed a mean of 1.3 of the 8 modules, so the trial tests the programme as offered, with low take-up.
Checked
6 Oct 2026
Back to text↩ 1
C4Supported
In a randomised trial of a wellness programme with nearly 5,000 US employees, the people who took part already had lower medical spending and healthier habits in the year before the programme.
Source
Jones D, Molitor D, Reif J (2019). What do workplace wellness programs do? Evidence from the Illinois Workplace Wellness Study. Quarterly Journal of Economics, 134(4), 1747-1791. Open source (abstract only)
Passage
We find strong patterns of selection: during the year prior to the intervention, program participants had lower medical expenditures and healthier behaviors than nonparticipants.
Limits
One US employer. A systematic review of participation found no consistent link between health and joining, so the size of this selection elsewhere is uncertain.
Checked
6 Oct 2026
Back to text↩ 1
C5Supported
A Cochrane review of 17 trials concluded that general health checks are unlikely to be beneficial.
Source
Krogsbøll LT, Jørgensen KJ, Gøtzsche PC (2019). General health checks in adults for reducing morbidity and mortality from disease. Cochrane Database of Systematic Reviews, CD009009. Open source (abstract only)
Passage
General health checks are unlikely to be beneficial.
Limits
Trials in the general adult population, not checks provided by employers. The outcomes are death, heart disease and stroke, not absence or wellbeing.
Checked
6 Oct 2026
Back to text↩ 1
C6Supported
In an English randomised trial, a telephone-first physiotherapy service gave faster access to treatment and outcomes equivalent to waiting-list care.
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 (abstract only)
Passage
Providing physiotherapy via PhysioDirect is equally clinically effective compared with usual waiting list-based care, provides faster access to treatment, appears to be safe, and is broadly acceptable to patients.
Limits
NHS community physiotherapy patients, not an employer-funded benefit. The median wait was 7 days against 34.
Checked
6 Oct 2026
Back to text↩ 1
C7Supported
For arm pain, immediate physiotherapy gave no added benefit over physiotherapy after a six to eight week wait.
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 (abstract only)
Passage
immediate physiotherapy confers no additional benefit in terms of disability, compared with physiotherapy delivered after 6-8 weeks waiting time.
Limits
One condition, UK community referrals. Advice to stay active while waiting did better than advice to rest.
Checked
6 Oct 2026
Back to text↩ 1
C8Supported
Denmark's tax exemption for employer-paid health insurance was argued partly on its potential to reduce long-term sickness absence.
Source
Pedersen KM (2011). Sickness absence and voluntary employer paid health insurance. COHERE working paper 2011:1, University of Southern Denmark. Open source (full text read)
Passage
The tax exempt status of this type of voluntary duplicate health insurance has been argued by reference to the potential for reducing long term sickness absence.
Where
Abstract
Limits
Denmark, where this insurance duplicates public cover, as it does in the UK. The policy argument is reported by the author.
Checked
6 Oct 2026
Back to text↩ 1
C9Supported
An analysis of two Danish surveys found no significant absence-reducing effect of employer-paid health insurance.
Source
Pedersen KM (2011). Sickness absence and voluntary employer paid health insurance. COHERE working paper 2011:1, University of Southern Denmark. Open source (abstract only)
Passage
No significant results are found on the absence reducing property of health insurance.
Where
Abstract
Limits
Observational survey data. In a matching analysis, one of the two datasets did show an effect, and the author notes that its specification was weaker.
Checked
6 Oct 2026
Back to text↩ 1
C10Supported
A Cochrane review of 14 trials found workplace interventions shortened time to first return to work (moderate quality) and reduced total time off sick (high quality).
Source
van Vilsteren M, van Oostrom SH, de Vet HCW, et al. (2015). Workplace interventions to prevent work disability in workers on sick leave. Cochrane Database of Systematic Reviews, CD006955. Open source (abstract only)
Passage
We found moderate-quality evidence that workplace interventions reduce time to first RTW, high-quality evidence that workplace interventions reduce cumulative duration of sickness absence
Limits
1,897 workers, mostly with musculoskeletal disorders, mostly outside the UK. The same review found very low quality evidence on lasting return, moderate quality evidence of more recurrences of sick leave, and no considerable effect for mental health problems or cancer.
Checked
6 Oct 2026
Back to text↩ 1
C11Supported
For depression, combining work-directed and clinical support probably reduces sick days, but after a year or more no more people are at work.
Source
Nieuwenhuijsen K, Verbeek JH, Neumeyer-Gromen A, et al. (2020). Interventions to improve return to work in depressed people. Cochrane Database of Systematic Reviews, CD006237. Open source (abstract only)
Passage
A combination of a work-directed intervention and a clinical intervention probably reduces the number of sickness absence days, but at the end of one year or longer follow-up, this does not lead to more people in the intervention group being at work.
Where
Authors' conclusions
Limits
45 trials of people with depressive disorders or high symptoms, not employer benefit schemes. The sick-day finding is moderate certainty; the finding on being at work after a year is high certainty.
Checked
6 Oct 2026
Back to text↩ 1
C12Supported
The reduction in sick days for depression translates to about 25 fewer days in a year.
Source
Nieuwenhuijsen K, Verbeek JH, Neumeyer-Gromen A, et al. (2020). Interventions to improve return to work in depressed people. Cochrane Database of Systematic Reviews, CD006237. Open source (abstract only)
Passage
This translates back to 0.5 fewer (95% CI -0.7 to -0.2) sick leave days in the past two weeks or 25 fewer days during one year (95% CI -37.5 to -11.8).
Where
Main results
Limits
The authors' back-translation of a pooled standardised estimate from 9 studies. The range runs from about 12 to about 38 days.
Checked
6 Oct 2026
Back to text↩ 1
C13Supported
Return-to-work coordination programmes, which insurers, healthcare providers and employers run, showed no benefit over usual practice across 14 trials.
Source
Vogel N, Schandelmaier S, Zumbrunn T, et al. (2017). Return-to-work coordination programmes for improving return to work in workers on sick leave. Cochrane Database of Systematic Reviews, CD011618. Open source (abstract only)
Passage
Offering return-to-work coordination programmes for workers on sick leave for at least four weeks results in no benefits when compared to usual practice.
Where
Authors' conclusions
Limits
12,568 workers in nine countries, mostly musculoskeletal problems. Coordination is one service; it is not the whole of what an insurer or employer may offer.
Checked
6 Oct 2026
Back to text↩ 1
C14Supported
Adding an early workplace dialogue to physiotherapy raised the share of patients with work ability at one year from 74% to 85%.
Source
Sennehed CP, Holmberg S, Axén I, et al. (2018). Early workplace dialogue in physiotherapy practice improved work ability at 1-year follow-up. Pain, 159(8), 1456-1464. Open source (abstract only)
Passage
Work ability was reached by significantly more patients in the intervention group (108/127, 85%) compared with the reference group (127/171, 74%) (P = 0.02).
Limits
One Swedish trial in primary care, neck and back pain, patients at risk of sick leave. Not an employer-funded benefit.
Checked
6 Oct 2026
Back to text↩ 1
C15Supported
An occupational health consultation for employees at high risk of absence cut mean sickness absence from 30 days to 19 over a year.
Source
Taimela S, Malmivaara A, Justén S, et al. (2008). The effectiveness of two occupational health intervention programmes in reducing sickness absence among employees at risk. Occupational and Environmental Medicine, 65(4), 236-241. Open source (abstract only)
Passage
In the trial for the high risk group the mean sickness absence was 30 days in the usual care group and 19 days in the intervention group
Limits
Finland, construction, service and maintenance staff, 88% men. Register data on absence.
Checked
6 Oct 2026
Back to text↩ 1
C16Supported
In the same study, phone advice for employees at intermediate risk did not reduce absence, mainly because few used it.
Source
Taimela S, Malmivaara A, Justén S, et al. (2008). The effectiveness of two occupational health intervention programmes in reducing sickness absence among employees at risk. Occupational and Environmental Medicine, 65(4), 236-241. Open source (abstract only)
Passage
The phone advice intervention for the intermediate risk group was not effective in controlling work loss primarily due to poor adherence.
Where
Conclusions
Limits
As C15.
Checked
6 Oct 2026
Back to text↩ 1
C17Supported
Across 81 randomised trials with 25,500 employed adults, workplace digital mental health programmes produced small reductions in depression, anxiety and stress.
Source
Stratton E, Morris RW, Milton A, et al. (2025). Optimizing workplace digital mental health interventions: systematic review and meta-analysis. Journal of Medical Internet Research, 27, e71253. Open source (abstract only)
Passage
We included 81 RCTs evaluating 98 different DMHIs in 25,500 participants. Small but significant pooled effect sizes were identified for all 3 outcomes
Limits
Mostly outside the UK, mostly against waitlist controls. The abstract prints the depression estimate with a positive sign against a wholly negative interval; the direction is a reduction.
Checked
6 Oct 2026
Back to text↩ 1
C18Supported
Most of the trials in that review were at high risk of bias.
Source
Stratton E, Morris RW, Milton A, et al. (2025). Optimizing workplace digital mental health interventions: systematic review and meta-analysis. Journal of Medical Internet Research, 27, e71253. Open source (abstract only)
Passage
In total, 71 (82.70%) trials had high risk bias, 4 (4.60%) had some concerns, and 11 (12.70%) had low risk bias, largely due to waitlist controls and inability to blind participants.
Limits
As C17.
Checked
6 Oct 2026
Back to text↩ 1
C19Supported
In a trial with 2,182 NHS staff in England, an unguided mindfulness app reduced stress more than an active control, with small effects.
Source
Taylor H, Cavanagh K, Field AP, Strauss C (2022). Health care workers' need for Headspace: findings from a multisite definitive randomized controlled trial of an unguided digital mindfulness-based self-help app to reduce healthcare worker stress. JMIR mHealth and uHealth, 10(8), e31744. Open source (abstract only)
Passage
Intention-to-treat analyses found that Headspace led to greater reductions in stress over time than Moodzone (b=-0.31, 95% CI -0.47 to -0.14; P<.001), with small effects.
Limits
Healthcare workers recruited online, 4.5 months. 41% of the app group met the per-protocol use threshold.
Checked
6 Oct 2026
Back to text↩ 1
C20Supported
The same trial found no effect of the app on sickness absence.
Source
Taylor H, Cavanagh K, Field AP, Strauss C (2022). Health care workers' need for Headspace: findings from a multisite definitive randomized controlled trial of an unguided digital mindfulness-based self-help app to reduce healthcare worker stress. JMIR mHealth and uHealth, 10(8), e31744. Open source (abstract only)
Passage
or sickness absence (γ=0.09, 95% CI -0.18 to 0.34)
Where
Results, the last item in the list of outcomes introduced by "but not for"
Limits
One trial, self-reported absence over 4.5 months.
Checked
6 Oct 2026
Back to text↩ 1
C21Supported
Across 11 randomised trials, therapist-guided online CBT did not reduce sickness absence in adults with common mental disorders.
Source
Udd-Granat L, Lahti J, Donnelly M, et al. (2023). Internet-delivered cognitive behavioral therapy (iCBT) for common mental disorders and subsequent sickness absence: a systematic review and meta-analysis. Scandinavian Journal of Public Health, 51(1), 137-147. Open source (abstract only)
Passage
iCBT did not appear to be effective in terms of reducing (largely self-assessed) SA in adults with common mental disorders.
Where
Conclusions
Limits
Clinical populations, mostly not employer schemes; absence mostly self-reported.
Checked
6 Oct 2026
Back to text↩ 1
C22Supported
Workplace insomnia programmes reduced insomnia symptoms across 12 studies with 827 workers, with evidence the authors rated moderate.
Source
Vega-Escaño J, Porcel-Gálvez AM, Diego-Cordero R, et al. (2020). Insomnia interventions in the workplace: a systematic review and meta-analysis. International Journal of Environmental Research and Public Health, 17(17), 6401. Open source (abstract only)
Passage
a moderate effect for the reduction of insomnia symptoms after the intervention (MD -2.08, CI 95%: [-2.68, -1.47]) and a non-significant degree of heterogeneity were obtained
Limits
Insomnia symptoms only. Effects on depression and anxiety are less certain, and most workplace trials found no change in sickness absence.
Checked
6 Oct 2026
Back to text↩ 1
C23Supported
Participation in workplace physical activity and nutrition programmes ranged from 10% to 64%, with a median of 33%.
Source
Robroek SJW, van Lenthe FJ, van Empelen P, Burdorf A (2009). Determinants of participation in worksite health promotion programmes: a systematic review. International Journal of Behavioral Nutrition and Physical Activity, 6, 26. Open source (abstract only)
Passage
Participation levels varied from 10% to 64%, with a median of 33% (95% CI 25-42%).
Limits
23 studies published from 1988 to 2007, mostly outside the UK.
Checked
6 Oct 2026
Back to text↩ 1
C24Supported
In an Australian public sector trial, staff given only a mindfulness app used it for 45 of 343 recommended minutes.
Source
Bartlett L, Martin AJ, Kilpatrick M, et al. (2022). Effects of a mindfulness app on employee stress in an Australian public sector workforce: randomized controlled trial. JMIR mHealth and uHealth, 10(2), e30272. Open source (abstract only)
Passage
App use time was considerably lower than recommended (app+classes: 120/343 minutes; app-only: 45/343 minutes).
Limits
211 participants, one public sector workforce.
Checked
6 Oct 2026
Back to text↩ 1
C25Supported
In that trial, self-guided app use alone produced no significant results, while the app with classes did.
Source
Bartlett L, Martin AJ, Kilpatrick M, et al. (2022). Effects of a mindfulness app on employee stress in an Australian public sector workforce: randomized controlled trial. JMIR mHealth and uHealth, 10(2), e30272. Open source (abstract only)
Passage
Including classes in the training protocol appears to have motivated engagement and led to benefits, whereas self-guided app use did not realize any significant results.
Where
Conclusions
Limits
As C24. Neither active group changed perceived stress, the primary outcome, at three months.
Checked
6 Oct 2026
Back to text↩ 1
C26Supported
In a 2020 survey of the workforce in Great Britain, half of employees (51%) had access to occupational health services in their current role.
Source
Moore G, Ezzard O (2023). Employee research phase 1: sickness absence, reasonable adjustments and occupational health. Quantitative research. Department for Work and Pensions research report 1021. Open source (full text read)
Passage
Half of employees (51%) had access to occupational health services in their current role
Where
Chapter 5, OH access amongst all workers
Limits
Self-reported, fieldwork 24 September to 8 October 2020, during the pandemic. Great Britain, not the UK. Main sample of 1,950 workers.
Checked
6 Oct 2026
Back to text↩ 1
C27Supported
In the same survey, 20% of employees did not know whether they had access to occupational health.
Source
Moore G, Ezzard O (2023). Employee research phase 1: sickness absence, reasonable adjustments and occupational health. Quantitative research. Department for Work and Pensions research report 1021. Open source (full text read)
Passage
A slightly higher proportion of employees reported not knowing whether they had access to occupational health services than the self-employed (20% vs 17%).
Where
Chapter 5, OH access amongst all workers
Limits
As C26.
Checked
6 Oct 2026
Back to text↩ 1
C28Supported
In the same survey, 70% of employees with occupational health access through their employer had not used it in the past 12 months.
Source
Moore G, Ezzard O (2023). Employee research phase 1: sickness absence, reasonable adjustments and occupational health. Quantitative research. Department for Work and Pensions research report 1021. Open source (full text read)
Passage
Of those employees that had access to OH services through their employer, 70% had not used OH services in the past 12 months.
Where
Chapter 5, Use of OH services and outcomes by employees
Limits
As C26. Not using a service in a year can mean not needing it.
Checked
6 Oct 2026
Back to text↩ 1
C29Supported
A Cochrane review of Mental Health First Aid training could draw no conclusions about its effects on the mental health of people in the communities where it was delivered.
Source
Richardson R, Dale HE, Robertson L, et al. (2023). Mental Health First Aid as a tool for improving mental health and well-being. Cochrane Database of Systematic Reviews, CD013127. Open source (abstract only)
Passage
We cannot draw conclusions about the effects of MHFA training on our primary outcomes due to the lack of good quality evidence.
Where
Authors' conclusions
Limits
21 trials, not workplace specific. The review found no published data on adverse effects. The training's effect on knowledge is a separate question.
Checked
6 Oct 2026
Back to text↩ 1
C30Supported
A review of eight workplace trials found that mental health screening alone does not improve worker mental health.
Source
Strudwick J, Gayed A, Deady M, et al. (2023). Workplace mental health screening: a systematic review and meta-analysis. Occupational and Environmental Medicine, 80(8), 469-484. Open source (abstract only)
Passage
The evidence supporting workplace mental health screening programmes is limited and available data suggest mental health screening alone does not improve worker mental health.
Limits
2,940 employees in 8 trials, certainty low to very low. Screening followed by facilitated access to treatment showed a small improvement.
Checked
6 Oct 2026
Back to text↩ 1

Back to the note320 questions about employee benefits