What the Evidence Says About Supervisor Training and RTW

The Uncomfortable Finding

Supervisor training is one of the most recommended interventions in disability management. Frameworks mention it. Guidelines call for it. Organisations that invest in RTW processes tend to arrive at the same conclusion: supervisors need to be trained.

The evidence agrees. And it disagrees. Simultaneously.

A meta-analysis by Gayed et al. (2018) pooled ten controlled trials of training managers in workplace mental health. Trained managers knew more about mental health at work, held less stigmatising attitudes, and reported more supportive behaviour towards employees with mental health problems.

What it could not show was any effect on the employees. Return to work was not among the outcomes. Too few trials measured sickness absence to pool them. The one employee outcome it could pool, psychological distress, showed no significant effect, and the authors call the employee findings preliminary.

This is not evidence that training fails. It is a clue to how training works.

Where Training Has Moved Outcomes

The studies lean one way, though not cleanly. The one study here of supervisor training across a whole employer evaluated it built into a broader system. Training delivered on its own has mostly improved what supervisors know and report, with one small exception.

Embedded in a System

Mustard et al. (2017) studied a large hospital that implemented a strengthened disability management policy including early contact protocols, supervisor training, and union integration. Mean duration of work-injury disability episodes fell from 19.4 days to 10.9 days. A 44% reduction. Durations fell across the sector over the same years: in 29 comparison hospitals, from 13.5 to 10.5 days. The policy was associated with the larger drop.

Macpherson et al. (2022) found a similar pattern at the level of a whole compensation system. Ontario’s Workplace Safety and Insurance Board introduced an integrated RTW and vocational rehabilitation programme. Among the longest construction claims referred to RTW specialists, it was associated with 274 fewer paid disability days at the 90th percentile. The paper does not mention supervisor training. What changed was the system around the claim.

Mustard’s study cannot separate the training’s share from everything else that changed. What was evaluated was the package: training as one part of a policy that also included early contact, union representatives in RTW planning, and RTW coordinators.

Standalone Training: Mostly Knowledge, Rarely Outcomes

Only one controlled test of training on its own has found fewer claims, and it is small. Shaw et al. (2006) gave 23 supervisors in one food-processing plant a four-hour workshop on communication and ergonomic accommodation: 11 straight away, 12 later as the control group. In the seven months after, new workers’ compensation claims fell 47% and active lost-time claims 18%, against 27% and 7% in the control group. It is an encouraging result from one plant, and among the studies here it is the exception.

Schaap et al. (2024) evaluated Mentorwijs, five weekly sessions for supervisors of employees with a work disability, delivered as a training on its own. They found a small positive effect on employment at 8 months but non-significant results at 3, 6, 9 and 12 months. Supervisors gained knowledge and self-efficacy, but their intention to apply it and what they reported doing did not change. Employee employment outcomes did not consistently change.

Vonderlin et al. (2023) studied a three-day mindfulness-based leadership programme for supervisors. Afterwards, the 13 trained supervisors had fewer sick days than matched controls (13.9 days vs. 32.9, d = 0.47), though the authors found this was driven mainly by sick days rising in the control group. The study also tracked their 196 employees: their sick days did not differ from controls’. No effect showed up among the employees.

Van Vilsteren et al. (2017) tested something broader than training: integrated care plus a participatory workplace intervention for workers with rheumatoid arthritis, in a randomised trial. After six months, workers in the programme rated their supervisors’ support slightly higher than the control group did, a small difference the authors treat with caution. Work instability and at-work productivity did not change. Adding parts around the supervisor does not guarantee outcomes either.

The evidence does not say supervisor training is ineffective. It says supervisor training on its own has rarely been shown to change RTW outcomes.

What Training Content Shows Promise

Not all training content is the same. Four content elements recur across the programmes studied here.

1. Participatory Problem-Solving

In the participatory approach, supervisor and worker each name the problems at work, then solve them together, rather than the supervisor applying a set plan. Earlier trials, in which an occupational health professional led the process for workers already on sick leave, found it improved return to work, as Ketelaar et al. (2017) summarise. Ketelaar and colleagues then trained supervisors to use it themselves, in a cluster RCT. It did not raise supervisors’ self-efficacy overall, but it did increase the number of employees they discussed work problems with.

The likely mechanism: participatory problem-solving puts the supervisor in the position of listening to what the worker actually needs, rather than applying a generic accommodation template.

2. Communication Skills for RTW Conversations

Shaw et al. (2006) and McLellan, Pransky, and Shaw (2001) both trained supervisors in how to respond when a worker reports pain or injury. Shaw’s controlled study found claims fell. McLellan’s pilot, with no control group, found supervisors more confident a month later; a year on, more of the supervisors who replied said lost work time had gone down than up. The content in both was practical. It was about how to have the conversation.

For musculoskeletal and other pain-related conditions, early contact with the worker by the workplace is one of the components with moderate evidence of shortening work disability (Franche et al., 2005); the strongest evidence in that review was for offers of accommodation and contact between the healthcare provider and the workplace. But early contact can only help if the supervisor knows what to say. Training that does not address communication skills misses the most likely route by which supervisors influence RTW.

3. Mental Health Literacy

In workplace mental health, training managers improves their knowledge, their attitudes and the supportive behaviour they report (Gayed et al., 2018). In a cluster randomised trial of HeadCoach, an online mental health training for managers, confidence rose straight after the course, and the managers’ reported responsive and preventive behaviours improved and held at four months (Gayed et al., 2019).

What it does for employees is not yet known. The review found too few studies of sickness absence to pool, and HeadCoach did not change the distress levels of the managers’ staff. Neither measured return to work. The reading here: mental health literacy is necessary but not sufficient. Organisational alignment and accommodation infrastructure must also be in place.

4. Access to Occupational Health Expertise

Van Hees et al. (2024) evaluated a workplace programme, run by occupational health professionals, to help supervisors support employees with common mental health problems. Supervisors’ self-reported supportive behaviour rose after it and was significantly higher at six months. In interviews and surveys, what made it work traced back to the professionals’ expertise and the supervisors’ trust in them, time set aside, and structural, low-key access to that expertise. The study had no training-only comparison, so it cannot show that ongoing access beats training alone. It shows what the supervisors’ experience pointed to.

The implication: training should not be a one-time event followed by unsupported practice. It should be connected to ongoing professional resources.

The Duration and Format Question

The evidence does not point to a single best training duration or format. The principle argued here: training must include active practice, not just information transfer.

• A four-hour workshop on communication and accommodation was followed by larger falls in claims than in a later-trained control group, in one plant (Shaw et al., 2006)

• Online modules raised managers’ confidence straight after the course and improved the supportive behaviours they reported, with no change in their staff’s distress (HeadCoach; Gayed et al., 2019)

• A three-day mindfulness-based leadership programme was followed by fewer supervisor sick days than in matched controls, mostly because the controls’ rose, and no difference in their employees’ sick days (Vonderlin et al., 2023)

The differentiator argued for here is not length. It is whether the training includes problem-solving practice with real or realistic scenarios and whether it is connected to an operational system that reinforces the trained behaviours.

What This Means for Your Organisation

If your organisation provides supervisor training on RTW accommodation, it is doing something the evidence supports. If that training is delivered without early contact protocols, accommodation planning tools, RTW coordination, and psychosocial assessment, the likely result is better knowledge without changed outcomes.

The question is not whether to train supervisors. The argument of this piece is that training pays off when the system supports what supervisors learned.

The question is whether you have the system.

References

Franche RL, Cullen K, Clarke J, et al. Workplace-based return-to-work interventions: a systematic review of the quantitative literature. Journal of Occupational Rehabilitation. 2005;15(4):607-631. doi:10.1007/s10926-005-8038-8

Gayed A, Bryan BT, LaMontagne AD, et al. A cluster randomized controlled trial to evaluate HeadCoach: an online mental health training program for workplace managers. Journal of Occupational and Environmental Medicine. 2019;61(7):545-551. doi:10.1097/jom.0000000000001597

Gayed A, Milligan-Saville JS, Nicholas J, et al. Effectiveness of training workplace managers to understand and support the mental health needs of employees: a systematic review and meta-analysis. Occupational and Environmental Medicine. 2018;75(6):462-470. doi:10.1136/oemed-2017-104789

Ketelaar SM, Schaafsma FG, Geldof MF, et al. Implementation of the participatory approach for supervisors to increase self-efficacy in addressing risk of sick leave of employees: results of a cluster-randomized controlled trial. Journal of Occupational Rehabilitation. 2017;27(2):247-257. doi:10.1007/s10926-016-9652-3

Macpherson RA, He A, Amick BC 3rd, et al. Evaluating effectiveness of an integrated return-to-work and vocational rehabilitation program on work disability duration in the construction sector. Scandinavian Journal of Work, Environment & Health. 2022;48(3):229-238. doi:10.5271/sjweh.4006

McLellan RK, Pransky G, Shaw WS. Disability management training for supervisors: a pilot intervention program. Journal of Occupational Rehabilitation. 2001;11(1):33-41. doi:10.1023/a:1016652124410

Mustard CA, Skivington K, Lay M, et al. Implementation of a disability management policy in a large healthcare employer: a quasi-experimental, mixed-methods evaluation. BMJ Open. 2017;7(6):e014734. doi:10.1136/bmjopen-2016-014734

Schaap R, Coenen P, Zwinkels W, et al. Training for supervisors to improve sustainable employment of employees with a work disability: a longitudinal effect and process evaluation from an intervention study with matched controls. Journal of Occupational Rehabilitation. 2024;34(1):180-196. doi:10.1007/s10926-023-10118-2

Shaw WS, Robertson MM, McLellan RK, et al. A controlled case study of supervisor training to optimize response to injury in the food processing industry. Work. 2006;26(2):107-114. doi:10.3233/wor-2006-00498

Van Hees S, Carlier B, Joosen M, et al. A workplace intervention to strengthen supervisor support for employees with common mental health problems: a mixed-methods realist evaluation. Scandinavian Journal of Work and Organizational Psychology. 2024;9(1):9. doi:10.16993/sjwop.249

van Vilsteren M, Boot CRL, Twisk JWR, et al. Effectiveness of an integrated care intervention on supervisor support and work functioning of workers with rheumatoid arthritis. Disability and Rehabilitation. 2017;39(4):354-362. doi:10.3109/09638288.2016.1145257

Vonderlin R, Schmidt B, Biermann M, et al. Improving health and reducing absence days at work: effects of a mindfulness- and skill-based leadership intervention on supervisor and employee sick days. Mindfulness. 2023;14(7):1751-1766. doi:10.1007/s12671-023-02172-x