Individual Placement and Support and Participatory Workplace Intervention on the Work Participation of People with Disabilities: A Randomised Controlled Trial.

The study was likely underpowered to detect the smaller effect sizes actually observed, as the control group had a higher rate of employment than expected. Implementation fidelity was sub-optimal; many PWI participants did not reach the stage where the workplace intervention could be delivered, and IPS integration with healthcare services was limited. Job coaches were not randomised, which may have introduced bias if coaches self-selected into interventions they were more motivated to deliver. The population included people with mixed physical, psychological, and social disabilities, which may make results less applicable to specific diagnostic groups like severe mental illness alone. Participants who could not understand the informed consent due to cognitive or language barriers were excluded, limiting generalisability to the most severely affected individuals.

The authors declared no conflicts of interest. The study was conducted within a large municipality in the Netherlands and approved by the Medical Ethics Committee of the VU University Medical Centre.

Do not conclude that IPS is ineffective for this population. The null primary result may be due to insufficient statistical power and sub-optimal implementation rather than a lack of therapeutic effect, especially given the significant positive finding on the secondary outcome of starting any work or education.