Background Post-intensive care syndrome (PICS), encompassing physical, psychological, and cognitive impairments, significantly compromises recovery in critical illness survivors. Although occupational therapy (OT) may mitigate functional decline, its efficacy in the intensive care unit (ICU) remains inadequately established. This systematic review and meta-analysis evaluated the impact of OT on clinical outcomes in critically ill adults. Methods We searched PubMed, Embase, Web of Science, CINAHL, Cochrane Library, CNKI, Wanfang, and China Biomedical Literature Service System from inception to August 5, 2025 for randomized controlled trials (RCTs) assessing ICU-based OT. Primary outcomes included activities of daily living (ADL), delirium incidence, grip strength, and mechanical ventilation duration. Data were synthesized using RevMan 5.3, with continuous outcomes expressed as mean difference (MD) or standardized mean difference (SMD) and dichotomous outcomes as relative risk (RR), all with 95% confidence intervals (CIs). The certainty of evidence for each outcome was assessed using the GRADE approach. Results Pooled analyses demonstrated that OT significantly enhanced ADL performance SMD = 0.72, 95% CI (0.40, 1.05), p 0.001, reduced the incidence of delirium RR = 0.44, 95% CI (0.30, 0.63), p 0.001, increased grip strength MD = 3.90 kg, 95% CI (2.03, 5.76), p 0.001, and shortened the duration of mechanical ventilation SMD = −0.68, 95% CI (−0.99, −0.37), p 0.001. The certainty of evidence (GRADE) was low for ADL, and moderate for delirium, grip strength, and mechanical ventilation duration. Subgroup analysis of stroke patients further demonstrated that OT significantly improved ADL performance SMD = 0.81, 95% CI (0.42, 1.20), p 0.001 and reduced delirium incidence RR = 0.39, 95% CI (0.21, 0.72), p = 0.003, suggesting a particularly beneficial effect in this vulnerable population. Conclusion This meta-analysis provides evidence that OT may enhance functional recovery, prevent delirium, and facilitate weaning in critically ill patients. However, the strength of the evidence is low to moderate, tempered by the limited number of trials, risk of bias, and observed heterogeneity. These findings underscore the need for further rigorous investigation to establish optimal protocols. Systematic review registration https://www.crd.york.ac.uk/PROSPERO/recorddashboard , Unique identifier: CRD42025312345.
Hua et al. (Thu,) studied this question.