Knowledge, Attitudes and Practices Regarding Physical Restraint Among ICU Nurses in China: A Mixed-Methods Systematic...

Physical restraint is still common in intensive care units (ICUs) despite well-documented adverse effects. ICU nurses' knowledge, attitudes and practices (KAP) directly affect how restraints are used.

To synthesize the evidence on physical restraint related KAP among ICU nurses in China and to identify gaps between knowledge and practice.

We searched 11 databases (CNKI, WanFang, CBM, VIP, Chaoxing, PubMed, Cochrane Library, Web of Science, MEDLINE via Ovid, CINAHL via EBSCO, Scopus) from inception to October 2025. The protocol was registered in PROSPERO (CRD420251154455). Methodological quality was assessed using Joanna Briggs Institute tools. We used a convergent segregated approach to integrate quantitative and qualitative findings.

Twenty studies were included (13 quantitative, 6 qualitative, 1 mixed methods). Knowledge scores ranged from 50.43% to 93.73% (median 63.37%), attitude scores from 38.92% to 70.98% (median 64.28%), and practice scores from 50.51% to 94.36% (median 73.64%). Systematic training, years of ICU experience, and critical care certification were significant influencing factors. The qualitative synthesis produced three themes: a knowledge-practice gap, attitudinal conflict between safety principles and emotional burden and a disconnect between standardized protocols and real-world practice.

The linear KAP model does not fully explain restraint practices in Chinese ICUs. Environmental and systemic factors such as staffing levels, workflow design and safety metrics mediate the relationship between knowledge and practice. Improving practice calls for unit-level interventions (a two-tiered assessment process integrated into workflow, daily multidisciplinary review) and nurse-level training (alternatives, ethical decision making). Future research should use validated instruments, include secondary hospitals and test interventions with controlled designs.

This study is a systematic review of previously published literature and did not involve direct participation of patients, service users, caregivers or members of the public in its design, conduct, analysis, interpretation or manuscript preparation.