Drivers of sustainable biomedical waste management practices among healthcare workers in Bangladesh: a cross-sectional structural equation modelling study

Abstract Background Sustainable biomedical waste management protects healthcare workers, patients, communities, and the environment. Appropriate practice depends not only on knowledge and attitudes but also on behavioural intention, perceived control, organisational support, and regulatory awareness. This study examined the drivers of sustainable biomedical waste-management practices among healthcare workers in Bangladesh using an extended theory of planned behaviour framework. Methods A cross-sectional, facility-based survey was conducted in Dhaka, Bangladesh, between May and September 2026 among 420 healthcare workers from 5 public hospitals, 10 private hospitals, and 13 specialised or diagnostic facilities. A questionnaire with five-point Likert-scale items measured six antecedent constructs, behavioural intention, and sustainable practice. Partial least squares structural equation modelling (SmartPLS 4.1.2.1) was used, with bootstrapping, PLSpredict, measurement invariance of composites testing, and permutation-based multigroup analysis. Results The model explained 41.4% of the variance in behavioural intention and 46.8% in sustainable practice. All six antecedents significantly predicted intention; perceived behavioural control was strongest (β = 0.277, p < 0.001), followed by training and organisational support (β = 0.227), regulatory awareness (β = 0.219), knowledge (β = 0.204), attitude (β = 0.159), and subjective norms (β = 0.087, p = 0.021). Behavioural intention was the strongest predictor of practice (β = 0.465, p < 0.001), followed by perceived behavioural control (β = 0.247, p < 0.001) and training and organisational support (β = 0.118, p = 0.005). Regulatory awareness had no significant direct effect on practice (β = 0.062, p = 0.110) but a significant indirect effect through intention. PLSpredict indicated meaningful out-of-sample predictive performance. No significant structural-path differences were detected across facility types. Conclusions Sustainable practice appears to be shaped by both motivational and enabling factors. Behavioural intention was the strongest predictor of practice, while perceived behavioural control and organisational support had important direct and indirect roles. Interventions should move beyond knowledge-based training and strengthen workers’ ability to act through supportive organisational systems, adequate resources, and effective implementation of regulatory requirements.