Background: Respiratory hazards are common in the workplace. Depending on the hazard and exposure, the health consequences may include: mild to life-threatening illnesses from infectious agents, acute effects ranging from respiratory irritation to chronic lung conditions, or even cancer from exposure to chemicals or toxins. Use of respiratory protective equipment (RPE) is an important preventive measure in many occupational settings. RPE only offers protection when worn properly, when removed safely and when it is either replaced or maintained regularly. The effectiveness of behavioural interventions either directed at employers or organisations or directed at individual workers to promote RPE use in workers remains an important unanswered question.
Objectives: To assess the effects of any behavioural intervention either directed at organisations or at individual workers on observed or self-reported RPE use in workers when compared to no intervention or an alternative intervention.
Search methods: We searched the Cochrane Work Group Specialised Register, the Cochrane Central Register of Controlled Trials (CENTRAL 2016, Issue 07), MEDLINE (1980 to 12 August 2016), EMBASE (1980 to 20 August 2016) and CINAHL (1980 to 12 August 2016).
Selection criteria: We included randomised controlled trials (RCTs), controlled before and after (CBA) studies and interrupted time-series (ITS) comparing behavioural interventions versus no intervention or any other behavioural intervention to promote RPE use in workers.
Data collection and analysis: Four authors independently selected relevant studies, assessed risk of bias and extracted data. We contacted investigators to clarify information. We pooled outcome data from included studies where the studies were sufficiently similar.
Main results: We included 14 studies that evaluated the effect of training and education on RPE use, which involved 2052 participants. The included studies had been conducted with farm, healthcare, production line, office and coke oven workers as well as nursing students and people with mixed occupations. All included studies reported the effects of interventions as use of RPE, as correct use of RPE or as indirect measures of RPE use. We did not find any studies where the intervention was delivered and assessed at the whole organization level or in which the main focus was on positive or negative incentives. We rated the quality of the evidence for all comparisons as low to very low. Training versus no trainingOne CBA study in healthcare workers compared training with and without a fit test to no intervention. The study found that the rate of properly fitting respirators was not considerably different in the workers who had received training with a fit test (RR 1.17, 95% Confidence Interval (CI) 0.97 to 1.10) or training without a fit test (RR 1.16, 95% CI 0.95 to 1.42) compared to those who had no training. Two RCTs that evaluated training did not contribute to the analyses because of lack of data. Conventional training plus additions versus conventional training aloneOne cluster-randomised trial compared conventional training plus RPE demonstration versus training alone and reported no significant difference in appropriate use of RPE between the two groups (RR 1.41, 95% CI 0.96 to 2.07).One RCT compared interactive training with passive training, with an information screen, and an information book. The mean RPE performance score for the active group was not different from that of the passive group (MD 2.10, 95% CI -0.76 to 4.96). However, the active group scored significantly higher than the book group (MD 4.20, 95% CI 0.89 to 7.51) and the screen group (MD 7.00, 95% CI 4.06 to 9.94).One RCT compared computer-simulation training with conventional personal protective equipment (PPE) training but reported only results for donning and doffing full-body PPE. Education versus no educationOne RCT found that a multifaceted educational intervention increased the use of RPE (risk ratio (RR) 1.69, 95% CI 1.10 to 2.58) at three years' follow-up when compared to no intervention. However, there was no difference between intervention and control at one year's, two years' or four years' follow-up. Two RCTs did not report enough data to be included in the analysis.Four CBA studies evaluated the effectiveness of education interventions and found no effect on the frequency or correctness of RPE use, except in one study for the use of an N95 mask (RR 4.56, 95% CI 1.84 to 11.33, 1 CBA) in workers. Motivational interviewing versus traditional lecturesOne CBA study found that participants given motivational group interviewing-based safety education scored higher on a checklist measuring PPE use (MD 2.95, 95% CI 1.93 to 3.97) than control workers given traditional educational sessions.
Authors' conclusions: There is very low quality evidence that behavioural interventions, namely education and training, do not have a considerable effect on the frequency or correctness of RPE use in workers. There were no studies on incentives or organisation level interventions. The included studies had methodological limitations and we therefore need further large RCTs with clearer methodology in terms of randomised sequence generation, allocation concealment and assessor blinding, in order to evaluate the effectiveness of behavioural interventions for improving the use of RPE at both organisational and individual levels. In addition, further studies should consider some of the barriers to the successful use of RPE, such as experience of health risk, types of RPE and the employer's attitude to RPE use.