Education, incentive, and engineering-based interventions to promote the use of seat belts

Abstract

Background: Over 1.3 million people die each year as a result of traffic collisions and hundreds of thousands of others are permanently and seriously injured. Most of these deaths occur in low- and middle-income countries, where mortality rates can be up to 10 times higher than those of some high-income countries. Seat belts are designed to accomplish two key functions – to prevent the occupant from being ejected from the vehicle by the force of impact, and to extend the time that the decelerating force is applied to a person. Seat belts also spread the area of impact both to larger and less vulnerable parts of the body. Since the 1950s, seat belts have been factory-fitted to most vehicles, and today around 90% of high-income countries have adopted seat belt legislation that makes it mandatory for some, if not all, vehicle occupants to wear seat belts. However, the simple passing of laws is not sufficient to ensure seat belt use, and, while the enforcement of seat belt laws does increase seat belt use, other interventions have been developed to encourage voluntary – and hence sustainable – behaviour change. Objectives: To evaluate the benefits of behavioural-change interventions (educational-based, incentive-based, engineering-based, or a combination, but not enforcement-based) that promote the use of seat belts, and to determine which types of interventions are most effective. Search methods: On 9 August 2022, we searched the Cochrane Central Register of Controlled Trials (CENTRAL), OvidSP Embase, OvidSP MEDLINE, 14 other databases, and clinical trials registers. We also screened reference lists and conference proceedings, searched websites of relevant organisations, and contacted road safety experts. The search was performed with no restrictions in terms of language and date of publication. Selection criteria: We included randomised controlled trials (RCTs), both individually randomised and cluster-randomised, that evaluated education, engineering, incentive-based interventions (or combinations) that promoted seat belt use. Data collection and analysis: Two review authors independently assessed the eligibility of RCTs, evaluated the risk of bias, and extracted data. We performed a narrative synthesis based on effect direction due to the heterogeneity observed between RCTs and reported the synthesis in accordance with reporting guidelines for systematic reviews without meta-analysis, as appropriate. We assessed the certainty of the evidence using the GRADE approach. We analysed data on the primary outcome, frequency of wearing a seat belt. None of the included RCTs reported the other primary outcome, crash-related injury rate or the secondary outcome, crash-related mortality rates of interest in this review. Main results: We included 15 completed RCTs (12 individual, parallel-group, and three cluster) that enroled 12,081 participants, published between 1990 and 2022. Four trials were published between 2019 and 2022, and the remaining trials were published 10 or more years ago. We also identified four ongoing RCTs. Thirteen RCTs were conducted in the USA. Trials recruited participants from various sites (worksites, schools, emergency departments, a residential retirement community, and primary care settings) and different age groups (adults, late adolescents, early adolescents, and dyads). Thirteen trials investigated educational interventions, one of which used education in addition to incentives (one of the intervention arms) measured through participant self-reports (12) and observation (one), and two trials investigated engineering-based interventions measured through in-vehicle data monitor systems at various follow-up periods (six weeks to 36 months). We grouped RCTs according to types of education-based interventions: behavioural education-based, health risk appraisal (HRA), and other education-based interventions. The evidence suggests that behavioural education-based (four trials) interventions may promote seat belt use and HRA interventions (one trial) likely promote seat belt use in the short term (six weeks to nine months). Four of the six trials that investigated behavioural education-based interventions found that the intervention compared to no or another intervention may promote seat belt use. These effects were measured through participant self-report and at various time points (six-week to 12-month follow-up) (low-certainty evidence). One of the three trials investigating HRA only or with additional intervention versus no or another intervention showed observed effects likely to promote seat belt use (moderate-certainty evidence). The evidence suggests that engineering-based interventions using vehicle monitoring systems (with in-vehicle alerts and with or without notifications/feedback) may promote the use of seat belts. One trial showed that engineering interventions (in-vehicle alerts and feedback) may promote seat belt use while the other showed unclear effects in two of the three intervention groups (low-certainty evidence). Both trials had small sample sizes and high baseline seat belt use. Authors' conclusions: The evidence suggests that behavioural education-based interventions may promote seat belt use and HRA (including incentives) with or without additional interventions likely promote seat belt use. Likewise, for engineering-based interventions using in-vehicle data monitor systems with in-vehicle alerts, with or without notifications/feedback the evidence suggests the interventions may promote the use of seat belts. Well-designed RCTs are needed to further investigate the effectiveness of education and engineering-based interventions. High-quality trials that examine the potential benefits of incentives to promote seat belt use, either alone or in combination with other interventions, as well as trials to investigate other types of interventions (such as technology, media/publicity, enforcement, insurance schemes, employer programmes, etc.) to promote the use of seat belts, are needed. Evidence from low- and middle-income economies is required to improve the generalisability of the data. In addition, research focused on determining which interventions or types of interventions are most effective in different population groups is needed.