Self-harm in adolescence and risk of crash: a 13-year cohort study of novice drivers in New South Wales, Australia.

INTRODUCTION: Self-harm and suicide are leading causes of morbidity and death for young people, worldwide. Previous research has identified self-harm is a risk factor for vehicle crashes, however, there is a lack of long-term crash data post licensing that investigates this relationship. We aimed to determine whether adolescent self-harm persists as crash risk factor in adulthood. METHODS: We followed 20 806 newly licensed adolescent and young adult drivers in the DRIVE prospective cohort for 13 years to examine whether self-harm was a risk factor for vehicle crashes. The association between self-harm and crash was analysed using cumulative incidence curves investigating time to first crash and quantified using negative binominal regression models adjusted for driver demographics and conventional crash risk factors. RESULTS: Adolescents who reported self-harm at baseline were at increased risk of crashes 13 years later than those reporting no self-harm (relative risk (RR) 1.29: 95% CI 1.14 to 1.47). This risk remained after controlling for driver experience, demographic characteristics and known risk factors for crashes, including alcohol use and risk taking behaviour (RR 1.23: 95% CI 1.08 to 1.39). Sensation seeking had an additive effect on the association between self-harm and single-vehicle crashes (relative excess risk due to interaction 0.87: 95% CI 0.07 to 1.67), but not for other types of crashes. DISCUSSION: Our findings add to the growing body of evidence that self-harm during adolescence predicts a range of poorer health outcomes, including motor vehicle crash risks that warrant further investigation and consideration in road safety interventions. Complex interventions addressing self-harm in adolescence, as well as road safety and substance use, are critical for preventing health harming behaviours across the life course.

Acculturation and risk of traffic crashes in young Asian-born Australian drivers.

The study examines changes over time in crash risk differences between young Australian drivers born in Asia and those born in Australia.Data from the 2003 baseline survey of the DRIVE cohort of 20 806 young drivers aged 17-24 years were linked to police, hospital and death data up until 2016. The association between country of birth and crash was investigated using flexible parametric survival models adjusted for confounders.Six months after baseline, the crash risk in Asian-born drivers was less than half that of their Australian-born counterparts (mean HR, MHR 0.41; 95% CI 0.29 to 0.57), only to increase steadily over time to resemble that of Australian-born drivers 13 years later (MHR 0.94; 95% CI 0.66 to 1.36).This is likely to be associated with acculturation and the adoption by young Asian-born Australian drivers of driving behaviour patterns akin to those born locally. This needs to be considered in future road safety campaigns.

Evidence for the 'safety in density' effect for cyclists: validation of agent-based modelling results.

The safety in numbers (SiN) effect for cyclists is widely observed but remains poorly understood. Although most studies investigating the SiN phenomenon have focused on behavioural adaptation to 'numbers' of cyclists in the road network, previous work in simulated environments has suggested SiN may instead be driven by increases in local cyclist spatial density, which prevents drivers from attempting to move through groups of oncoming cyclists. This study therefore set out to validate the results of prior simulation studies in a real-world environment. Time-gap analysis of cyclists passing through an intersection was conducted using 5 hours of video observation of a single intersection in the city of Melbourne, Australia, where motorists were required to 'yield' to oncoming cyclists. Results demonstrated that potential collisions between motor vehicles and cyclists reduced with increasing cyclists per minute passing through the intersection. These results successfully validate those observed under simulated conditions, supporting evidence of a proposed causal mechanism related to safety in density rather than SiN, per se. Implications of these results for transportation planners, cyclists and transportation safety researchers are discussed, suggesting that increased cyclist safety could be achieved through directing cyclists towards focused, strategic corridors rather than dispersed across a network.

Crash characteristics of on-road single-bicycle crashes: an under-recognised problem.

Compared with crashes with motor vehicles, single-bicycle crashes are an under-recognised contributor to cycling injury and the aetiology is poorly understood. Using an in-depth crash investigation technique, this study describes the crash characteristics and patient outcomes of a sample of cyclists admitted to hospital following on-road bicycle crashes. Enrolled cyclists completed a structured interview, and injury details and patient outcomes were extracted from trauma registries. Single-bicycle crashes (n=62) accounted for 48% of on-road crashes and commonly involved experienced cyclists. Common single-bicycle crash types included loss-of-control events, interactions with tram tracks, striking potholes or objects or resulting from mechanical issues with the bicycle. To address single-bicycle crashes, targeted countermeasures are required for each of these specific crash types.

Distracting and risky behaviours while cycling: a comparison of group and non-group riders in Western Australia.

BACKGROUND: Use of mobile phones and portable audio equipment and alcohol are known to negatively affect cycling ability. Evidence suggests that cyclists may be less likely to engage in these behaviours while riding in a group; however, it is unknown whether group riders are also at reduced risk when participating in non-group riding. OBJECTIVE: To examine the association between group riding participation and the use of mobile phones and portable audio equipment and alcohol while non-group riding in Perth, Western Australia. METHODS: A cross-sectional analysis of an online questionnaire was undertaken. Group and exclusive non-group riders were compared and separate binary logistic regression models were used to examine the association between group riding participation and the use of mobile phones and portable audio equipment and alcohol while non-group riding, controlling for gender, age, education and frequency of non-group riding. RESULTS: Participants included 365 cyclists: 187 exclusive non-group riders (51.2%) and 178 group riders (48.8%). Group riders were less likely to have possibly cycled while over the legal blood alcohol limit in the past 12 months (OR: 0.56, 95% CI 0.34 to 0.92) and were less likely to ever use portable audio equipment (OR: 0.57, 95% CI 0.34 to 0.94) than exclusive non-group riders, while participating in non-group riding. Group riding status was not associated with mobile phone use. CONCLUSIONS: This study provides early evidence that there may be differences between group and non-group riders that impact on their safety behaviours while participating in non-group riding.

The role of sleepiness, sleep disorders, and the work environment on heavy-vehicle crashes in 2 Australian states.

Heavy-vehicle driving involves a challenging work environment and a high crash rate. We investigated the associations of sleepiness, sleep disorders, and work environment (including truck characteristics) with the risk of crashing between 2008 and 2011 in the Australian states of New South Wales and Western Australia. We conducted a case-control study of 530 heavy-vehicle drivers who had recently crashed and 517 heavy-vehicle drivers who had not. Drivers' crash histories, truck details, driving schedules, payment rates, sleep patterns, and measures of health were collected. Subjects wore a nasal flow monitor for 1 night to assess for obstructive sleep apnea. Driving schedules that included the period between midnight and 5:59 am were associated with increased likelihood of crashing (odds ratio = 3.42, 95% confidence interval: 2.04, 5.74), as were having an empty load (odds ratio = 2.61, 95% confidence interval: 1.72, 3.97) and being a less experienced driver (odds ratio = 3.25, 95% confidence interval: 2.37, 4.46). Not taking regular breaks and the lack of vehicle safety devices were also associated with increased crash risk. Despite the high prevalence of obstructive sleep apnea, it was not associated with the risk of a heavy-vehicle nonfatal, nonsevere crash. Scheduling of driving to avoid midnight-to-dawn driving and the use of more frequent rest breaks are likely to reduce the risk of heavy-vehicle nonfatal, nonsevere crashes by 2-3 times.

Work-related road traffic injury: a multilevel systems protocol.

BACKGROUND: Although road traffic injury is reported as the leading cause of work-related death in Australia, it is not clear, due to limitations in previous methods used, just how large a burden it is. Many organisations are unaware of the extent of work-related road traffic injury and, importantly, what can be done to reduce the burden. The proposed research will (i) estimate the prevalence of work-related road traffic injury and (ii) identify the organisational determinants associated with work-related road traffic injury. METHODS AND DESIGN: The current study is designed to enumerate the problem and identify the individual driver-level, the supervisor-level and organisational-level factors associated with work-related road traffic injury. The multilevel systems protocol will involve a series of cross-sectional surveys administered to drivers of fleet vehicles (n=1200), supervisors of the drivers (n=1200) and senior managers (n=300) within the same organisation. DISCUSSION: The novel use of the multilevel systems protocol is critical to be able to accurately assess the specific determinants of driving safety within each context of an organisation. RESULTS: The results are expected to highlight that reducing injury in the workplace requires more than just individual compliance with safety procedures. It will also establish, for the first time, an occupational translation taskforce to ensure that the research findings are adopted into work-place practice and thereby directly contribute to reductions in work-related road traffic injury.

Evaluation of the Vietnamese A6 mortality reporting system: injury as a cause of death.

BACKGROUND: Despite the fact that the A6 mortality reporting system has been operating for almost 20 years in Vietnam, there has been no systematic evaluation of the system. This study assesses the completeness, sensitivity and positive predictive value of the system in relation to injury related mortality. METHODS: Evaluation of the A6 system was undertaken in three (geographically distributed) provinces in Vietnam. Deaths identified in the A6 system were compared with deaths identified by an independent consensus panel to determine the per cent completeness of the A6 system. Verbal autopsies (VA) were conducted for all identified deaths from the consensus panels, and the sensitivity and positive predictive value of the A6 system were assessed using the VAs as the reference. RESULTS: 5273 deaths were identified from the A6 system with a further 340 cases identified by the independent consensus panel (total n=5613). Injury related deaths accounted for 13.6% (n=763) of all deaths with an overall injury mortality rate of 55.3 per 100 000 person years. The per cent completeness of the A6 system in relation to injury deaths was 93.9% with a sensitivity of 75.4%, specificity of 98.4% and positive predictive value of 88.4%. CONCLUSIONS: The A6 mortality reporting system is embedded within the commune health system and is the lead mortality reporting system for the Ministry of Health. The system performs well in relation to its completeness and classification of injury related deaths. With further enhancements and ongoing support from government and donor agencies, the A6 system will be a valuable resource for identifying and planning preventive strategies targeting the leading causes of injury related deaths in Vietnam.

Compliance with return-to-play regulations following concussion in Australian schoolboy and community rugby union players.

BACKGROUND: There is a risk of concussion when playing rugby union. Appropriate management of concussion includes compliance with the return-to-play regulations of the sports body for reducing the likelihood of premature return-to-play by injured players. PURPOSE: To describe the proportion of rugby union players who comply with the sports body's regulations on returning to play postconcussion. STUDY DESIGN: Prospective cohort study. METHODS: 1958 community rugby union players (aged 15-48 years) in Sydney (Australia) were recruited from schoolboy, grade and suburban competitions and followed over ≥1 playing seasons. Club doctors/physiotherapists/coaches or trained injury recorders who attended the game reported players who sustained a concussion. Concussed players were followed up over a 3-month period and the dates when they returned to play (including either a game or training session) were recorded, as well as any return-to-play advice they received. RESULTS: 187 players sustained ≥1 concussion throughout the follow-up. The median number of days before players returned to play (competition game play or training) following concussion was 3 (range 1-84). Most players (78%) did not receive return-to-play advice postconcussion, and of those who received correct advice, all failed to comply with the 3-week stand-down regulation. CONCLUSIONS: The paucity of return-to-play advice received by community rugby union players postconcussion and the high level of non-compliance with return-to-play regulations highlight the need for better dissemination and implementation of the return-to-play regulations and improved understanding of the underlying causes of why players do not adhere to return-to-play practices.

Mild traumatic brain injury among a cohort of rugby union players: predictors of time to injury.

This study reports the time to sustain a mild traumatic brain injury (mTBI) among a cohort of community rugby union players. Demographic and player characteristics were collected and players followed up for between one and three playing seasons. 7% of the cohort sustained an mTBI within 10 h of game time, increasing twofold to 14% within 20 h. The mean time to first mTBI was 8 h with an SD of 6.2 (median 6.8 h; IQR: 2.9-11.7 h). Players reporting a recent history of concussion were 20% more likely to sustain an mTBI after 20 h of game time compared with those with no recent history of concussion. Players were likely to sustain an mTBI in shorter time if they trained for <3 h/week (HR=1.48, p=0.03) or had a body mass index <27 (HR=1.77, p=0.007). The findings highlight modifiable characteristics to reduce the likelihood of shortened time to mTBI.

Motorcycle rider training for the prevention of road traffic crashes.

BACKGROUND: Riding a motorcycle (a two-wheeled vehicle that is powered by a motor and has no pedals) is associated with a high risk of fatal crashes, particularly in new riders. Motorcycle rider training has therefore been suggested as an important means of reducing the number of crashes, and the severity of injuries. OBJECTIVES: To quantify the effectiveness of pre- and post-licence motorcycle rider training on the reduction of traffic offences, traffic crash involvement, injuries and deaths of motorcycle riders. SEARCH STRATEGY: We searched the Cochrane Injuries Group Specialised Register, CENTRAL (The Cochrane Library 2008, Issue 3), TRANSPORT, MEDLINE, EMBASE, CINAHL, WHOLIS (World Health Organization Library Information System), PsycInfo, LILACS (Latin American and Caribbean Health Sciences), ISI Web of Science: Social Sciences Citation Index (SSCI), ERIC, ZETOC and SIGLE. Database searches covered all available dates up to October 2008. We also checked reference lists of relevant papers and contacted study authors in an effort to identify published, unpublished and ongoing trials related to motorcycle rider training. SELECTION CRITERIA: We included all relevant intervention studies such as randomised and non-randomised controlled trials, interrupted time-series and observational studies such as cohort and case-control studies. DATA COLLECTION AND ANALYSIS: Two review authors independently analysed data about the study population, study design and methods, interventions and outcome measures as well as data quality from each included study, and compared the findings. We resolved differences by discussion with a third review author. MAIN RESULTS: We reviewed 23 studies: three randomised trials, two non-randomised trials, 14 cohort studies and four case-control studies. Five examined mandatory pre-licence training, 14 assessed non-mandatory training, three of the case-control studies assessed 'any' type of rider training, and one case-control study assessed mandatory pre-licence training and non-mandatory training. The types of assessed rider training varied in duration and content.Most studies suffered from serious methodological weaknesses. Most studies were non-randomised and controlled poorly for confounders. Most studies also suffered from detection bias due to the poor use of outcome measurement tools such as the sole reliance upon police records or self-reported data. Small sample sizes and short follow-up time after training were also common. AUTHORS' CONCLUSIONS: Due to the poor quality of studies identified, we were unable to draw any conclusions about the effectiveness of rider training on crash, injury, or offence rates. The findings suggest that mandatory pre-licence training may be an impediment to completing a motorcycle licensing process, possibly indirectly reducing crashes through a reduction in exposure. It is not clear if training (or what type) reduces the risk of crashes, injuries or offences in motorcyclists, and a best rider training practice can therefore not be recommended. As some type of rider training is likely to be necessary to teach motorcyclists to ride a motorcycle safely, rigorous research is needed.

Barriers to, and factors associated, with observed motorcycle helmet use in Vietnam.

This study investigated barriers to, and factors associated with, observed motorcycle helmet use among motorcyclists in Hai Duong Province, Vietnam. The findings highlighted an array of factors associated with observed helmet use namely, support for universal helmet legislation and a positive attitude towards what might be perceived as negative attributes of helmet use such as inconvenience and discomfort in hot weather. As well, older age (greater than 25 years in age), riding on a compulsory road, being a driver, trips of greater than 10 km, higher levels of education (having a university degree and higher) were found to be key determinants of helmet use. Despite over 95% of motorcyclists disagreeing with the statement that wearing a helmet does not reduce the severity of head injury in a crash, most motorcyclists believed that helmets did not need to be worn for a short trip. Overall, only 23% of motorcyclists were observed wearing a helmet. The authors conclude that efforts to increase helmet use need to focus on the necessity for universal helmet legislation in association with identifying solutions to reduce the negative attitudes towards helmet use.

Cyclist exposure to the risk of car door collisions in mixed function activity centers: A study in Melbourne, Australia.

OBJECTIVE: The objective of this study is to describe key risks related to dooring collisions on roads in mixed function activity centers by examining video footage recorded onboard the bicycles of cyclists riding through this road environment. The study aims to enhance our understanding of the risk associated with cyclist door collisions on these roads and to provide a focus for future studies that aim to identify measures that enhance cyclist safety. METHOD: The study measured 4 key risk exposures (per hour and kilometer), namely, on-street parked cars and 3 conditions associated with parked cars being accessed or egressed: door opened (a) after the cyclist passes (give-way event); (b) in the path of the cyclist without collision (obstruction event); and (c) in the path of the cyclist with collision (collision event). Exposure to the risk factors was measured using video footage recorded onboard the bicycles of adult cyclists (n = 25) as they rode through this road environment. The average speed of cyclists was also estimated from these video data and related to the measures of exposure. RESULTS: The cycling experiences of the participants were observed over 3 h 58 min and 84.0 km. On average, the sample was exposed to 1,166 parked car events per hour (55 per kilometer), 6.9 give-way events per hour (0.3 per kilometer), 2.3 obstruction events per hour (0.1 per kilometer), and no collision events. There were 9 instances of obstruction events. In most cases, the cyclist was clearly visible (n = 7; 77.8%) and modified his or her position to avoid the door (n = 8; 88.9%). Moreover, the door was opened by the driver in nearly all cases (n = 8; 88.9%), primarily to exit the vehicle (n = 4; 44.4%) or enter the vehicle (n = 3; 33.3%). The average speed of cyclists was 22.7 km/h (SD = 4.1 km/h), and average speed tended to reduce as exposure to parked car events increased. CONCLUSION: Cyclists seldom interact with a vehicle occupant accessing or egressing a parked car, yet a concerning proportion of these interactions involve the occupant opening the door in the path of the cyclist, rather than waiting for the cyclist to pass. This suggests that there is still considerable need to identify measures that increase the likelihood that a vehicle occupant will look for cyclists before opening the car door, particularly in road environments where longer-term solutions such as physical separation are not readily achieved.

The contribution of passengers versus mobile phone use to motor vehicle crashes resulting in hospital attendance by the driver.

There is evidence that mobile phone use while driving (including hands-free) is associated with motor vehicle crashes. However, whether the effects of mobile phone use differ from that of passengers in the vehicle remains unclear. The aim of this research was to estimate the risk of crash associated with passenger carriage and compare that with mobile phone use. A case-control study ('passenger study') was performed in Perth, Western Australia in 2003 and 2004. Cases were 274 drivers who attended hospital following a motor vehicle crash and controls were 1096 drivers (1:4 matching) recruited at service stations matched to the location and time and day of week of the crash. The results were compared with those of a case-crossover study ('mobile phone study') undertaken concurrently (n=456); 152 cases were common to both studies. Passenger carriage increased the likelihood of a crash (adjusted odds ratio (adj. OR), 95% confidence interval (95% CI), 1.6, 1.1-2.2). Drivers carrying two or more passengers were twice as likely to crash as unaccompanied drivers (adj. OR 2.2, 95% CI 1.3-3.8). By comparison, driver's use of a mobile phone within 5 min before a crash was associated with a fourfold increased likelihood of crashing (OR 4.1, 95% CI 2.2-7.7). Passenger carriage and increasing numbers of passengers are associated with an increased likelihood of crash, though not to the same extent as mobile phone use. Further research is needed to investigate the factors underlying the increased risks.

The prevalence of, and factors associated with, serious crashes involving a distracting activity.

The study's objectives were to determine the prevalence and types of distracting activities involved in serious crashes, and to explore the factors associated with such crashes. We interviewed 1367 drivers who attended hospital in Perth, Western Australia between April 2002 and July 2004 following a crash. A structured questionnaire was administered to each driver and supplementary data were collected from ambulance and medical records. Over 30% of drivers (433, 31.7%) cited at least one distracting activity at the time of crashing and driver distraction was reported to have contributed to 13.6% of all crashes. The major distracting activities were conversing with passengers (155, 11.3%), lack of concentration (148, 10.8%) and outside factors (121, 8.9%). Using logistic regression, a distracting activity at the time of a crash was significantly more likely among drivers with shorter driving experience (0-9 years, 38.3% versus >or=30 years, 21.0%, p<0.001). Distracting activities at the time of serious crashes are common and can cause crashes, and the types of activities reported are varied. Increased driver awareness of the adverse consequences of distracted driving with a focus on novice drivers, enforcement of existing laws (e.g. those requiring a driver to maintain proper control of a vehicle), and progress on engineering initiatives (such as collision warning systems) are needed to reduce injury.

Determinants of the occupational environment and heavy vehicle crashes in Western Australia: A case-control study.

OBJECTIVE: To determine the association between a heavy vehicle driver's work environment, including fatigue-related characteristics, and the risk of a crash in Western Australia. METHODS: This case-control study included 100 long-haul heavy vehicle drivers who were involved in a police-reported crash in WA and 100 long-haul heavy vehicle drivers recruited from WA truck stops, who were not involved in a crash in the previous 12 months. Driver demographics and driving details, work environment, vehicle and sleep-related characteristics were obtained using an interviewer-administered questionnaire. Drivers were tested for obstructive sleep apnoea using an overnight diagnostic device. Conditional multiple logistic regression analysis was undertaken to determine work environment-related factors associated with crash involvement. RESULTS: After accounting for potential confounders, driving a heavy vehicle with an empty load was associated with almost a three-fold increased crash risk compared to carrying general freight (adjusted OR: 2.93, 95% CI: 1.17-7.34). Driving a rigid heavy vehicle was associated with a four-fold increased risk of crashing compared to articulated heavy vehicles (adjusted OR: 4.08, 95% CI: 1.13-14.68). The risk of crashing was almost five times higher when driving more than 50% of the trip between midnight and 5.59am (adjusted OR: 4.86, 95% CI: 1.47-16.07). Furthermore, the risk of crashing significantly increased if the time since the last break on the index trip was greater than 2h (adjusted OR: 2.18, 95% CI: 1.14-4.17). Drivers with more than 10 years driving experience were 52% less likely to be involved in a crash (adjusted OR: 0.48, 95% CI: 0.23-0.99). CONCLUSION: The results provide support for an association between a driver's work environment, fatigue-related factors, and the risk of heavy vehicle crash involvement. Greater attention needs to be paid to the creation of a safer work environment for long distance heavy vehicle drivers.

Estimating population-based injury prevention practices.

This study piloted a survey suitable to generate statewide prevalence estimates for injury prevention practices and/or behaviours. A number of prevalence estimates of injury prevention practices and/or behaviours were generated in the pilot. For example, the prevalence of a household smoke alarm was high (732/1000 households) and yet only 44% of households had tested the functionality of the alarm in the 12 months preceding the survey. Findings of this kind point to an increasing need for primary care practitioners to play a leading role in targeting preventive strategies for injury. In addition, the survey methods and approach highlight the value of such information to health departments and non-government agencies in terms of targeting priority injury risk behaviours and the evaluation of injury-related intervention programmes. The survey has been designed so the methods can be applied nationally and internationally.

Modelling bivariate count series with excess zeros.

Bivariate time series of counts with excess zeros relative to the Poisson process are common in many bioscience applications. Failure to account for the extra zeros in the analysis may result in biased parameter estimates and misleading inferences. A class of bivariate zero-inflated Poisson autoregression models is presented to accommodate the zero-inflation and the inherent serial dependency between successive observations. An autoregressive correlation structure is assumed in the random component of the compound regression model. Parameter estimation is achieved via an EM algorithm, by maximizing an appropriate log-likelihood function to obtain residual maximum likelihood estimates. The proposed method is applied to analyze a bivariate series from an occupational health study, in which the zero-inflated injury count events are classified as either musculoskeletal or non-musculoskeletal in nature. The approach enables the evaluation of the effectiveness of a participatory ergonomics intervention at the population level, in terms of reducing the overall incidence of lost-time injury and a simultaneous decline in the two mean injury rates.

Obstructive sleep apnea, health-related factors, and long distance heavy vehicle crashes in Western Australia: a case control study.

STUDY OBJECTIVES: To determine the association between obstructive sleep apnea (OSA), health-related factors and the likelihood of heavy vehicle crashes in Western Australia (WA). METHODS: This case-control study included 100 long-haul heavy vehicle drivers who were involved in a police-reported crash in WA during the study period (cases) and 100 long-haul heavy vehicle drivers recruited from WA truck stops, who were not involved in a crash during the past year (controls). Driver demographics, health, and fatigue-related characteristics were obtained using an interviewer administered questionnaire. Drivers were tested for OSA using a diagnostic Flow Wizard. Logistic regression was used to determine health-related factors associated with crash involvement among long distance heavy vehicle drivers. RESULTS: Heavy vehicle drivers diagnosed with OSA through the use of the FlowWizard were over three times more likely to be involved in a crash than drivers without OSA (adjusted OR: 3.42, 95% CI: 1.34-8.72). The risk of crash was significantly increased if heavy vehicle drivers reported a diagnosis of depression (adjusted OR: 6.59, 95% CI: 1.30-33.24) or had not completed fatigue management training (adjusted OR: 6.05, 95% CI: 1.80-20.24). Crash risk was 74% lower among older drivers (> 35 years) than younger drivers (adjusted OR: 0.25, 95% CI: 0.08-0.82). CONCLUSION: The results suggest that more rigorous screening and subsequent treatment of OSA and depression by clinicians as well as compulsory fatigue management training may reduce crashes among heavy vehicle drivers. COMMENTARY: A commentary on this article appears in this issue on page 409.

Geospatial analysis of cyclist injury trends: an investigation in Melbourne, Australia.

OBJECTIVE: This study applied geospatial analysis to explore spatial trends in cycling-related injury in Melbourne, Australia, in order to identify an area where injury density was reducing against expectation. The crash characteristics and cycling environment of the identified area were examined to better understand factors related to cycling safety. METHOD: Two methods were used to examine spatial trends in cycling-related injury. Firstly, cycling injury density was calculated using a kernel density estimation method for the years 2000 to 2011. This was used to examine patterns in injury density across Melbourne over an extended time period. Secondly, absolute change in injury density was calculated between 2005 and 2011. From this, a geographical area presenting a reduced injury density was selected for a case study, and crash characteristics of the area were obtained for the observational period. This led to discussion on which changes to the cycling environment, if any, may be associated with the reduced injury rate. RESULTS: Injury density in Melbourne had been progressively increasing between 2000 and 2011, with a nearly 3-fold increase in the peak injury density over that period. Decreases were observed in some locations between 2005 and 2011, and a geographical area to the southeast of Melbourne experienced a more significant decrease than others. This appeared to be associated with a combination of behavior and road infrastructure change, although a lack of data to verify change in cycling exposure prevented more definitive associations from being established. CONCLUSION: The apparent positive response of the injury rate to behavior and road infrastructure interventions is promising, yet the injury rate is unlikely to achieve the government's road safety target of 30% reduction in serious injuries by 2022. Moreover, the number of injuries sustained at the most common crash location appears to be increasing. Further research is necessary to discern which specific features of the urban road infrastructure have an effect on the risk of injury to a cyclist and which combination of features is consistent with a safe cycling environment.