Accident; analysis and prevention
-
Injury narratives are now available real time and include useful information for injury surveillance and prevention. However, manual classification of the cause or events leading to injury found in large batches of narratives, such as workers compensation claims databases, can be prohibitive. In this study we compare the utility of four machine learning algorithms (Naïve Bayes, Single word and Bi-gram models, Support Vector Machine and Logistic Regression) for classifying narratives into Bureau of Labor Statistics Occupational Injury and Illness event leading to injury classifications for a large workers compensation database. ⋯ Integrating LR into this ensemble mix improved performance only slightly. For large administrative datasets we propose incorporation of methods based on human-machine pairings such as we have done here, utilizing readily-available off-the-shelf machine learning techniques and resulting in only a fraction of narratives that require manual review. Human-machine ensemble methods are likely to improve performance over total manual coding.
-
Pedestrian crashes are an important issue globally as pedestrians are a highly vulnerable road user group, accounting for approximately 35% of road deaths worldwide each year. In highly motorised countries, pedestrian distraction by hand held technological devices appears to be an increasing factor in such crashes. An online survey (N=363) was conducted to 1) obtain prevalence information regarding the extent to which people cross the road while simultaneously using mobile phones for potentially distracting activities; 2) identify whether younger adult pedestrians are more exposed to/at risk of injury due to this cause than older adults; and 3) explore whether the Theory of Planned Behaviour (TPB) might provide insight into the factors influencing the target behaviours. ⋯ Moreover, high exposure was associated with stronger intentions to use a smart phone while crossing, and the effect was large, suggesting high frequency mobile phone use may lead to riskier habits, such as failing to interrupt use while crossing the road. Interventions should target pedestrians under 30 years old and aim to strengthen negative attitudes towards using smart phones while crossing, or to challenge the perceived advantages or emphasise the disadvantages of using one's phone while crossing in order to reduce intentions to do so. Young people's perceptions that others in their social group approve of smart phone use while crossing could also be an important factor to address.
-
This study compared the ability of five injury severity measures, namely the Abbreviated Injury Scale to the Head (AIS-H), Glasgow Coma Scale (GCS), Glasgow Outcome Scale (GOS), Extended Glasgow Outcome Scale (GOSE), and Injury Severity Score (ISS), to predict return-to-work after a traumatic brain injury (TBI). Furthermore, factors potentially associated with return-to-work were investigated. In total, 207 individuals aged ≤65 years newly diagnosed with a TBI and employed at the time of injury were recruited and followed-up for 1year by telephone every 3 months. ⋯ A multivariable analysis revealed that individuals with higher injury severity as measured by the ISS (hazard ratio [HR], 0.94; 95% confidence interval [CI], 0.92-0.97), a lack of autonomy in transportation (HR, 2.55; 95% CI, 1.23-5.32), cognitive impairment (HR, 0.47; 95% CI, 0.28-0.79), and depression (HR, 0.97; 95% CI, 0.95-0.99) were significantly less likely to be employed after a TBI. In conclusion, of the five injury severity measures, the ISS may be the most capable measure of predicting return-to-work after a TBI. In addition to injury severity, autonomy in transportation, cognitive function, and the depressive status may also influence the employment status during the first year after a TBI.
-
This paper presents an analysis of drivers' stopping and traversing behaviors during inter-green periods. Eight intersections were observed in Changchun, China both with and without countdown timers and/or video surveillance during summer and winter. The impacts of the devices on the drivers' behavior were examined and compared between the two seasons from a safety perspective. ⋯ Three impacts are studied, including the profile of approaching speeds, the stop/go decision, and the maximum acceleration and deceleration. The findings revealed that installing both a countdown timer and CCTV in summer, or either of the devices in winter can increase drivers' stopping tendency and hence reduce red-light violations. Especially on an icy road during winter, a countdown timer can help smooth decelerations, which tend to begin earlier than at the intersections without the device, reducing the incidence of sudden braking.
-
Long-haul truck drivers in the United States suffer disproportionately high injury rates. Sleep is a critical factor in these outcomes, contributing to fatigue and degrading multiple aspects of safety-relevant performance. Both sleep duration and sleep quality are often compromised among truck drivers; however, much of the efforts to combat fatigue focus on sleep duration rather than sleep quality. Thus, the current study has two objectives: (1) to determine the degree to which sleep impacts safety-relevant performance among long-haul truck drivers; and (2) to evaluate workday and non-workday sleep quality and duration as predictors of drivers' safety-relevant performance. ⋯ Sleep quality appears to be better associated with safety-relevant performance among long-haul truck drivers than sleep duration. Comprehensive and multilevel efforts are needed to meaningfully address sleep quality among drivers.