Dataset Information

A systematic review of statistical models and outcomes of predicting fatal and serious injury crashes from driver crash and offense history data

ABSTRACT: Background Expenditure on driver-related behavioral interventions and road use policy is often justified by their impact on the frequency of fatal and serious injury crashes. Given the rarity of fatal and serious injury crashes, offense history, and crash history of drivers are sometimes used as an alternative measure of the impact of interventions and changes to policy. The primary purpose of this systematic review was to assess the rigor of statistical modeling used to predict fatal and serious crashes from offense history and crash history using a purpose-made quality assessment tool. A secondary purpose was to explore study outcomes. Methods Only studies that used observational data and presented a statistical model of crash prediction from offense history or crash history were included. A quality assessment tool was developed for the systematic evaluation of statistical quality indicators across studies. The search was conducted in June 2019. Results One thousand one hundred and five unique records were identified, 252 full texts were screened for inclusion, resulting in 20 studies being included in the review. The results indicate substantial and important limitations in the modeling methods used. Most studies demonstrated poor statistical rigor ranging from low to middle quality. There was a lack of confidence in published findings due to poor variable selection, poor adherence to statistical assumptions relating to multicollinearity, and lack of validation using new data. Conclusions It was concluded that future research should consider machine learning to overcome correlations in the data, use rigorous vetting procedures to identify predictor variables, and validate statistical models using new data to improve utility and generalizability of models. Systematic review registration PROSPERO CRD42019137081

SUBMITTER: Slikboer R

PROVIDER: S-EPMC7523043 | biostudies-literature | 2020 Jan

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:ObjectivesIn this study, we aimed to determine whether three minimum legal drinking age 21 (MLDA-21) laws-dram shop liability, responsible beverage service (RBS) training, and state control of alcohol sales-have had an impact on underage drinking and driving fatal crashes using annual state-level data, and compared states with strong laws to those with weak laws to examine their effect on beer consumption and fatal crash ratios.MethodsUsing the Fatality Analysis Reporting System, we calculated the ratio of drinking to nondrinking drivers under age 21 involved in fatal crashes as our key outcome measure. We used structural equation modeling to evaluate the three MLDA-21 laws. We controlled for covariates known to impact fatal crashes including: 17 additional MLDA-21 laws; administrative license revocation; blood alcohol concentration limits of.08 and.10 for driving; seat belt laws; sobriety checkpoint frequency; unemployment rates; and vehicle miles traveled. Outcome variables, in addition to the fatal crash ratios of drinking to nondrinking drivers under age 21 included state per capita beer consumption.ResultsDram shop liability laws were associated with a 2.4% total effect decrease (direct effects: β =.019, p =.018). Similarly, RBS training laws were associated with a 3.6% total effect decrease (direct effect: β =.048, p =.001) in the ratio of drinking to nondrinking drivers under age 21 involved in fatal crashes. There was a significant relationship between dram shop liability law strength and per capita beer consumption, F (4, 1528) = 24.32, p <.001, partial η(2) =.016, showing states with strong dram shop liability laws (Mean (M) = 1.276) averaging significantly lower per capita beer consumption than states with weak laws (M = 1.340).ConclusionsDram shop liability laws and RBS laws were both associated with significantly reduced per capita beer consumption and fatal crash ratios. In practical terms, this means that dram shop liability laws are currently associated with saving an estimated 64 lives in the 45 jurisdictions that currently have the law. If the remaining 6 states adopted the dram shop law, an additional 9 lives could potentially be saved annually. Similarly, RBS training laws are associated with saving an estimated 83 lives in the 37 jurisdictions that currently have the laws. If the remaining 14 states adopted these RBS training laws, we estimate that an additional 28 lives could potentially be saved.

Project description:ObjectiveAccurate road crash reporting is essential for evaluating road safety interventions and plans. Under-reporting of road traffic crashes, injuries, and fatalities in police records has been widely described. The aim of this study was to apply and evaluate a community crash recording system, and assess the quality of the data in comparison with traffic police data in Nepal.MethodsThe crash data collection methodology involved recruiting Local Record Keepers working and living at locations known to be at a high risk of crashes. Six shopkeepers were recruited at three crash locations and trained to complete collision incident forms for crashes occurring over one year on the section of road visible from their premises. Manual traffic volume and pedestrian counts, and spot speed surveys were conducted. Data were compared with District police records for the same period and locations.ResultsOver 12 months, 110 crashes were recorded by the Local Record Keepers. Of these, 70 resulted in 145 injuries (5 fatalities, 62 severe, and 78 minor injuries), while 40 resulted only in property damage. Comparable police data recorded 23 crashes, of which 18 crashes resulted in 27 injuries (8 fatalities, 13 serious, and 6 minor injuries), and 5 crashes in property damage only. The difference in recording of fatal and serious injuries was statistically significant (χ2(1) = 19.94, p < 0.001). The police reporting rate was highest for fatalities (62.5%) but only 11.6% and 7.1% for property damage cases and minor injuries respectively, and 3.8% for single-vehicle crashes. Compared to the Local Record Keeper data, the overall police crash reporting rate was 19.7%.ConclusionsLocal Record Keepers' recording of road traffic crashes and casualties is feasible and provides a more complete record than routinely collected police data. The low reporting rate in the police records of minor injury, property damage, and single-vehicle crashes suggest significant underestimation and bias in the reporting of the actual burden of road traffic crashes. Local Record Keeper recording is a viable method for validating police reports.

Dataset Information

A systematic review of statistical models and outcomes of predicting fatal and serious injury crashes from driver crash and offense history data

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets