Traffic Inj Prev [JOURNAL]
Sun
200 papers
Exploring the latent factors affecting motorcycle crashes by using latent class clustering analysis and multinomial logit models.
Fang CW, Ou YK, Chen JJ … +1 more , Chen WR
Traffic Inj Prev · 2026 Feb · PMID 41747272 · Publisher ↗
OBJECTIVE: Motorcycle crashes constitute a critical safety issue in Taiwan, particularly in Tainan, a city characterized by exceptionally high motorcycle ownership, dense traffic environments, and complex street network...
OBJECTIVE: Motorcycle crashes constitute a critical safety issue in Taiwan, particularly in Tainan, a city characterized by exceptionally high motorcycle ownership, dense traffic environments, and complex street network designs. This study aimed to uncover latent crash typologies and identify key determinants of motorcycle injury severity across diverse crash contexts. METHODS: A total of 673 motorcycle-involved crashes from the Tainan City Traffic Accident Investigation Committee were analyzed. Latent Class Clustering (LCC) was used to classify crashes into six heterogeneous clusters based on rider demographics, roadway characteristics, environmental conditions, and crash configurations. Subsequently, cluster-specific Multinomial Logit (MNL) models were estimated to assess how these factors influence the likelihood of mild, moderate, or severe injury outcomes. RESULTS: Six distinct crash scenarios were identified, reflecting meaningful heterogeneity across rider age, lighting conditions, lane configurations, and roadway speed limits. Across clusters, older riders (≥60 years) had a substantially higher probability of sustaining severe injuries, while crashes occurring on roadways with ≤40 km/h speed limits showed unexpectedly elevated severity risks, likely attributable to narrow lanes and high conflict density. Seasonal effects were also observed, with winter crashes demonstrating significantly higher odds of severe injury. Notably, several risk factors exerted different magnitudes and directions of influence across clusters, highlighting strong context dependency in motorcycle crash severity mechanisms. CONCLUSIONS: This study provides robust empirical evidence that motorcycle crash severity in Tainan is shaped by a combination of rider characteristics, environmental conditions, and underlying latent crash patterns. The findings support the development of targeted, cluster-specific safety measures, including risk-tailored rider education, intersection and lighting improvements, and enhanced speed management strategies. Such interventions can help reduce motorcycle crash incidence and mitigate injury severity in high-risk urban environments.
Machine learning and Geographic Information Systems (GIS)-based model for road crash severity prediction and behavioural pattern analysis.
A Al-Mashagba L, Sumari P, Mashagba M … +4 more , Al Abri F, Mashagba HA, Abd Aziz AB, Al-Bawri SS
Traffic Inj Prev · 2026 Feb · PMID 41707203 · Publisher ↗
OBJECTIVES: Road traffic injuries constitute a major public health problem in low- and middle-income countries, including Jordan. This study develops a hybrid machine learning and Geographic Information Systems (GIS)-bas...
OBJECTIVES: Road traffic injuries constitute a major public health problem in low- and middle-income countries, including Jordan. This study develops a hybrid machine learning and Geographic Information Systems (GIS)-based framework to predict crash severity and identify behavioral, environmental, and infrastructural factors contributing to injury outcomes. METHODS: A dataset of 11,345 crashes from the Jordan Traffic Institute (2018) was analyzed. After encoding, imputation, normalization, and outlier treatment, a two-stage analytical design was applied. First, association rule mining (Apriori; minimum support 0.05, confidence 0.60) was used to uncover dominant behavioral and environmental patterns linked to different injury levels. Second, Decision Tree, Random Forest, and AdaBoost models were trained (70/30 split) to classify severity (slight, medium, severe, fatal). Model performance was evaluated using accuracy, precision, recall, F1-score, and confusion matrices. GIS-based kernel density estimation was used to detect spatial hotspots of severe and fatal crashes. RESULTS: Association rules revealed strong links between nighttime driving, young drivers, and lack of seatbelt use with fatal outcomes. GIS analysis identified high-risk clusters on major urban area in Zarqa, Awajan, and Al-Rusayfah. The Random Forest model achieved the highest predictive performance (accuracy: 98.5-99.9%). Key predictors included crash type, vehicle speed, time of day, driver age, and road characteristics. CONCLUSIONS: The proposed hybrid ML-GIS framework offers highly accurate severity prediction and reveals spatial and behavioral patterns critical for targeted safety interventions. Findings highlight the dominant influence of temporal and infrastructural factors over driver characteristics, supporting evidence-based engineering and enforcement strategies in Jordan.
Mapping clusters of road traffic accident fatalities and injuries: A spatiotemporal study in Isfahan Province, Iran.
Rahmati M, Aghamohammadi H, Behzadi S … +1 more , Alesheikh AA
Traffic Inj Prev · 2026 Feb · PMID 41707200 · Publisher ↗
OBJECTIVE: Road traffic accidents (RTAs) pose a significant challenge to global public health. In Iran, Isfahan Province experiences a high rate of RTAs, leading to many injuries and fatalities. Despite this, no comprehe...
OBJECTIVE: Road traffic accidents (RTAs) pose a significant challenge to global public health. In Iran, Isfahan Province experiences a high rate of RTAs, leading to many injuries and fatalities. Despite this, no comprehensive spatiotemporal analysis of RTA outcomes has been conducted in this region. To address this research gap, this study presents the first disaggregate-level spatiotemporal analysis of RTA injuries and fatalities in Isfahan Province by applying advanced spatial statistical techniques to identify critical patterns in RTA-related outcomes. METHODS: The research uses RTA data for the years 2017-2019 to investigate the spatiotemporal patterns of fatalities and injuries caused by RTAs. The analysis focuses on major suburban roadways in Isfahan Province, Iran. To measure overall spatial autocorrelation, Global Moran's I was employed, while localized hotspots were identified using the Getis-Ord Gi* statistic. Analyses were conducted on the whole dataset as well as separately for each year and each season to examine temporal variations. RESULTS: Moran's I test showed a strong spatial clustering pattern for RTA fatalities during the winter season. The spatial patterns during the other seasons did not reveal clustering, indicating randomness. Spatial clustering for the injury cases was consistently present across most temporal scales except during the autumn and winter seasons, where patterns approached randomness. Furthermore, Getis-Ord Gi* hotspot analysis detected obvious geographic hotspots and coldspots for both fatalities and injuries at different temporal scales, providing overall insights into high-risk areas. CONCLUSION: This study contributes to the field of traffic safety by providing the first spatiotemporal analysis of RTA injuries and fatalities in Isfahan Province, addressing a significant research gap. The application of spatial statistical methods at seasonal and annual scales unveiled distinct patterns of high-risk areas and differences in the spatial distributions of injuries and fatalities. The findings will help shift from broad policies to targeted, data-driven strategies that can guide enforcement, optimal placement of emergency stations, and resource allocation tailored to identified high-risk areas to enhance road safety and reduce RTA-related injuries and fatalities in the region.
Modeling driver information transmission during turning at unsignalized intersections: a driving simulator study.
Ni D, Zhou Y
Traffic Inj Prev · 2026 Feb · PMID 41707221 · Publisher ↗
OBJECTIVE: The proportion of older drivers has increased with the aging population. In order to solve the problem of high accident tendency at unsignalized intersections caused by improper driving behavior of older drive...
OBJECTIVE: The proportion of older drivers has increased with the aging population. In order to solve the problem of high accident tendency at unsignalized intersections caused by improper driving behavior of older drivers (OD-ER), this study analyzed turning behavior differences between OD-ERs and young and middle-aged drivers (YM-ERs), and then a petri net was used to describe the change in the behavior characteristics of two driver groups. The shortcoming of OD-ER's behavior can be found through the information transfer model, thereby improving driving safety. METHODS: Virtual scenarios of unsignalized intersections were built for drivers who were recruited from the general population for the study. Data on vehicle operation, driver visual behavior and physiological information were collected and analyzed. A petri net was applied to construct an information transfer model of turning behavior for two driver groups. RESULTS: YM-ERs performed better than OD-ERs in three types of data. The most significant difference was observed in the vehicle operation data, for which all examined metrics showed statistically significant differences ( < 0.05). The information transfer model indicated that the primary limitation for OD-ERs is saccade behavior, which introduced additional transmission delays and constrained information flow compared with YM-ERs, particularly during left turning scenarios. CONCLUSION: OD-ERs exhibit weaker turning ability than YM-ERs in terms of vehicle operation, information processing and judgment. Driver's turning behavior characteristic can be described from the turning information transfer model and then evaluate the driving quality, OD-ERs can identify their areas of focus from the graph for training and improvement of driving safety. It can be used for the application of improvement face to the older driver when they enter into the unsignalized intersection.
Partial cannabis legalization and the increase of the THC threshold in road traffic: a statistical analysis of traffic cases before and after legal changes.
Wohlfarth A, Franz T, Skopp GA … +1 more , Musshoff F
Traffic Inj Prev · 2026 Feb · PMID 41707218 · Publisher ↗
OBJECTIVE: On April 1, 2024, cannabis was partially legalized in Germany, accompanied by an increase in the statutory Δ9-tetrahydrocannabinol (THC) threshold for administrative traffic offenses from 1.0 to 3.5 ng THC/mL...
OBJECTIVE: On April 1, 2024, cannabis was partially legalized in Germany, accompanied by an increase in the statutory Δ9-tetrahydrocannabinol (THC) threshold for administrative traffic offenses from 1.0 to 3.5 ng THC/mL serum in August 2024. A key legislative objective was to protect frequent cannabis users who separate consumption from driving but still display THC levels exceeding 1.0 ng/mL from disproportionate sanctions. The law mandates a statistical evaluation of its impact after three years. Since no national statistics differentiate substance-impaired driving offenses by a specific drug class, we analyzed all traffic-related blood samples containing THC or THC-carboxylic acid detected in our laboratory over four years. METHODS: Included were cases from three federal states between April 1, 2021, and March 31, 2025, submitted for toxicological analysis with the legal context of administrative (§24 German Road Traffic Act) or criminal offenses (§316/§315c German Criminal Code). We examined case frequencies, age distributions, co-use of other substances, time intervals between incident and blood collection, consumption patterns, THC and THC-carboxylic acid concentrations, and frequencies of sanctions and driving errors. RESULTS: Among 48,058 total cases, 83% were administrative and 17% criminal offenses. 46% of drivers were aged 21-30 years; the proportion of those below 21 years decreased from 18.2% to 13.3%. In administrative offenses, THC alone was detected in 74% of cases, compared to only 30% in criminal cases. The time from incident to blood sampling was under 1.5 h in 82% of cases. Based on THC-carboxylic acid concentrations, four consumption categories were established: occasional, regular, repeated and chronic. The distribution across these groups was 47%, 22%, 19%, and 11%, respectively and median THC concentrations were 1.3, 4.5, 8.4, and 14.9 ng/mL serum. Remarkably, median THC levels in administrative and criminal offenses were identical: 3.44 ng/mL serum. Across individual years, the highest THC medians in serum occurred in the year following legalization (3.95 and 3.97 ng/mL). Applying the new 3.5 ng/mL limit to our case cohort would exempt 899 frequent and 11,855 occasional users with serum THC levels between 1.0 and < 3.5 ng/mL from sanctions. CONCLUSION: Our data indicate a modest upward shift in THC concentrations, continuing beyond April 2024, alongside a demographic shift toward older drivers. The nearly identical THC medians between drivers with and without conspicuous driving performance challenge the validity of the new threshold. Notably, almost half of the unimpaired drivers exceeded 3.5 ng/mL, whereas half of the impaired drivers remained below it. 7.4% of the frequent users benefited from the raised threshold, while at the same time 58.3% of occasional users-whose driving ability at THC concentrations between 1 and <3.5 ng/mL may be compromised-fell out of legal sanction. While substantial doubt remains if the initial goal was achieved, the question of an appropriate THC threshold remains unresolved and calls for alternative legal approaches.
Improved Balanced Random Forest (iBRF): A flexible hybrid resampling-bagging framework for crash severity classification in imbalanced datasets.
Mohammadpour SI, Vahedi J
Traffic Inj Prev · 2026 Feb · PMID 41707211 · Publisher ↗
OBJECTIVE: Class imbalance presents a notable challenge in accurately classifying severe crash outcomes, which typically constitute the minority class. Ignoring class imbalance may lead to biased or misleading interpreta...
OBJECTIVE: Class imbalance presents a notable challenge in accurately classifying severe crash outcomes, which typically constitute the minority class. Ignoring class imbalance may lead to biased or misleading interpretations. Moreover, relying on traditional resampling techniques results in potential issues like overfitting, and information loss. Hence, this study proposes the Improved Balanced Random Forest (iBRF), a novel ensemble framework designed to combine various resampling techniques with ensemble learning to derive a model that addresses the challenge of class imbalance. METHODS: The iBRF framework utilizes all minority cases and bootstrapped samples of the majority. SMOTE augments additional minority‑class instances, while the Neighborhood Cleaning Rule (NCR) and Random Undersampling (RUS) remove majority‑class instances to reduce noise and class overlap. The balanced data is utilized to train a decision tree in each iteration. Finally, a weighted soft voting based on the G-mean of each tree generates the final predictions. The performance of iBRF is tested using G-mean, and Matthews Correlation Coefficient (MCC) on holdout data (30%) and compared against other machine learning models designed for imbalanced datasets. RESULTS: iBRF achieves better performance than other resampling (SMOTE, RUS, NCR, and CTGANs) and state-of-the-art ensemble learning approaches (XGBoost, RF, LightGBM, RUSBoost, SMOTEBagging, SMOTEBoost, UnderBagging, OverBagging, and OverBoost) based on balanced metrics. The iBRF algorithm substantially improves the absolute G‑Mean of the simple RF and SMOTE‑RF models by 9.30 and 4.50 percentage points, respectively. CONCLUSIONS: The iBRF algorithm not only enhances classification performance but also helps identify risk factors associated with the minority class, which almost always correspond to severe‑outcome crashes and the focus of studies.
Study on the risk Assessment of vehicle groups considering driving styles.
Xiong X, Chen M, Cai Y … +2 more , Liu Q, Chen L
Traffic Inj Prev · 2026 Feb · PMID 41672723 · Publisher ↗
OBJECTIVE: To address the limited consideration of driving style diversity in group-level risk propagation analysis, a Dynamic Interaction Potential (DIP)-based framework is proposed to capture the risk evolution pattern...
OBJECTIVE: To address the limited consideration of driving style diversity in group-level risk propagation analysis, a Dynamic Interaction Potential (DIP)-based framework is proposed to capture the risk evolution patterns of vehicles with different driving styles. METHODS: Vehicle groups are identified by DBSCAN using the vehicle trajectories from HighD. To describe vehicle interactions within groups, DIP is constructed based on motion conflict and behavioral uncertainty. Fluctuation intensity and fluctuation rate are derived from DIP for risk measurement and driving style classification. The roles of different styles in risk propagation are quantified using Network analysis. A comprehensive vehicle group risk index is proposed that incorporates style-specific effects. RESULTS: Risk propagation network analysis reveals the distinct roles of driving styles, identifying aggressive drivers as key risk initiators within vehicle groups. The proposed style-aware risk index (CRAI_Style) demonstrates superior predictive performance compared to baseline indicators. In the one-second prediction window, the ROC-AUC for CRAI_Style is 0.845, followed by its style-agnostic counterpart CRAI (0.745) and traditional conflict-based indicator TTC (0.675). CRAI_Style consistently maintains the highest predictive accuracy when the prediction window extends to three seconds. CONCLUSION: These findings reveal the differentiated roles of driving styles in risk propagation, where aggressive drivers act as key risk initiators and conservative drivers as risk suppressors, and demonstrate that incorporating this heterogeneity significantly enhances the accuracy and robustness of vehicle group risk assessment.
Estimated real-world safety benefits associated with blind spot monitoring systems in Australasian passenger vehicles.
Keall MD, Newstead S
Traffic Inj Prev · 2026 Feb · PMID 41672590 · Publisher ↗
OBJECTIVE: Blind spot monitoring systems help drivers avoid collisions with another vehicle when changing lanes. The systems provide visual and/or haptic warnings; active systems additionally avoid collision by applying...
OBJECTIVE: Blind spot monitoring systems help drivers avoid collisions with another vehicle when changing lanes. The systems provide visual and/or haptic warnings; active systems additionally avoid collision by applying brakes or steering the vehicle. This study aimed to estimate real-world effectiveness of these technologies as installed in the Australasian light vehicle fleet in Australasian road and driving conditions. METHODS: Police-recorded crash data were studied for the years 2019-2023 from the Australian states Victoria, Queensland, South Australia, New South Wales and Western Australia and also New Zealand. Using quasi-induced exposure analysis, rates of lane change crash involvements were studied for vehicles manufactured from 2018, classified by whether they were fitted with blind spot monitoring systems or not. RESULTS: We found a statistically significant 15% reduction (95% CI 26%-3%) in lane change crashes for vehicles equipped with a blind spot monitoring system. There was indicative evidence that the system was more effective for male than female drivers. For crashes involving injury, the associated reduction was estimated to be larger: a 24% reduction (95% CI 38%-6%). Active systems were rare in our data, but the analysis suggested they may be even more effective than the warning systems. CONCLUSIONS: Despite the relatively low prevalence of lane change crashes generally, the estimated 15% reduction is significant. In the Australasian crash data analyzed, older drivers had an elevated rate of lane change crashes and may consequently gain greater benefit from blind spot monitoring systems.
How periodic decorations in extra-long highway tunnels affect driving safety: A simulation study.
Sun H, Ma F, Wang X … +2 more , Du Z, Xu R
Traffic Inj Prev · 2026 Feb · PMID 41672485 · Publisher ↗
OBJECTIVE: The enclosed environment and monotonous driving conditions of extra-long tunnels often induce driver fatigue and unstable operations, undermining both safety and comfort. As an effective intervention, tunnel l...
OBJECTIVE: The enclosed environment and monotonous driving conditions of extra-long tunnels often induce driver fatigue and unstable operations, undermining both safety and comfort. As an effective intervention, tunnel landscape decoration zones are designed to alleviate fatigue, enhance alertness, and improve driving continuity. To investigate the influence of tunnel landscape decoration zones on driving behavior, this study evaluates their comprehensive effects across vehicle operation characteristics, physiological responses, and visual behavior, while proposing optimization strategies. METHOD: This study used the under-construction Qingdao Jiaozhou Bay Second Subsea Tunnel in China as a case study. Twelve simulated landscape schemes were constructed within a driving simulation environment, culminating in a total of 28 samples. Multidimensional driver data-including vehicle operation, eye movement, and heart rate-were collected and analyzed. RESULT: Results show that patterned vaults and periodic sidewall design significantly improve operational stability, while a proper pattern spacing provides the optimal balance between visual stimulation and cognitive load. Among all tested conditions, featuring sidewall-consistent vault patterns, periodic sidewall design, and 15 m spacing achieved the best performance in both safety and comfort. CONCLUSION: These findings provide theoretical foundations and practical guidance for the systematic design and environmental evaluation of extra-long tunnels.
Effectiveness and heterogeneous impacts of traffic safety education on non-motorized vehicle violations.
Shi D, Cong H, Chen C … +4 more , Zhu J, Wang J, Wang E, Su P
Traffic Inj Prev · 2026 Feb · PMID 41672046 · Publisher ↗
OBJECTIVE: While Traffic Safety Education is regarded as a cornerstone intervention for reducing traffic violations, the persistence of its effects and the heterogeneity of its impact across populations remain inadequate...
OBJECTIVE: While Traffic Safety Education is regarded as a cornerstone intervention for reducing traffic violations, the persistence of its effects and the heterogeneity of its impact across populations remain inadequately investigated. This study evaluates the differential effectiveness and temporal dynamics of TSE among recidivist non-motorized vehicle users in City A Hangzhou, China, including riders of bicycles, electric bicycles and tricycles, and aims to identify subgroups that would benefit most from precision intervention strategies. METHODS: Traffic Safety Education is a mandatory intervention administered by traffic police, combined in-person and online modules on traffic rules and safety. Completion was required for all offenders identified through on-site or electronic monitoring. We developed a multidimensional analytical framework integrating segmented linear regression through interrupted time-series with event-study designs to estimate both immediate and sustained effects of TSE interventions. Kaplan-Meier survival analysis and first-order Markov transition matrices were employed to quantify behavioral persistence and transition patterns. The analysis utilized 68,426 officially documented violation records from 13,312 individuals who completed TSE between 2021 and 2023. Subgroup analyses by age, gender, and violation type were conducted to examine heterogeneous post-intervention behavioral trajectories. RESULTS: Substantial heterogeneity in TSE effectiveness was observed across demographic and behavioral subgroups. Recidivism declined steadily with age, from 43.7% among riders aged 18-25 to 36.1% among those over 60, indicating stronger behavioral self-regulation in older cohorts. Violations such as 'disobeying traffic signals' and 'failing to wear helmets' exhibited the highest recurrence rates, suggesting ingrained habits less amenable to short-term educational efforts. Female riders demonstrated a six-month recidivism rate approximately 10 percentage points lower than males. Overall, the violation-free survival probability declined to 56.6% within six months post-intervention, reflecting an initial improvement in compliance that attenuated over time. Transition analysis further indicated that most relapses occurred within the first three months, highlighting the short-term nature of behavioral modification. CONCLUSIONS: The results reveal significant heterogeneity in behavioral responses to TSE and limited durability of its deterrent effects. To foster long-term behavioral change, TSE programs should transition from uniform delivery to precision-targeted approaches, focusing on high-risk subgroups-especially younger riders, habitual offenders, and individuals with rapid relapse patterns. Integrating sustained educational reinforcement with targeted enforcement and personalized feedback may enhance behavioral retention and improve overall traffic safety outcomes.
A dual-model approach to traffic risk assessment of mountainous expressway tunnel groups.
Meng Y, Deng Y, Zhang X … +4 more , Li B, Qing G, Chen F, Wang K
Traffic Inj Prev · 2026 Feb · PMID 41672045 · Publisher ↗
OBJECTIVE: To improve the traffic risk conditions of mountainous expressway tunnel sections, it is necessary to conduct safety risk assessments and adopt different countermeasures according to the assessed risk levels. M...
OBJECTIVE: To improve the traffic risk conditions of mountainous expressway tunnel sections, it is necessary to conduct safety risk assessments and adopt different countermeasures according to the assessed risk levels. METHODS: An evaluation system was established with 4 primary indicators-tunnel condition, traffic characteristics, operational environment, and safety facilities-and 16 secondary indicators. Safety status was divided into 5 risk levels. To assign indicator weights objectively, information entropy was used to improve the traditional CRITIC method. Two assessment models based on extension matter-element theory and set pair analysis were then developed to form a dual-verification mechanism: the former handles indicator-grade incompatibility correlation functions, while the latter treats assessment uncertainty using multiple connection numbers. RESULTS: Fifteen tunnels on the Guangzhou-Kunming Expressway in Yunnan Province were selected as evaluation objects. The improved CRITIC method effectively reduced subjective bias, with key indicator weights adjusted by up to 10% for more objective weighting. The extension matter-element model and set pair analysis (SPA) model yielded highly consistent dual assessment results (agreement rate >80%). Most tunnels were classified as low-risk, while several long tunnels were categorized as medium-risk. The SPA model showed greater advantages in describing risk evolution trends, clearly characterizing transitions between adjacent risk levels potential series. CONCLUSION: The improved CRITIC method significantly enhances the objectivity of indicator weighting, making it more consistent with actual tunnel conditions. The combined application of the extension matter-element model and the set pair analysis model form a complementary dual verification mechanism. Case studies verified that this integrated evaluation system can accurately determine tunnel safety levels and provide a reliable basis for developing targeted risk prevention and control measures.
Socioeconomic influences on drunk driving recidivism: an integrated Prototype Willingness Model approach.
Jou RC, Hsu HW
Traffic Inj Prev · 2026 Feb · PMID 41637574 · Publisher ↗
OBJECTIVE: Drunk driving offenders represent a high-risk group for repeated violations. This study employs an integrated Prototype Willingness Model (PWM) framework to examine whether socioeconomic characteristics signif...
OBJECTIVE: Drunk driving offenders represent a high-risk group for repeated violations. This study employs an integrated Prototype Willingness Model (PWM) framework to examine whether socioeconomic characteristics significantly influence various psychological constructs, and provides policy recommendations to inform more effective prevention strategies. METHODS: Multivariate Analysis of Variance (MANOVA) was used to analyze the effects of socioeconomic variables on multiple psychological constructs, including attitude, subjective norms, perceived behavioral control, prototype, drunk driving behavioral intention, and behavioral willingness. Post hoc tests were conducted to explore group differences across socioeconomic variables. The study sample comprised individuals who were apprehended for drunk driving and attended road safety courses at regulatory agencies. RESULTS: The results indicate no significant overall differences in the integrated PWM dimensions between first-time and repeat drunk driving offenders, except for significant differences in attitude and recidivism behavior. Socioeconomic variables-including gender, age, marital status, education level, monthly income, and age at first alcohol consumption-showed significant effects on the integrated PWM dimensions, reflecting group differences in psychological constructs. These findings highlight the associations between socioeconomic factors and psychological mechanisms, providing an empirical basis for developing targeted prevention and intervention strategies. CONCLUSIONS: This study reveals that drunk driving recidivism is influenced by the interplay of psychological constructs and socioeconomic characteristics, with attitude, subjective norms, and perceived behavioral control playing key roles in decision-making. Applying MANOVA to the PWM framework facilitates identification of group differences and supports multi-level, targeted interventions. It is recommended that government implement integrated measures combining legal enforcement, education, and social counseling to reduce recidivism and enhance traffic safety. Given the study's cross-sectional design, reliance on self-reported data, and regional limitations, future research should adopt longitudinal designs and larger samples to improve generalizability and policy applicability.
Stable factors associated with traffic conflict risk in expressway weaving segments: A causal modeling framework.
Zhao J, Jiang H, Zhao X … +3 more , Luo G, Liu Q, Zheng Q
Traffic Inj Prev · 2026 Feb · PMID 41637573 · Publisher ↗
OBJECTIVES: This study aims to identify stable factors associated with traffic conflict risk in expressway weaving segments, with a particular focus on addressing the challenge of unobserved data distribution bias betwee...
OBJECTIVES: This study aims to identify stable factors associated with traffic conflict risk in expressway weaving segments, with a particular focus on addressing the challenge of unobserved data distribution bias between training and test datasets, which can compromise model reliability. METHODS: To mitigate distribution bias and enhance result robustness, a causally regularized logistic model (CRLM) with a global causal regularizer was employed. To validate the stability of the CRLM, multi-dataset validation and model parameter consistency tests were conducted using five datasets collected from the field and simulation in two weaving types. Meanwhile, classic logistic regression (LR) and eXtreme Gradient Boosting (XGBoost) were developed for comparison. RESULTS: In the multi-dataset validation test, the average area under receiver operating characteristic curve (AUC) of the CRLMs is close to that of the XGBoosts, but with a lower standard deviation, suggesting that the CRLM provides more stable predictive performance across different combinations of training and testing datasets. In the model parameter consistency test, the CRLM can identify more stable factors across heterogeneous traffic environments. Furthermore, the causal mechanisms underlying traffic conflict risk in Type A and Type B weaving segments are distinct. The hazardous traffic flow characteristics for each weaving type were discussed in detail. CONCLUSIONS: These findings provide a novel and robust methodological framework for traffic conflict risk analysis. In addition, the model results have practical implications for developing proactive traffic control strategies and enhancing automated driving systems (ADS) to improve traffic safety in expressway weaving segments.
Perceived platform surveillance, mental health and road risk among couriers and food delivery workers: Evidence from Wuhan, China.
Chen L, Wang R
Traffic Inj Prev · 2026 Feb · PMID 41637572 · Publisher ↗
OBJECTIVES: While platform surveillance of work processes is common in the gig economy, the subjective perception of surveillance varies among workers with different characteristics. This study uses the survey data of co...
OBJECTIVES: While platform surveillance of work processes is common in the gig economy, the subjective perception of surveillance varies among workers with different characteristics. This study uses the survey data of couriers and food delivery workers to explore the impact of perceived platform surveillance on road risks (risky riding and traffic crashes), and examines the mediating effect of mental health in this relationship, attempting to establish an explanatory framework among these factors. METHODS: The survey was conducted using a respondent-driven sampling method and included 801 couriers and food delivery workers in Wuhan, China. Propensity score matching was used to construct comparable perceived and unperceived groups prior to subsequent logit regression analysis. Causal mediation analysis was employed to examine the mediating effect of mental health in the relationship of perceived platform surveillance and road risk. RESULTS: The findings revealed that couriers and food delivery workers with higher socioeconomic status and greater work experience were more likely to perceive platform surveillance, which led them to higher odds of risky riding and traffic crashes. Mental health, as a mediator of the negative impact of perceived platform surveillance on road risk, contributed 14.2% of the total effect on risky riding and 17.2% of the total effect on traffic crashes, respectively. CONCLUSIONS: The study underscored the heterogeneity in the perception of platform surveillance among couriers and food delivery workers, which contributed to disparities in mental health and road risk consequences. Policy implications included enhancing the ethical accountability of platform algorithms, improving riders' safety awareness, and establishing a collaborative, multi-stakeholder road safety system.
Injury crash risk of battery and hybrid electric passenger cars.
Koisaari T, Kari T, Vahlberg T
Traffic Inj Prev · 2026 Feb · PMID 41637566 · Publisher ↗
OBJECTIVES: We examined at-fault injury crashes of four passenger car populations: Hybrid Electric Vehicles (HEVs), Plug-in Hybrid Electric Vehicles (PHEVs), Battery Electric Vehicles (BEVs) and traditional internal comb...
OBJECTIVES: We examined at-fault injury crashes of four passenger car populations: Hybrid Electric Vehicles (HEVs), Plug-in Hybrid Electric Vehicles (PHEVs), Battery Electric Vehicles (BEVs) and traditional internal combustion engine vehicles (ICEVs). For these populations, crash rates were calculated in relation to both registration years and mileage. Finally, controlled crash rate ratios were calculated to compare the crash risk between electric vehicles (EVs) and ICEVs. METHODS: Studied car populations were identified and their vehicle information for the period of 2019-2023, including the mileage (76 billion kilometers for all cars during the study period), was drawn from the national Vehicular and Driver Data Register. In addition, cars in the study populations were identified from the motor liability insurance (MLI) database and the crash data for them was retrieved (11,388 motor vehicle occupant injury crashes in total). Crash rates and crash rate ratios were calculated to evaluate the crash risk of EVs. Negative binomial regression was used to model crash involvement rate ratios both per registration year and per mileage for EVs, controlling the age and gender of the vehicle owner and vehicle size. RESULTS: Only battery electric vehicles showed significantly different crash rates than ICEVs per mileage, although the result was weakly significant -15% [-28%; 0%]. There were no significant differences in crash rates per registration years. In addition, there were only a few significant differences in crash circumstances between EVs and ICEVs. On average, the motor vehicle occupant injury crash rate of ICEVs was 151 crashes per billion kilometers and 2.37 crashes per thousand registration years. CONCLUSIONS: Our results indicate that, when measured by motor vehicle occupant injury crash rate, passenger cars-regardless of powertrain-have not become safer in Finland compared to the situation ten years ago. However, the current crash rate of BEVs is lower than that of ICEVs. Previous studies suggest that some of the differences in crash rate may be explained by varying usage conditions, which our findings support. Part of the difference may be explained by differences in driver populations, which should be investigated further.
Revealing the unspoken: drunk driving behavior of Vietnamese students.
Tran TKM, Goto D, Manh Hung N … +1 more , Tran D
Traffic Inj Prev · 2026 Feb · PMID 41637564 · Publisher ↗
OBJECTIVE: The increasing rate of alcohol related traffic accidents, especially among adolescents, has become a major safety concern. While numerous studies have examined drunk driving among young people using direct que...
OBJECTIVE: The increasing rate of alcohol related traffic accidents, especially among adolescents, has become a major safety concern. While numerous studies have examined drunk driving among young people using direct questioning methods, such behaviors are often viewed as sensitive, increasing the risk of underreporting. This work is to examine the actual prevalence of misreporting regarding drunk driving and riding with a drunk driver, while accounting for potential misreporting bias. METHOD: The study employed a list experiment, an indirect questioning technique, on a sample of 615 undergraduates in Vietnam. RESULTS: Our findings reveal that the prevalence of drunk driving among male students is nearly three times higher when measured using the list experiment compared to direct questioning. Furthermore, students from single-parent or parentless households exhibit considerably greater misreporting bias and a higher prevalence of drunk driving than those from families with both parents. CONCLUSION: This study suggests that list experiments are an appropriate and effective method for estimating the prevalence of sensitive alcohol consumption-related travel behaviors.
A hybrid decision support system, focused on in-depth road risk assessment by integrating iRAP and fuzzy AHP methods.
Gherbi FZ, Oulha R, Hamadouche MA … +1 more , Lallam M
Traffic Inj Prev · 2026 Feb · PMID 41637557 · Publisher ↗
OBJECTIVE: This study aims to assess and model road risk by addressing the multidimensional complexity of road accidents, which result from the interaction of infrastructure characteristics, traffic composition, driver b...
OBJECTIVE: This study aims to assess and model road risk by addressing the multidimensional complexity of road accidents, which result from the interaction of infrastructure characteristics, traffic composition, driver behavior, and environmental conditions. The main objective is to develop an integrated analytical framework capable of identifying high-risk road segments, quantifying their overall risk levels, and prioritizing intervention strategies to enhance safety. The proposed approach seeks to overcome the limitations of traditional geometry-based methods by providing an adaptive decision-support tool that better reflects the contextual and operational realities of road networks. METHODS: This study employs a hybrid methodology combining the International Road Assessment Program (iRAP) and the Fuzzy AHP (Fuzzy AHP) to analyze to 999 road sections (100 m each) located in various Algerian regions: Mascara, Ghazaouet, Djebahia, and Blida. This combination enables the development of a simplified road risk assessment model for each 100-meter section. The steps include identifying the factors affecting safety, building a hierarchical model, applying the fuzzy method to evaluate the main and sub-factors and calculating their weights, and finally classifying the sections from least to most risky based on a star scale derived from the iRAP methodology. RESULTS: The analysis revealed that infrastructure-related factors exert the strongest influence on road risk, followed by human and traffic-related variables, while environmental factors showed the least impact. The hybrid model corrected several safety overestimations observed in iRAP by incorporating behavioral and dynamic parameters. A strong correlation between the hybrid model's risk levels and actual accident data confirmed its predictive reliability and robustness. CONCLUSION: The proposed hybrid methodology offers a reliable and adaptable tool for road safety assessment. By integrating infrastructure, human, traffic, and environmental factors, it provides a realistic view of road risk. It also suggests targeted measures-such as lane separation, improved visibility, and speed management to enhance safety, especially in developing regions with limited data.
Modeling and analysis of influencing factors of primary accidents to secondary accidents based on structural equation model.
Gong X, Liu R, Tang J
Traffic Inj Prev · 2026 Jan · PMID 41569228 · Publisher ↗
OBJECTIVES: Secondary accidents generate a serious threat to road safety and traffic efficiency due to their compounded impact on congestion and emergency response. Due to the higher risks and difficulty for prevention o...
OBJECTIVES: Secondary accidents generate a serious threat to road safety and traffic efficiency due to their compounded impact on congestion and emergency response. Due to the higher risks and difficulty for prevention of secondary accidents, the study aims to compose a method to mining the potential relationship of the occurrence of secondary accidents. METHODS: This study quantifies the influence of primary accidents factors on the secondary accidents occurrence by using Structural Equation Model (SEM). Based on the Statewide Integrated Traffic Records System (SWITRS) accident record and roadway geometry data from OpenStreetMap (OSM), secondary accidents are first identified using a fixed spatiotemporal threshold approach. Then, a multidimensional dataset is constructed by integrating variables related to driving behavior, environmental conditions, and roadway features. SEM includes a measurement model for primary accidents, which defines five latent variables: driving behavior, environment, spatiotemporal features, primary accident features, and high-risk indicators. The measurement model for secondary accidents is developed as well, which defines the latent variable representing secondary accidents. Finally, a structural model is developed to analyze the relationships among the latent variables across the two measurement models. RESULTS: The research results show that "PCF_viol_category" (0.9767), weather (0.7657), direction (0.9781), MVIW (0.847) and "State_hwy_ind" (0.5968) of primary accidents, have the highest positive impact value among the factors. While collision time (0.9947), weather (0.9018) and curve (0.3355) are positively significant among the factors in secondary accidents. Besides, high-risk (-0.4995) and driving behavior (-0.1222) have negative impact value, indicating that proactive interventions in high-risk scenarios and safe driving behaviors effectively mitigate the probability of secondary accidents. CONCLUSIONS: The findings highlight the critical roles of driving behavior regulation and freeway accident management in preventing the escalation of secondary accidents. Environmental conditions and roadway geometric features substantially influence secondary accident risk, underscoring the need for targeted safety management strategies.
Relationship between occupant compartment intrusion and injury by crash counterpart relevant to MASH requirements.
Mattos G, Kenney T, Friedman K … +1 more , Paver J
Traffic Inj Prev · 2026 Jan · PMID 41569227 · Publisher ↗
OBJECTIVE: This study used recent crash data to quantify the correlation between MASH intrusion limits and injury outcomes for crashes with roadside hardware, vehicles, or fixed objects. The study also provides additiona...
OBJECTIVE: This study used recent crash data to quantify the correlation between MASH intrusion limits and injury outcomes for crashes with roadside hardware, vehicles, or fixed objects. The study also provides additional descriptive statistics and characteristics about the intrusion environment, resulting injuries, and the associated crash environment. METHODS: This study uses the latest crash data from NHTSA's CISS database to investigate the relationship and potential correlations between injuries and vehicle intrusion disaggregated by crash counterpart: roadside hardware, vehicles, or fixed objects. Statistical analyses are used to quantify the parameter effects and their significance in predicting injury as a direct result of intrusion into the occupant compartment. The data are summarized at the occupant level and selected crashes are analyzed in detail to highlight potential injury mechanisms. RESULTS: Occupant compartment intrusion in crashes is relatively rare, generally, and even more rare in crashes with roadside hardware. For crashes with roadside hardware that resulted in occupant compartment intrusion, the crash was most often more severe than the test-level 3 conditions required by MASH. While the severity of injury for injured occupants in roadside hardware crashes was lower compared to those in fixed object and vehicle crashes, the rate of lower extremity injuries was much higher. Occupants that experienced intrusion at their seat location that exceeded MASH thresholds were more likely (OR = 16.02 to 24.57) to suffer a severe or greater injury than those that did not. CONCLUSIONS: Roadside hardware crashes are distinct from fixed-object and vehicle-to-vehicle crashes with regard to occupant compartment intrusion and injury outcomes. The MASH intrusion requirements are a significant predictor of injury outcome. These results can be used to enhance and improve the requirements of MASH and may lead to additional and more robust occupant injury metrics.
Risk-rule filtered LSTM-dueling DQN for autonomous lane change decision.
Chen Z, Zhu K
Traffic Inj Prev · 2026 Jan · PMID 41556779 · Publisher ↗
OBJECTIVE: Autonomous vehicle lane-change decision making has long been a prominent research topic in intelligent transportation systems. To enhance both the safety and efficiency of lane changes in dynamic traffic envir...
OBJECTIVE: Autonomous vehicle lane-change decision making has long been a prominent research topic in intelligent transportation systems. To enhance both the safety and efficiency of lane changes in dynamic traffic environments, we propose a Risk-Rule Filtered Long Short-Term Memory Dueling Deep Q-Network (LSTM-DDQN) method for autonomous lane-change decision making under complex traffic scenarios. METHODS: This approach achieves intelligent decision making by integrating a risk-rule filter with a reinforcement-learning framework. First, the model employs a Time-to-Collision (TTC)-based risk filtering mechanism to select target vehicles that pose potential collision risks to the ego vehicle as the basis for decision making. Next, the LSTM network processes the filtered observation sequences to extract dynamic traffic-context information with temporal features. These features are then input into a Dueling DQN architecture to separately estimate the state-value function and the action-advantage function, thereby optimizing the final lane-change action selection. RESULTS: Compared to a standard Dueling DQN baseline implemented under identical simulation settings (same architecture, hyper-parameters, and training protocol but without the LSTM temporal module and TTC-based risk filter), the proposed LSTM-DDQN achieves a 10.9% increase in normalized average single-step reward. CONCLUSIONS: The proposed model was validated on the Simulation of Urban Mobility (SUMO) platform, underscoring its superior performance in improving lane-change safety.
About
- Frequency
- Sun
- Papers found
- 200
- RSS feed
- Subscribe