Modelling Road Crash Fatalities in Albania: A Cross-Sectional Study of Risk Factors Using Advanced Count Regression

Modelling Road Crash Fatalities in Albania: A Cross-Sectional Study of Risk Factors Using Advanced Count Regression | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Modelling Road Crash Fatalities in Albania: A Cross-Sectional Study of Risk Factors Using Advanced Count Regression Raimonda Dervishi, Fabiana Çullhaj, Agbata Benedict Celestine, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7860402/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract The global escalation in road crash has led to a substantial public health burden, with annual fatalities reaching millions worldwide and causing extensive economic and social disruption. While existing research has largely focused on predicting the likelihood of road crash fatalities as a classification problem, relatively few studies have investigated the underlying relationships among the complex factors contributing to such fatalities. This study aims to identify the significant determinants of road crash fatalities in Albania by employing advanced modelling approaches. The analysis is based on official statistics from the Albanian Institute of Statistics ( INSTAT ) covering seven years from 2018 to 2024. During this period, a total of 13,203 individuals were involved in road traffic accidents, with 1,349 fatalities recorded. Although the overall number of accidents has decreased, the fatality rate has shown a concerning upward trend.To better understand the dynamics of road crash fatalities ( RAF s), this study examines the impact of driver behaviours, time factors, and road characteristics. Two regression approaches Poisson–Lognormal ( PL ) and Conway–Maxwell–Poisson ( CMP ) regression models are applied to model the number of fatalities concerning these key factors. These models are particularly suited to handling the over–dispersed nature of the data, in contrast to the traditional Poisson regression model, which assumes equi–dispersion. The CMP-Poisson model outperformed both, offering better model fit and more robust estimation under dispersion, validating its application in traffic fatality research with count data.By quantifying the effects of these risk factors, this study provides valuable insights for the development of targeted interventions and policy strategies aimed at reducing road crash fatalities in Albania. Road crash fatalities (RCF) Statistical modelling Driver behaviours CMP regression Over–dispersed data 1. Introduction Road traffic accidents result in an estimated 20 to 50 million injuries globally each year, alongside approximately 1.3 million fatalities [ 1 ]. Alarmingly, nearly 90% of these deaths occur in low- and middle-income countries, underscoring a major global health and development disparity [ 1 ]. From 2015 to 2030, road injuries both fatal and non-fatal are projected to cost the global economy around USD 1.8 trillion, highlighting the immense financial burden of traffic-related harm ([ 1 , 2 ]). Road traffic accidents remain a major threat to public safety across all nations and continue to challenge global injury prevention strategies [ 3 ]. In this context, it has been argued that developing nations should adopt infrastructure and safety frameworks modelled after those in high-income countries [ 4 ]. Indeed, traffic-related deaths remain a leading cause of mortality worldwide [ 5 , 6 ]. At the European level, the European Commission has emphasized safe and efficient mobility as part of its transport integration and sustainability agenda, with the Vision Zero strategy aiming to eliminate road fatalities by 2030. Albania has adopted several reforms in line with EU standards, including stricter traffic law enforcement, harmonized penalties, enhanced driver training, adoption of vehicle safety technologies, and modernization of infrastructure. The European Commission has also introduced directives to strengthen road safety management through standardized audits [ 7 , 8 ]. Despite a decline in accident cases in the past decade, road safety remains a concern in Albania [ 9 ]. The country is far from achieving the WHO target of reducing road crash fatalities by 50% by 2030 [ 10 ]. Data show that driver violations account for the largest share of accidents with loss of life, followed by pedestrian non-compliance [ 11 ]. This trend is especially concerning given Albania’s decreasing population alongside growing motorization. Road traffic accidents are driven by complex and interrelated factors, making risk factor analysis essential for targeted prevention [ 12 , 13 ]. While progress has been made in accident modelling, further research is needed [ 14 ], particularly on risk factors and mitigation strategies, which remain a priority in traffic safety research [ 15 ]. In Albania, studies indicate progress but also highlight persistent challenges in improving road safety [ 10 , 16 ]. Authorities stress the importance of education, infrastructure upgrades, and consistent enforcement, supported by investment in safety-focused transport strategies [ 17 ]. According to the WHO, road traffic injuries are a leading cause of death among young people worldwide, and Albania is no exception. National road networks, driver behaviour, and environmental conditions all contribute to accident frequency. A cohesive policy framework emphasizing safety, sustainability, governance, and integration is required, alongside both domestic reforms and international cooperation. A range of factors including driver characteristics, road conditions, and time affect accident severity [ 18 ]. Driver errors, such as speeding, substance use, and failure to use protective measures, are among the most critical [ 19 – 22 ]. Other significant factors include age, driving time, and location [ 20 ]. However, existing research on road safety in Albania is limited, with few studies using advanced regression approaches such as Poisson–Lognormal (PL) or Conway–Maxwell–Poisson (CMP) models. Prior work has applied Bayesian Poisson models to predict risk levels based on time, location, speeding, and infrastructure [ 23 ], while others examined specific factors like driver behavior or weather conditions [ 9 , 24 ]. Yet, a comprehensive analysis considering multiple interacting factors remains lacking. The present study addresses this gap by applying PL and CMP regression models to road crash fatality data from Albania (2018–2024). Using official statistics from the Institute of Statistics (INSTAT), we investigate how road conditions, driver behaviour, and temporal factors contribute to fatal accidents. This study contributes to the growing literature on road safety in developing contexts and aims to provide evidence that can inform targeted interventions and policy decisions for improving traffic safety outcomes in Albania. 2. Related Works Accurate modeling and prediction of road accidents are fundamental to understanding the causal mechanisms behind crash severity and the effectiveness of safety interventions. Traffic accident prediction models are commonly employed to examine relationships between crash severity outcomes and contributing variables, including driver behavior, vehicle characteristics, road and time factors, and environmental and traffic conditions. Preventative strategies targeting road fatalities and injuries often require a multidisciplinary approach involving transport agencies, police, public health authorities, and educational institutions. These interventions range from infrastructure redesign and vehicle safety improvements to post-crash care enhancement and behavioral reforms through education and enforcement [ 26 ]. Over the past few decades, a broad range of statistical and regression-based approaches have been developed to model crash data. Among them, the Poisson, Poisson-Lognormal, Negative Binomial, and Conway–Maxwell–Poisson ( CMP ) regression models are commonly employed due to their adaptability to real-world data characteristics, particularly the count nature of accident data [ 27 ]. The traditional Poisson regression model is widely used because of its simplicity and interpretability. It assumes that the mean and variance of crash frequency are equal (equi–dispersion). However, real-world fatality datasets often violate this assumption due to over–dispersion or under–dispersion, which limits the applicability of the Poisson model in many contexts [ 28 ]. To address over–dispersion where the variance exceeds the mean researchers have utilized models such as the Conway-Maxwell-Poisson and Poisson-Lognormal regressions. These models introduce additional parameters to account for unobserved heterogeneity in accident data. The study [ 29 ] introduced the CMP regression model as a flexible alternative capable of accommodating both over- and under-dispersed data, a limitation for which neither the traditional Poisson nor the Negative Binomial models are well suited. Several empirical studies have applied Poisson models to real-world crash datasets. As shown in [ 30 ], the frequency of accident involvement using Poisson-family models, identifying demographic characteristics such as age and gender as significant predictors. Their study highlighted the impact of geometric and traffic-related features on accident occurrence. Similarly, the authors in [ 31 ] advocated for Poisson regression in highway safety research, emphasizing its statistical advantages over linear models in modeling accident frequencies. In their analysis of traffic accident severity in Saudi Arabia, [ 32 , 33 ] employed both Poisson and logistic regression models to assess the relative impact of various risk factors on crash outcomes. Further improvement was introduced by [ 34 ], who developed a mixed generalized ordered response model to analyze injury severity among pedestrians and cyclists. Although this study focused on vulnerable road users, its methodology applies to broader crash severity analysis. More recently, [ 35 ] employed a random parameter logit model incorporating heterogeneity in both the mean and variance to investigate driver injury severity in Florida work–zone crashes, differentiating the influence of geometric versus non-geometric (mainly behavioral) factors. The model results encompassed an array of factors, including spatial characteristics, vehicle characteristics, environmental conditions, geometric features, crash attributes, traffic patterns, and driver traits. These factors were intricately connected with human factors, particularly driver behaviors and characteristics, playing a pivotal role in work-zone crashes tied to geometric and non-geometric attributes. A significant contribution to flexible modeling is the Conway–Maxwell–Poisson regression, which offers a robust framework for handling data characterized by over–, under–, or equi–dispersion. In a comparative study [ 36 ], the model was evaluated using both simulated and real accident data from Canada and South Korea. Their findings demonstrated the model’s capability to provide reliable estimates across various dispersion scenarios, although limitations were noted for small sample sizes or datasets with low mean crash frequencies. Nevertheless, the CMP model’s computational burden was found to be manageable, and its predictive performance was comparable to or better than competing models such as the Poisson–Gamma, making it a viable option for applied crash data analysis in transportation safety research. A study [ 37 ]applied the Poisson, CMP, and Zero–Inflated Conway-Maxwell Poisson ( ZICMP ) models to analyze road crash fatalities in Thailand, where the data showed a rare pattern of under–dispersion with excessive zero counts. Their study indicated that although the ZICMP model marginally outperformed the standard CMP model, the necessity for zero–inflation modeling depends on the specific research question rather than solely on the data structure. Road and environmental conditions, as well as seasonal patterns (such as during the Songkran festival), were identified as significant contributors to fatal crashes. Another study [ 38 ] conducted a comparative analysis using Poisson, Poisson–Lognormal, and Negative Binomial regression models to examine road crash data from North Central Nigeria, evaluating the contributions of key factors Driver Error, Faulty Vehicle, and Road Condition to fatalities. Their analysis determined that the Poisson-Lognormal model offered the best statistical fit based on Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), and pseudo R-squared values. The study found that driver error was the most significant predictor of fatalities, contributing to a 2.17% increase in fatality risk per unit increase. Faulty vehicles were also substantial, whereas road condition did not exhibit a statistically meaningful effect. Similarly, studies by [ 28 ] and [ 29 ] further validated the usefulness of Poisson–family models in regions with high accident data rates. Moreover, [ 28 ] examined accident data in a subregion of Ghana, confirming the presence of over–dispersion in fatality counts and advocating for the use of over–dispersion–tolerant models. 3. Trends in Road crash Fatalities in Albania Road crash fatalities result from a multifaceted interplay of human, environmental, vehicular, and institutional factors. Understanding these causes is essential for designing effective interventions aimed at reducing mortality on the roads. Although accidents may occur due to unforeseeable events, the literature consistently identifies a set of recurring—and often preventable causes of fatal outcomes in road accidents [ 39 , 40 ]. While road accidents can arise from multiple causes, human behavior consistently emerges as the most influential, responsible for over half of collisions on its own. When coupled with factors like road conditions or vehicle issues, human error is involved in the vast majority of cases, reaching nearly 93% [ 27 ]. Speeding continues to be a primary cause of fatal crashes, as excessive velocity reduces a driver’s ability to respond to hazards and increases the severity of impacts. The World Health Organization (2023) reports that speeding contributes to approximately 30% of road traffic deaths globally [ 40 ]. In Albania, speed–related incidents are particularly prevalent on rural and regional roads, where enforcement is inconsistent and signage is often inadequate. Driving under the influence of alcohol or drugs is another major factor. Substance impairment diminishes cognitive and motor functions, heightening the risk of fatal outcomes. The European Transport Safety Council (2023) highlights a troubling pattern of alcohol-related road deaths in Southeast Europe, including Albania, where young male drivers are disproportionately involved in fatal incidents. Additionally, distracted driving stemming from mobile phone use, in-vehicle technologies, or fatigue has become an increasingly prominent risk factor, especially in urban environments where multitasking is common [ 41 ]. Institutional and systemic limitations significantly contribute to the persistence of elevated road traffic fatality rates. In Albania, although legal frameworks exist to regulate driver behaviour and enhance traffic safety, enforcement remains inconsistent, particularly in rural and peri-urban areas where oversight is weaker [ 42 ]. Weak institutional coordination, limited capacity for traffic law enforcement, and insufficient investment in road safety infrastructure further undermine efforts to mitigate risk. Moreover, deficiencies in driver education and public awareness campaigns reduce the effectiveness of existing policies. Of critical concern is the inadequacy of post-crash emergency response systems. Delays in medical intervention following traffic collisions often due to logistical constraints or poor coordination between emergency services frequently lead to fatalities that might otherwise have been preventable [ 42 , 43 ]. These institutional shortcomings underscore the need for a more integrated, well–funded, and professionally managed road safety system. Over the recent years, Albania has witnessed fluctuating yet persistently concerning trends in road crash fatalities. While certain years have reflected modest declines in fatality rates, the broader trajectory highlights the ongoing difficulty in achieving sustained and meaningful reductions. This challenge is driven by a complex combination of factors, including inadequate and underdeveloped infrastructure, rapid urban expansion, increased motorization, and widespread risky road user behaviours. Together, these elements continue to undermine national road safety efforts and pose significant barriers to progress. Nevertheless, some infrastructure-focused initiatives have shown measurable results. For instance, a World Bank-supported program implemented through the International Road Assessment Programme (IRAP) led to a 23% reduction in road deaths on primary and secondary road corridors in Albania saving an estimated 68 lives annually by improving road design and safety features [ 42 ]. This demonstrates that, despite structural and behavioural challenges, targeted engineering interventions can yield meaningful progress toward fatality reduction goals. Failure to use protective equipment such as seatbelts and helmets further exacerbates fatality risks. According to WHO (2023), proper seatbelt use can reduce the risk of death by 45–50% for front-seat occupants and up to 75% for those in the rear. Despite awareness campaigns in Albania, usage remains inconsistent, particularly among motorcyclists and drivers in rural areas [ 40 ]. National statistics indicate that fatality rates declined slightly between 2018 and 2020, likely reflecting the effects of early enforcement initiatives and public safety campaigns. However, this progress was reversed in 2023 and 2024, with a notable increase in accident frequency. This recent rise appears to correlate with Albania’s growing vehicle fleet and the resurgence of high-risk driving behaviours. According to WHO (2023), Albania’s road traffic death rate stood at approximately 10.8 deaths per 100,000 population in 2021[ 40 ]. INSTAT (2024) reported that 86.4% of road accidents in 2023 were attributed to driver violations, with drivers aged 25–34 comprising the most affected demographic group. Despite some recent improvements in pedestrian safety evidenced by a reduced share of pedestrian-involved accidents in 2024 the broader picture remains dominated by human factors and increased motorization. Disaggregated data further reveal shifting patterns in the distribution of serious injuries across user groups. Between 2018 and 2024, pedestrians accounted for 25.3% of serious injuries, followed closely by passengers (25.2%) and motorcyclists (23.7%) [ 25 ]. Previous empirical studies underscore the direct correlation between vehicle volume and fatality rates [ 44 , 45 ]. Environmental and/or road factors have been consistently identified as statistically significant contributors to road crash fatalities in Albania and beyond [ 37 , 54 , 55 ]. Furthermore, variables such as time of day, seasonal trends, road type, and intersection design are increasingly acknowledged by both global and national studies as influential determinants of crash severity [ 9 , 16 , 24 ]. A nuanced understanding of these environmental and roadway dimensions is therefore essential for the development of data-driven safety interventions and infrastructure planning. Between 2018 and 2024, Albania recorded a total of 16,192 road traffic accidents, involving an estimated 24,020 individuals. These incidents resulted in 2,199 fatalities and 19,622 injuries. However, it is important to note that official figures may underestimate the true burden due to underreporting, especially in cases where individuals fled the scene or the incident was not formally documented [ 25 ]. A comprehensive summary of fatality characteristics by key variables is provided in Table 1 . Table 1 Summary of fatality characteristics by key variables, 2018–2024. Variables No of Fatalities (%) Variables No of Fatalities (%) Drivers experience Seasons of accidents Without License 105 (7.78%) Autumn 338 (25.1%) 0–3 years 238 (17.64%) Spring 308 (22.8%) 3–6 years 247 (18.31%) Summer 356 (26.4%) 6–9 years 219 (16.23%) Winter 347 (25.7%) Above 9 505 (37.43%) Time of accidents Missing 35 (2.59%) 00:00–06:00 149 (11.0%) Sex of drivers 06:00–08:00 70 (5.2%) Male 1047 (77.6%) 08:00–12:00 192 (14.2%) Female 267 (19.8%) 12:00–14:00 123 (9.1% ) Missing 35 (2.6%) 14:00–17:00 221 (16.4%) The age group of drivers 17:00–19:00 266 (19.7%) 0–24 246 (18.2%) 19:00–00:00 328 (24.3%) 25–34 510 (37.8%) Road geometry 35–44 221 (16.4%) Straight road 968 (71.8%) 45–59 180 (13.3%) Curved 216 (16.0%) Above 59 157 (11.6%) T-Intersection 104 (7.7%) Missing 35 (2.6%) Roundabout 32 (2.4%) Day of the week of accidents Other 29 (2.1%) Monday 199 (14.8%) Road type Tuesday 208 (15.4%) Highway 80 (5.9%) Wednesday 181 (13.4%) Interurban Road 613 (45.4%) Thursday 182 (13.5%) Urban Road 446 (33.1%) Friday 208 (15.4%) Local Road 210 (15.6%) Saturday 182 (13.5%) Sunday 189 (14.0%) 4. Modelling Road Crash Fatalities Data Over the years, numerous data-related and methodological issues have been highlighted in the literature concerning accident frequency analysis. A core objective of model-building techniques in this context is to develop an accurate and parsimonious representation of the relationship between accident outcomes—typically measured as count data and a set of explanatory variables [ 46 ]. As emphasized in recent studies, crash prediction plays a fundamental role in the broader road safety management process, facilitating proactive interventions and infrastructure planning [ 47 ]. Given that the dependent variable in this study namely, the number of road crash fatalities is discrete and demonstrates under–dispersion (the variance exceeds the mean), traditional Poisson regression may not suffice. To address this, two more flexible count regression models were applied: the Poisson-Lognormal ( PL ) and the Conway–Maxwell–Poisson ( CMP ) models. These models allow for a more accurate estimation of significant risk factors by accounting for unobserved heterogeneity and the potential for both over–dispersion and under–dispersion in the data. 4.1 Poisson-Lognormal Regression Model The PL model extends the standard Poisson framework by incorporating a log–normally distributed random effect to account for unobserved heterogeneity. This approach is effective in modelling over–dispersed count data and estimating how each covariate influences the expected number of fatalities. This model assumes a lognormal distribution for the error term, offering enhanced flexibility in modelling variance heterogeneity and multilevel data structures [ 46 , 47 ]. The model specification is: Y i ~ Poisson ( λ i ), log( λ i ) = X i T β + ℇ i , ℇ i ~ N (0, σ 2 ) where Y i denotes the fatality count for accident i, X i is the vector of covariates, β the regression coefficients, and ℇ i ​ the normally distributed random effect. However, a significant limitation of the Poisson-lognormal model is the lack of a closed-form expression for its likelihood function, complicating parameter estimation and computational implementation [ 47 ]. Despite this, it remains a valuable tool when over–dispersion or complex random effects are present. A few researchers have suggested the use of the Poisson-lognormal regression for modelling accident data [ 52 ]. 4.2 Conway-Maxwell Poisson Regression Model Standard Poisson regression assumes equi–dispersion, where the mean equals the variance a condition rarely met in real-world crash data. The Conway–Maxwell–Poisson ( CMP ) distribution generalizes the Poisson to accommodate both over–dispersion and under–dispersion, making it well-suited for traffic crash modelling [ 48 ]. The CMP–Poisson regression framework offers greater flexibility by incorporating a dispersion parameter that adjusts variance independently from the mean. This model is particularly useful when accident data exhibit under–dispersion, a scenario where traditional Poisson models perform poorly [ 49 ]. It generalizes the Poisson distribution with an additional dispersion parameter ν , allowing for flexible modelling of both over–dispersion and under dispersion: Y i ~ CMP( λ i , ν ), log( λ i ) = X i T β Nevertheless, estimating CMP regression models can be computationally intensive, and small sample sizes or low mean counts can adversely affect parameter stability [ 50 , 51 ]. Such challenges have limited widespread adoption despite the theoretical advantages. Some researchers have suggested the Conway–Maxwell–Poissonmethod for modelling accident data [ 29 , 37 , 53 ] 4.3Assessment of Model Fit and Selection Criteria To assess model adequacy and comparative performance, we employed three standard statistical metrics: the Akaike Information Criterion ( AIC ), the Bayesian Information Criterion ( BIC ), and the log-likelihood value. These measures provide a consistent framework for evaluating the trade-off between model fit and complexity. The Akaike Information Criterion, proposed by [ 56 ],estimates the relative quality of models by penalizing the number of parameters while prioritizing goodness-of-fit.It is defined as: AIC = –Log L + 2k Where Log L is the maximized log-likelihood and k is the number of model parameters. The Bayesian Information Criterion, introduced by [ 57 ], imposes a stronger penalty for model complexity based on sample size n : BIC = – Log L + k Log n This renders the BIC criterion more conservative than AIC when selecting among competing models, as it imposes a stronger penalty for model complexity. While AIC tends to prioritize predictive accuracy, BIC emphasizes parsimony, especially as sample size increases [ 58 ]. The log-likelihood measures the probability of the observed data given the model parameters. A higher log-likelihood value indicates a better fit [ 58 ]. Formally, it is expressed as: LogL(θ∣y i ) = ∑ ​log f(y i ∣θ) Where y i are observed values, and f(y i ​∣θ) is the likelihood function given parameters θ. 4.4Computational Environment All statistical analyses were performed using the R statistical software environment [ 59 ]. A variety of specialized R packages were employed to model fatal traffic accidents using count data frameworks. The standard Poisson regression model was estimated via the glm() function. To address over-dispersion and accommodate non–equidispersed count outcomes, we implemented the Conway–Maxwell–Poisson (CMP-Poisson) model using the glmmTMB package [ 60 ]. For the Poisson-lognormal model, we applied a generalized linear mixed model with random intercepts using the same framework, allowing us to model latent heterogeneity between groups. Model selection and comparative performance were assessed using the log-likelihood value, Akaike Information Criterion ( AIC ), and Bayesian Information Criterion ( BIC ). These statistics were calculated using functions from the performance and insight packages [ 61 ]. Data management and visualization were conducted using dplyr and ggplot2 [ 62 ], ensuring a reproducible and transparent workflow. 5. Estimated Model Results This study employs a cross-sectional design, analyzing all road traffic accident cases and their associated fatalities over eight years. Each observation in the dataset represents a single traffic accident, with the primary outcome variable being the count of fatalities per accident. Independent variables are grouped into three main categories: driver,time, and road factors. Driver factors consisted of driver experience (categorized as without license, 0–3 years, 3–6 years, 6–9 years, and above 9 years), driver age groups (0–24, 25–34, 35–44, 45–59, and above 59), and driver sex (male, female). Time factors included time of accident (divided into time intervals ranging from 00:00 to 24:00 hours), day of the week, and season of the accident (spring, summer, autumn, and winter). Road factors comprised road type (highway, interurban roadway, urban roadway, and local roads) and road geometry (straight road, curved, roundabout, T-intersection, and other). The dependent variable in this study was defined as the number of fatalities recorded within each observational group. To adjust for differing levels of exposure across groups, the total number of road traffic accidents was incorporated as an offset term in the regression models, using its natural logarithm. This approach allowed for modelling fatality risk relative to accident frequency. The explanatory variables were selected based on their established relevance in the road safety literature and the availability of corresponding data from the Albanian Institute of Statistics ( INSTAT ). These covariates were included in the modelling framework to assess their contribution to variations in fatality counts per accident. A summary of the main findings is provided in Table 2 . Table 2 Coefficients(p-value) from the Poisson, PL,and CMP regression models. Variables Poisson PL CMP Drivers' experience category Intercept -3.3844 (< 0.0001) -3.3845 (< 0.0001) -3.3814 (< 0.0001) Without License + 0.1063 (0.36416 ) + 0.1060 (< 0.0001) + 0.0069 (< 0.0001) 0–3 years + 0.2744 (< 0.0001) + 0.2744 (< 0.0001) + 0.2747 (< 0.0001) 3–6 years -0.1769 (0.0514) -0.1770 (0.0513 ) -0.1777 (< 0.0001) 6–9 years + 0.0792 (0.3971) + 0.0793 (0.3970) + 0.0791 (< 0.0001) Above 9 + 0.0961 (0.2212 ) + 0.0962 (0.2213) + 0.0961 (< 0.0001) Missing -0.6374 (0.0042) -0.6374 (0.3642) -0.6559 (< 0.0001) Sex of drivers category Female + 1.3083 (< 0.0001) + 1.3081 (< 0.0001) + 0.3516 (< 0.0001) Male + 1.7007 (< 0.0001) + 1.7006 (< 0.0001) + 1.6994 (< 0.0001) Missing + 0.3963 (0.0274) + 0.3962 (0.0275) + 0.3739 (< 0.0001) The age group of the driver category 0–24 + 0.3543 (< 0.0001) + 0.3544 (< 0.0001) + 0.3521 (< 0.0001) 25–34 + 0.3851 (< 0.0001) + 0.3850 (< 0.0001) + 0.3824 (< 0.0001) 35–44 -0.2084 (< 0.0001) -0.2084 (0.0335) -0.2124 (< 0.0001) 45–59 -0.1097 (0.2364) -0.1099 (0.2364) -0.1140 (< 0.0001) Above 59 + 0.2049 (< 0.0001) + 0.2049 (0.0448) + 0.2022 (< 0.0001) Missing -0.5575 (< 0.0001) -0.5576 (< 0.0001) -0.5634 (< 0.0001) Day of the week of the accident category Monday -0.0572 (0.5639 ) -0.0582 (0.5639) -0.0575 (< 0.0001) Tuesday + 0.0335 (0.7320) + 0.0337 (0.7320) + 0.0313 (< 0.0001) Wednesday -0.1040 (0.3060) -0.1044 (0.3060) -0.1055 (< 0.0001) Thursday -0.1161 (0.2523) -0.1160 (0.2523) -0.1194 (< 0.0001) Friday + 0.2255 (0.0202) + 0.2256 (0.0202) + 0.2250 (< 0.0001) Saturday -0.0964 (0.3422) -0.0965 (0.3422) -0.0987 (< 0.0001) Sunday -0.0397 (0.6911) -0.0399 (0.6911) -0.0426 (< 0.0001) Seasons of accidents category Autumn + 0.2947 (< 0.0001) + 0.2948 (< 0.0001) + 0.2950 (< 0.0001) Spring + 0.0430 (0.5845) + 0.0431 (0.5845) + 0.0431 (< 0.0001) Summer -0.0414 (0.5809) -0.0419 (0.5809) -0.0416 (< 0.0001) Winter + 0.0716 (0.3465) + 0.0719(0.3465) + 0.0712 (< 0.0001) Time of accidents category 00:00–06:00 -0.3293 (< 0.0001) -0.3294 (0.0019 ) -0.3258 (< 0.0001) 06:00–08:00 -0.7385 (< 0.0001) -0.7386 (< 0.0001) -0.7358 (< 0.0001) 08:00–12:00 -0.9838 (< 0.0001) -0.9838 (< 0.0001) -0.9805 (< 0.0001) 12:00–14:00 -0.9652 (< 0.0001) -0.9653 (< 0.0001) -0.9622 (< 0.0001) 14:00–17:00 -0.7248 (< 0.0001) -0.7249 (< 0.0001) -0.7213 (< 0.0001) 17:00–19:00 -0.3553 (< 0.0001) -0.3555 (< 0.0001) -0.3521(< 0.0001) 19:00–00:00 -0.4401 (< 0.0001) -0.4405 (< 0.0001) -0.4378 (< 0.0001) Road geometry category Straight road -0.2596 (< 0.0001) -0.2597 (< 0.0001) -0.2596 (< 0.0001) T-Intersection -0.8992 (< 0.0001) -0.8993 (< 0.0001) -0.9000 (< 0.0001) Roundabout -0.1136 (0.5485) -0.1135 (0.5486) -0.1150 (< 0.0001) Other + 0.3857 (0.0511) + 0.3858 (0.0513) + 0.3876 (< 0.0001) Road type category Highway -0.4626 (< 0.0001) -0.4627 (< 0.0001) -0.4647 (< 0.0001) Interurban Road -0.3442 (< 0.0001) -0.3443 (< 0.0001) -0.3465 (< 0.0001) Urban Road -1.2375 (< 0.0001) -1.2375 (< 0.0001) -1.2401 (< 0.0001) Local Road -0.5056 (< 0.0001) -0.5055 (< 0.0001) -0.5087 (< 0.0001) Table 3 Estimated parameters (p-value), Log-Likelihood value, AIC, and BIC criteria for the advanced regression model. Poisson PL CMP Β -3.3844 (p < 0.0001) -3.3845 (p < 0.0001) -3.3814 (p < 0.0001) Θ NA ≈ 0 0.120 AIC 381.26 383.30 137.3 BIC NA 458 212 Log-Likelihood Value -190.6 -148.6 -25.6 To evaluate the adequacy of the fitted regression models, three widely used model selection criteria were employed: the Akaike Information Criterion (AIC), the Bayesian Information Criterion (BIC), and the log-likelihood value. These indicators provide a unified framework for comparing models with differing numbers of parameters by balancing model complexity with goodness–of–fit. As shown in Table 3 , the standard Poisson regression model yielded an AIC of 381.3 and a log–likelihood of − 1 48.6, indicating poor fit due to over–dispersion. The Poisson-lognormal model, although slightly more flexible, failed to significantly improve the fit, as evidenced by negligible variance (≈ 0) in the random intercept. In contrast, the standardCOM-Poisson model outperformed all others, with an AIC of 137.3, BIC of 212.0, and log-likelihood of − 25.6. Moreover, its estimated dispersion parameter (θ = 0.0019) confirmed its capacity to address severe over–dispersion in the data. These findings support the use of the CMP–Poisson model as the most parsimonious and best–fitting model for analyzing over–dispersed fatality count data in road traffic research. 5.1 Discussion This study examined fatal road traffic accidents in Albania between 2018 and 2024 using advanced count regression models to identify significant risk factors and evaluate model performance under various dispersion assumptions. Key demographic, temporal, and environmental variables were found to be significantly associated with fatal outcomes. Notably, over–dispersion in the count data highlighted the necessity of employing flexible models such as the Conway–Maxwell–Poisson (CMP–Poisson), which outperformed traditional alternatives in terms of fit and parameter stability [ 63 , 64 ]. Driver Experience . Unlicensed drivers and observations with missing license status were linked to a markedly higher risk of fatality (CMP: β = − 2.0356 and − 4.2947, respectively; p < 0.001). In contrast, drivers with more than nine years of experience demonstrated a protective effect (β = +1.2491, p < 0.0001), aligning with international evidence associating experience with improved hazard response [ 40 ]. Driver Age and Sex . The age group 25–34 showed the highest association with fatal crashes (β = +1.2686, p < 0.0001), suggesting increased exposure or risk-taking tendencies. Male drivers were significantly more likely to be involved in fatal incidents (β = +2.7039, p < 0.0001), reflecting global gender-based disparities in driving behaviour and crash involvement [ 40 , 65 , 66 ]. Temporal Patterns . Fatality rates tended to rise during evening hours (19:00–00:00) and in the summer and autumn seasons, periods commonly associated with greater traffic density, fatigue, and travel activity. Early morning hours and midweek days were comparatively safer, suggesting time-specific patterns that could inform enforcement efforts [ 67 ]. Road Infrastructure. Higher fatality risks were observed on straight and interurban roads (β = +2.5534 and + 1.6233), while roundabouts and T-intersections exhibited strong protective effects (β = − 4.3028 and − 2.0496). These findings reinforce the established role of traffic–calming infrastructure in reducing crash severity [ 68 ]. The CMP-Poisson model effectively addressed the observed over–dispersion and provided more reliable estimates than the standard Poisson or Poisson-lognormal models. Its use in this context underscores the value of adopting flexible modelling strategies in traffic safety research when equi–dispersion assumptions are violated. Overall, the results highlight the need for targeted interventions addressing unlicensed drivers, young male demographics, high–risk periods, and specific roadway types. 5.2 Limitations and Future Research This study is subject to several limitations. First, the use of aggregated group-level data may mask important within-group variations, limiting the ability to detect individual-level risk patterns. Additionally, the absence of critical variables such as vehicle type, alcohol or drug involvement, and enforcement intensity constrains the comprehensiveness of the analysis. Future research should consider multilevel modelling frameworks that account for regional and individual-level variability, allowing for a more nuanced understanding of crash determinants. Incorporating longitudinal data would also enable the evaluation of temporal trends and the effectiveness of safety policies over time. Moreover, integrating additional variables such as weather conditions, vehicle specifications, and emergency response times could enhance model accuracy. Further studies should also examine the spatial and temporal distribution of fatal crashes using geospatial methods and explore the impact of recent infrastructure and legislative interventions in Albania. Access to more detailed, individual-level crash records would be essential for developing targeted, evidence-based road safety strategies. 5.3 Conclusions This study examined fatal road traffic accidents in Albania between 2018 and 2024 using advanced count regression models to identify significant risk factors and evaluate model performance under various dispersion assumptions. Key demographic, temporal, and environmental variables were found to be significantly associated with fatal outcomes. Notably, over–dispersion in the count data highlighted the necessity of employing flexible models such as the Conway–Maxwell–Poisson (CMP–Poisson), which outperformed traditional alternatives in terms of fit and parameter stability [ 63 , 64 ]. Our findings regarding the influence of driver characteristics are consistent with previous research. Unlicensed and inexperienced drivers were associated with higher fatality risks, while those with more than nine years of experience showed a protective effect. This aligns with international evidence linking driving experience to improved hazard perception and reduced crash involvement [ 40 ]. Similarly, our results identified young drivers, particularly those aged 25–34, and male drivers as the most at-risk groups. These outcomes correspond with global studies that highlight young males as more prone to risk-taking behaviours and greater exposure to fatal crashes [ 65 , 66 ]. Temporal patterns observed in our study also resonate with earlier findings. We found that fatality risks increase during evening hours (19:00–00:00) and in summer and autumn. These results echo prior studies emphasizing that night-time driving and seasonal variations (linked to higher traffic volumes, fatigue, or risky behaviour) significantly contribute to accident severity [ 67 ]. Such similarities reinforce the importance of targeted interventions during high-risk times. Road infrastructure factors emerged as significant determinants of fatal accidents. Straight and interurban roads showed higher fatality risks, whereas roundabouts and T-intersections provided protective effects. These results are in agreement with studies that demonstrate the traffic-calming role of roundabouts and improved intersection design in reducing crash severity [ 68 ]. Our findings add further evidence from Albania, underscoring the role of infrastructure interventions in mitigating risks. From a methodological perspective, our study confirms the suitability of flexible regression models for analyzing road crash fatality data. The CMP regression model, which performed better than the standard Poisson and Poisson-lognormal models, effectively addressed the issue of over-dispersion in the dataset. This supports previous applications of the CMP model in traffic safety research, where it has been shown to yield more reliable estimates under similar data conditions [ 63 , 64 ]. The consistency of our findings with prior studies highlights both the robustness of the identified risk factors and the applicability of advanced statistical methods in capturing the complexity of accident data. At the same time, our analysis provides new evidence specific to Albania, where empirical studies remain limited, thereby contributing to a deeper understanding of risk determinants in developing contexts. Declarations Ethics approval and consent to participate Not applicable Consent for Publication Not applicable Competing Interests The authors declare that they have no competing interests. Funding This research received no funding. Hence the corresponding author hereby request 100% APC waiver for the publication of this article. Author Contribution Raimonda Dervishi, , Data collection, Conceptualization, Model formulation, Methodology, Writing Original draft preparation, Analysis, Investigation.. Fabiana Çullhaj, Literature review, Analysis, Investigation, Model formulation, Methodology, writing introduction, Editing Agbata, Benedict Celestine Model formulation, Methodology, Conceptualization, Methodology, Writing Original draft preparation, Analysis. Sander Kovaci, Editing, Literature review, writing introduction. Writing Original draft preparation. Acknowledgements Not applicable Data Availability Table 2 and 3 of this publication present the data used in this investigation. Upon reasonable request, the associated author will provide the simulation's source code. References World Health Organization (2018). Global status report on road safety 2018 . https://www.who.int/publications/i/item/9789241565684 Chen, S. (2019). The global macroeconomic burden of road injuries: Estimates and projections for 166 countries. The Lancet Planetary Health , 3 (9), e390–e398. World Health Organization (2013). Projections of mortality and causes of death, 2015 and 2030 . http://www.who.int/healthinfo/ Joshi, A. K., Joshi, C., Singh, M., & Singh, V. (2014). Road traffic accidents in hilly regions of Northern India: What has to be done? World Journal of Emergency Medicine , 5 (2), 112–115. Nyamuame, G. Y., Aglina, M. K., Akple, M. S., Philip, A., & Klomegah, W. (2015). Analysis of road traffic accidents trend in Ghana: Causing factors and preventive measures. International Journal of Engineering Sciences & Management Research , 127–132. http://www.ijesmr.com Ahmed, S., Hossain, M. A., & Sabuj, S. R. (2023). A study on road accident prediction and contributing factors using explainable machine learning models: Analysis and performance. Transportation Research Interdisciplinary Perspectives , 19 , 100814. Archer, J. (2005). Indicators for Traffic Safety Assessment and Prediction and Their Application in Micro-Simulation Modelling: A Study of Urban and Suburban Intersections . Royal Institute of Technology. European Commission, Directorate-General for Mobility and Transport (2024). Transport in the European Union – Current trends and issues . https://data.europa.eu/doi/10.2832/131741 Dervishi, R., Kasharaj, J., & Kondi, I. (2021). Estimating of road accidents during the COVID-19 pandemic year in Albania. International Journal of Ecosystems and Ecology Science , 11 (3), 489–494. https://doi.org/10.31407/ijees11.318 Dervishi, R., Kovaci, S., Agbata, B. C., Gurmu, E. D., & Emadifar, H. (2025). Novel application of time-weighted statistical techniques for trend analysis in roadway accident fatalities: A case study from Albania. Edelweiss Applied Science and Technology , 9 (5), 2535–2548. https://doi.org/10.55214/25768484.v9i5.7516 United Nations Economic Commission for Europe (2018). Road safety performance review: Albania . https://digitallibrary.un.org/record/3813635?ln=en Sohail, A., Cheema, M. A., Ali, M. E., Toosi, A. N., & Rakha, H. A. (2023). Data-driven approaches for road safety: A comprehensive systematic literature review. Safety Science , 158 , 105949. https://doi.org/10.1016/j.ssci.2022.105949 Ulu, M., & &Türkan, Y. S. (2024). Bibliometric analysis of traffic accident prediction studies from 2003 to 2023: Trends, patterns and future directions. Promet – Traffic & Transportation , 36 (5), 833–851. https://doi.org/10.7307/ptt.v36i5.576 Zheng, L., Sayed, T., & Mannering, F. (2020). Modeling traffic conflicts for use in road safety analysis: A review of analytic methods and future directions. Analytic Methods in Accident Research . https://doi.org/10.1016/j.amar.2020.100142 Park, S. (2012). Analysis of injury severity in traffic crashes: A case study of Korean expressways. KSCE Journal of Civil Engineering , 16 (7), 1280–1288. https://doi.org/10.1007/s12205-012-1527-3 Kovaci, S., Meco, Z. M., Agbata, B. C., Malotaj, M., & Dervishi, R. (2025). A data-driven study of road traffic accidents in Albania: Nonlinear regression applications in the transportation system. Edelweiss Applied Science and Technology , 9 (5), 1545–1556. http://dx.doi.org/10.55214/25768484.v9i5.7215 Ministry of Infrastructure and Energy (2020). Sectoral strategy of transport & action plan 2016–2020: 3rd monitoring report . Wali, B., Khattak, A. J., & Karnowski, T. (2020). The relationship between driving volatility in time to collision and crash-injury severity in a naturalistic driving environment. Analytic Methods in Accident Research . Amirifar, H., Ghahfarkhi, M. R., & &Chaleshtari, A. H. (2018). A study of factors affecting road accidents in Isfahan Province. In The Second International Conference on New Horizons in Engineering Sciences , Iran. Singh, D., Singh, S. P., Kumaran, M., & Goel, S. (2015). Epidemiology of road traffic accident deaths in children in Chandigarh zone of North West India. Egyptian Journal of Forensic Sciences , 6 , 255–260. https://doi.org/10.1016/j.ejfs.2015.01.008 Rasool, F. A. A., Alekri, F. A., Nabi, H. A., Naiser, M. J., Shamlooh, N. M., Alnashaba, S. A., Alherz, Z. A., et al. (2015). Prevalence and behavioral risk factors associated with road traffic accidents among medical students of Arabian Gulf University in Bahrain. International Journal of Medical Science and Public Health , 4 , 933–938. https://doi.org/10.5455/ijmsph.2015.14022015189 Zagorac, S., Markovic-Denic, L., & Nikolic, V. (2025). Trends in road traffic crash fatalities in Belgrade: A twelve-year retrospective analysis (2010–2021). Journal of Prevention . https://doi.org/10.1007/s10935-025-00846-x Dervishi, R., & Cenaj, E. (2021). Traffic accident characteristics of Albania and their prediction using Bayesian approach. International Scientific Journal Trans & MOTAUTO World , 6 (4), 151–153. Raço, E., Beliu, E., Haxhi, K., & &Zaçaj, O. (2025). Statistical modeling of weather impact on road accidents: A nonlinear distributed lag analysis in Tirana, Albania. WSEAS Transactions on Environment and Development , 21 , 196–205. https://doi.org/10.37394/232015.2025.21.17 Institute of Statistics Republic of Albania (INSTAT). Statistics of road accidents . https://www.instat.gov.al/ Abdulhafedh, A. (2017). Road accident prediction models: Different statistical modeling approaches. Journal of Transportation Technologies , 7 (2), 190–205. https://doi.org/10.4236/jtts.2017.72014 American Association of State Highway and Transportation Officials (AASHTO) (2010). Highway Safety Manual . Washington, D.C. Frempong, N. K., Yirenkyi, P. A., Nakua, E., & Dontwi, I. K. (2016). Modeling Road accident Fatalities in a Subregion of Ghana. American Journal of Scientific and Industrial Research , 7 (4), 88–94. https://doi.org/10.5251/ajsir.2016.7.4.88.94 Lord, D., Guikema, S. D., & Geedipally, S. R. (2008). Application of the Conway-Maxwell-Poisson generalized linear model for analyzing motor vehicle crashes. Accident Analysis & Prevention , 40 (3), 1123–1134. Abdel-Aty, M., & Radwan, A. E. (2000). Modeling traffic accident occurrence and involvement. Accident Analysis & Prevention , 32 (5), 633–642. https://doi.org/10.1016/S0001-4575(99)00094-9 Jovanis, P., & Chang, H. (1986). Modeling the relationship of accidents to miles traveled. Transportation Research Record , 1068 , 42–51. Al-Ghamdi, A. S. (1994). Comparison of Accident Rates Using the Likelihood Ratio Testing Technique. Transportation Research Record , 1401 , 50–54. Al-Ghamdi, A. S. (2002). Using logistic regression to estimate the influence of accident factors on accident severity. Accident Analysis & Prevention , 34 (6), 729–741. Eluru, N., Bhat, C. R., & Hensher, D. A. (2008). A mixed generalized ordered response model for examining pedestrian and bicyclist injury severity level in traffic crashes. Accident Analysis & Prevention , 40 (3), 1033–1054. Islam, M. (2024). Assessing the effects of geometry and non-geometry related factors in work-zone crashes. Traffic Injury Prevention . https://doi.org/10.1080/15389588.2024.2321914 Hermans, E., Brijs, T., Wets, G., & Vanhoof, K. (2006). Explanatory modeling of traffic crash data using the COM–Poisson model. Transportation Research Board 85th Annual Meeting , Paper No. 06-1034. Simmachan, T., Wongsai, N., Wongsai, S., & Lerdsuwansri, R. (2022). Modeling road accident fatalities with underdispersion and zero-inflated counts. Plos One , 17 (11), e0269022. https://doi.org/10.1371/journal.pone.0269022 Usman, A., & Alhassan, S. (2024). Fit a Poisson regression, negative binomial regression and Poisson log-normal regression models: An application to road accident data in North Central Nigeria. International Journal of Modeling and Applied Science Research , 6 (9). https://cambridgeresearchpub.com/ijmasr/article/view/412 Høye, A. (2016). The effects of speed on accidents and accident severity: A review. Accident Analysis & Prevention , 95 , 162–170. https://doi.org/10.1016/j.aap.2016.07.006 World Health Organization (WHO). (2023). Global status report on road safety 2023 . World Health Organization. https://www.who.int/publications/i/item/9789240072250 European Transport Safety Council (2023). PIN Flash Report 43: Reducing deaths in single-vehicle collisions . https://etsc.eu/pin International Road Assessment Programme (iRAP) (2024). Albania cuts road deaths by 23% with innovative road maintenance programme . https://irap.org/2024/02/albania-cuts-road-deaths-by-23-with-innovative-road-maintenance-programme/ World Health Organization. (2015). Global status report on road safety 2015 . WHO. https://www.who.int/publications/i/item/9789241565066 Cenaj, E., & Dervishi, R. (2024). Road accident fatalities forecasting models using Smeed’s regression analysis: A case study. European Journal of Engineering and Technology Research , 9 (6), 20–24. https://doi.org/10.24018/ejeng.2024.9.6.3213 Zhou, M., & Sisiopiku, V. P. (1997). Relationship between volume-to-capacity ratios and accident rates. Transportation Research Record , 1581 (1), 47–52. https://doi.org/10.3141/1581-07 Kassa, S. M., Lachore, T. S., Fetene, A. T., Sewnet, H., & &Demlew, M. B. (2023). Developing a correlation model between the frequency of road traffic accidents and road geometric design elements. Public Infrastructure Studies , 70 (4). https://doi.org/10.31075/PIS.70.04.01 Basu, S., & Saha, P. (2017). Regression models of highway traffic crashes: A review of recent research and future research needs. Procedia Engineering , 187 , 59–66. https://doi.org/10.1016/j.proeng.2017.04.351 Sellers, K. F., & Premeaux, B. (2021). Conway–Maxwell–Poisson regression models for dispersed count data. Wiley Interdisciplinary Reviews: Computational Statistics , 13 (3), e1533. https://doi.org/10.1002/wics.1533 Abdella, G. M., Kim, J., Al-Khalifa, K. N., & Hamouda, A. M. (2019). Penalized Conway-Maxwell-Poisson regression for modelling dispersed discrete data: The case study of motor vehicle crash frequency. Safety Science , 120 , 157–163. https://doi.org/10.1016/j.ssci.2019.06.036 Li, X., Lord, D., Zhang, Y., & Xie, Y. (2009). Predicting motor vehicle crashes using support vector machine models. Accident Analysis & Prevention , 40 (5), 1611–1618. https://doi.org/10.1016/j.aap.2009.07.009 Lord, D. (2006). Modeling motor vehicle crashes using Poisson-gamma models: Examining the effects of low sample mean values and small sample size on the estimation of the fixed dispersion parameter. Accident Analysis & Prevention , 38 (4), 751–766. https://doi.org/10.1016/j.aap.2005.11.008 Mitra, S., & Washington, S. (2007). On the nature of over-dispersion in motor vehicle crash prediction models. Accident Analysis & Prevention , 39 (3), 459–468. https://doi.org/10.1016/j.aap.2006.08.002 Khan, N. M., Soobhug, A. D., & Jannoo, Z. (2020). Modeling road traffic accidents in Mauritius using clustered longitudinal COM-Poisson with gamma random effects. Communications in Statistics: Case Studies, Data Analysis and Applications, 7 (2), 113–127. https://doi.org/10.1080/23737484.2020.1842269 Abdulla, R., Qader, B., & Sdiq, K. (2023). Traffic Accident Traits and Driver Characteristics Implication on Road Accidents using Descriptive Analysis: A Cross Sectional Study in Sulaymaniyah, Iraq. Engineering Technology and Applied Science Research , 13 (2), 10372–10376. https://dx.doi.org/10.48084/etasr.5669 Fowode, K. V., Nwaogazie, I. L., & Anyanwu, B. O. (2023). Prevalence and Factors Causing Road Traffic Accidents in Lagos State, Nigeria. Asian Journal of Advanced Research and Reports , 17 (6), 36–44. https://doi.org/10.9734/ajarr/2023/v17i6488 Akaike, H. (1974). A new look at the statistical model identification. IEEE Transactions on Automatic Control , 19 (6), 716–723. https://doi.org/10.1109/TAC.1974.1100705 Schwarz, G. (1978). Estimating the dimension of a model. The Annals of Statistics , 6 (2), 461–464. https://doi.org/10.1214/aos/1176344136 Burnham, K. P., & Anderson, D. R. (2002). Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach (2nd ed.). Springer. R Core Team. (2024). R: A Language and Environment for Statistical Computing . R Foundation for Statistical Computing. https://www.R-project.org/ Brooks, M. E., Kristensen, K., van Benthem, K. J., Magnusson, A., Berg, C. W., Nielsen, A., Skaug, H. J., Maechler, M., & Bolker, B. M. (2017). glmmTMB balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling. The R Journal , 9 (2), 378–400. https://doi.org/10.32614/RJ-2017-066 Lüdecke, D., Ben-Shachar, M., Patil, I., Waggoner, P., & Makowski, D. (2021). performance: An R package for assessment, comparison and testing of statistical models. Journal of Open Source Software , 6 (60), 3139. https://doi.org/10.21105/joss.03139 Wickham, H. (2016). ggplot2: Elegant graphics for data analysis . Springer. https://ggplot2.tidyverse.org Shmueli, G., Minka, T. P., Kadane, J. B., Borle, S., & Boatwright, P. (2005). A useful distribution for fitting discrete data: Revival of the Conway-Maxwell-Poisson distribution. Journal of the Royal Statistical Society: Series C (Applied Statistics) , 54 (1), 127–142. https://doi.org/10.1111/j.1467-9876.2005.00474.x Sellers, K. F., & Shmueli, G. (2010). A flexible regression model for count data. The Annals of Applied Statistics , 4 (2), 943–961. https://doi.org/10.1214/10-AOAS325 Elvik, R. (2010). The stability of long-term trends in the real risk of traffic injury for different road user groups. Accident Analysis & Prevention , 42 (2), 523–532. https://doi.org/10.1016/j.aap.2009.09.001 Moran, C., Bennett, J. M., & Prabhakharan, P. (2020). The relationship between cognitive function and hazard perception in younger drivers. Transportation Research Part F: Traffic Psychology and Behaviour , 74 , 104–119. Ackaah, W., & Salifu, M. (2011). Crash prediction model for two-lane rural highways in the Ashanti Region of Ghana. IATSS Research , 35 (1), 34–40. https://doi.org/10.1016/j.iatssr.2011.02.001 Federal Highway Administration (2017). Roundabouts: An informational guide (2nd ed., FHWA-SA-17-067). U.S. Department of Transportation. https://safety.fhwa.dot.gov/intersection/roundabouts/ Hilbe, J. M. (2011). Negative binomial regression (2nd ed.). Cambridge University Press. https://doi.org/10.1017/CBO9780511973420 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7860402","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":551292455,"identity":"62e98837-0245-475b-bfbe-5d60daac2b2b","order_by":0,"name":"Raimonda Dervishi","email":"","orcid":"","institution":"Polytechnic University of Tirana","correspondingAuthor":false,"prefix":"","firstName":"Raimonda","middleName":"","lastName":"Dervishi","suffix":""},{"id":551292456,"identity":"91f9c248-a60c-4d41-976c-18e2e8b5f0d8","order_by":1,"name":"Fabiana Çullhaj","email":"","orcid":"","institution":"AleksanderMoisiu University of Durrës","correspondingAuthor":false,"prefix":"","firstName":"Fabiana","middleName":"","lastName":"Çullhaj","suffix":""},{"id":551292457,"identity":"546293d0-3ab5-42e3-8689-f431f61e4fae","order_by":2,"name":"Agbata Benedict Celestine","email":"data:image/png;base64,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","orcid":"","institution":"Confluence University of Science and Technology","correspondingAuthor":true,"prefix":"","firstName":"Agbata","middleName":"Benedict","lastName":"Celestine","suffix":""},{"id":551292458,"identity":"a1f4cf0a-dfeb-48d4-85fe-b8cb21bdac65","order_by":3,"name":"Sander Kovaci","email":"","orcid":"","institution":"Polytechnic University of Tirana","correspondingAuthor":false,"prefix":"","firstName":"Sander","middleName":"","lastName":"Kovaci","suffix":""}],"badges":[],"createdAt":"2025-10-14 15:53:23","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7860402/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7860402/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":97137302,"identity":"b1884923-e49d-4ab4-a24b-954ebb5366ec","added_by":"auto","created_at":"2025-12-01 09:57:33","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":72726,"visible":true,"origin":"","legend":"","description":"","filename":"NRASTAPAPERRevised30.7.20251Copy.docx","url":"https://assets-eu.researchsquare.com/files/rs-7860402/v1/336117c01a66116ad7d9f8c8.docx"},{"id":96967647,"identity":"9792979e-1bee-424a-a083-0e238ccbbdb6","added_by":"auto","created_at":"2025-11-28 06:56:07","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6527,"visible":true,"origin":"","legend":"","description":"","filename":"256a10c657f8424683dde9589a89bbb3.json","url":"https://assets-eu.researchsquare.com/files/rs-7860402/v1/cdaadd20611a3962625fab22.json"},{"id":96967646,"identity":"13986b2d-d776-48cd-a589-7c78cbb18bea","added_by":"auto","created_at":"2025-11-28 06:56:06","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":169033,"visible":true,"origin":"","legend":"","description":"","filename":"256a10c657f8424683dde9589a89bbb31enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7860402/v1/e475ae06408a4ecb1f204b06.xml"},{"id":97137278,"identity":"905d17ee-b9cd-4580-88f8-eb1cb5c68d84","added_by":"auto","created_at":"2025-12-01 09:57:31","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":171030,"visible":true,"origin":"","legend":"","description":"","filename":"256a10c657f8424683dde9589a89bbb31structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7860402/v1/0c4b7f0aa1417c6111ba5b7f.xml"},{"id":96967650,"identity":"7dd37d8d-607b-455c-b615-29bfb031da5c","added_by":"auto","created_at":"2025-11-28 06:56:07","extension":"html","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":184082,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7860402/v1/214d1c9ae8160748ce16b278.html"},{"id":97144646,"identity":"28f38ddd-a632-4abe-b71e-52ba7418328f","added_by":"auto","created_at":"2025-12-01 10:11:33","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1035754,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7860402/v1/e53dbe51-583d-4f94-b1ce-82ca9ef5a32a.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Modelling Road Crash Fatalities in Albania: A Cross-Sectional Study of Risk Factors Using Advanced Count Regression","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eRoad traffic accidents result in an estimated 20 to 50\u0026nbsp;million injuries globally each year, alongside approximately 1.3\u0026nbsp;million fatalities [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Alarmingly, nearly 90% of these deaths occur in low- and middle-income countries, underscoring a major global health and development disparity [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. From 2015 to 2030, road injuries both fatal and non-fatal are projected to cost the global economy around USD 1.8 trillion, highlighting the immense financial burden of traffic-related harm ([\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]). Road traffic accidents remain a major threat to public safety across all nations and continue to challenge global injury prevention strategies [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. In this context, it has been argued that developing nations should adopt infrastructure and safety frameworks modelled after those in high-income countries [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Indeed, traffic-related deaths remain a leading cause of mortality worldwide [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. At the European level, the European Commission has emphasized safe and efficient mobility as part of its transport integration and sustainability agenda, with the Vision Zero strategy aiming to eliminate road fatalities by 2030. Albania has adopted several reforms in line with EU standards, including stricter traffic law enforcement, harmonized penalties, enhanced driver training, adoption of vehicle safety technologies, and modernization of infrastructure. The European Commission has also introduced directives to strengthen road safety management through standardized audits [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Despite a decline in accident cases in the past decade, road safety remains a concern in Albania [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. The country is far from achieving the WHO target of reducing road crash fatalities by 50% by 2030 [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Data show that driver violations account for the largest share of accidents with loss of life, followed by pedestrian non-compliance [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. This trend is especially concerning given Albania\u0026rsquo;s decreasing population alongside growing motorization. Road traffic accidents are driven by complex and interrelated factors, making risk factor analysis essential for targeted prevention [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. While progress has been made in accident modelling, further research is needed [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e], particularly on risk factors and mitigation strategies, which remain a priority in traffic safety research [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eIn Albania, studies indicate progress but also highlight persistent challenges in improving road safety [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. Authorities stress the importance of education, infrastructure upgrades, and consistent enforcement, supported by investment in safety-focused transport strategies [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. According to the WHO, road traffic injuries are a leading cause of death among young people worldwide, and Albania is no exception. National road networks, driver behaviour, and environmental conditions all contribute to accident frequency. A cohesive policy framework emphasizing safety, sustainability, governance, and integration is required, alongside both domestic reforms and international cooperation. A range of factors including driver characteristics, road conditions, and time affect accident severity [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Driver errors, such as speeding, substance use, and failure to use protective measures, are among the most critical [\u003cspan additionalcitationids=\"CR20 CR21\" citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. Other significant factors include age, driving time, and location [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eHowever, existing research on road safety in Albania is limited, with few studies using advanced regression approaches such as Poisson\u0026ndash;Lognormal (PL) or Conway\u0026ndash;Maxwell\u0026ndash;Poisson (CMP) models. Prior work has applied Bayesian Poisson models to predict risk levels based on time, location, speeding, and infrastructure [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e], while others examined specific factors like driver behavior or weather conditions [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Yet, a comprehensive analysis considering multiple interacting factors remains lacking.\u003c/p\u003e\u003cp\u003eThe present study addresses this gap by applying PL and CMP regression models to road crash fatality data from Albania (2018\u0026ndash;2024). Using official statistics from the Institute of Statistics (INSTAT), we investigate how road conditions, driver behaviour, and temporal factors contribute to fatal accidents. This study contributes to the growing literature on road safety in developing contexts and aims to provide evidence that can inform targeted interventions and policy decisions for improving traffic safety outcomes in Albania.\u003c/p\u003e"},{"header":"2. Related Works","content":"\u003cp\u003eAccurate modeling and prediction of road accidents are fundamental to understanding the causal mechanisms behind crash severity and the effectiveness of safety interventions. Traffic accident prediction models are commonly employed to examine relationships between crash severity outcomes and contributing variables, including driver behavior, vehicle characteristics, road and time factors, and environmental and traffic conditions. Preventative strategies targeting road fatalities and injuries often require a multidisciplinary approach involving transport agencies, police, public health authorities, and educational institutions. These interventions range from infrastructure redesign and vehicle safety improvements to post-crash care enhancement and behavioral reforms through education and enforcement [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. Over the past few decades, a broad range of statistical and regression-based approaches have been developed to model crash data. Among them, the Poisson, Poisson-Lognormal, Negative Binomial, and Conway\u0026ndash;Maxwell\u0026ndash;Poisson (\u003cem\u003eCMP\u003c/em\u003e) regression models are commonly employed due to their adaptability to real-world data characteristics, particularly the count nature of accident data [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. The traditional Poisson regression model is widely used because of its simplicity and interpretability. It assumes that the mean and variance of crash frequency are equal (equi\u0026ndash;dispersion). However, real-world fatality datasets often violate this assumption due to over\u0026ndash;dispersion or under\u0026ndash;dispersion, which limits the applicability of the Poisson model in many contexts [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. To address over\u0026ndash;dispersion where the variance exceeds the mean researchers have utilized models such as the Conway-Maxwell-Poisson and Poisson-Lognormal regressions. These models introduce additional parameters to account for unobserved heterogeneity in accident data. The study [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e] introduced the CMP regression model as a flexible alternative capable of accommodating both over- and under-dispersed data, a limitation for which neither the traditional Poisson nor the Negative Binomial models are well suited.\u003c/p\u003e\u003cp\u003eSeveral empirical studies have applied Poisson models to real-world crash datasets. As shown in [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e], the frequency of accident involvement using Poisson-family models, identifying demographic characteristics such as age and gender as significant predictors. Their study highlighted the impact of geometric and traffic-related features on accident occurrence. Similarly, the authors in [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e] advocated for Poisson regression in highway safety research, emphasizing its statistical advantages over linear models in modeling accident frequencies. In their analysis of traffic accident severity in Saudi Arabia, [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e] employed both Poisson and logistic regression models to assess the relative impact of various risk factors on crash outcomes.\u003c/p\u003e\u003cp\u003eFurther improvement was introduced by [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e], who developed a mixed generalized ordered response model to analyze injury severity among pedestrians and cyclists. Although this study focused on vulnerable road users, its methodology applies to broader crash severity analysis. More recently, [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e] employed a random parameter logit model incorporating heterogeneity in both the mean and variance to investigate driver injury severity in Florida work\u0026ndash;zone crashes, differentiating the influence of geometric versus non-geometric (mainly behavioral) factors. The model results encompassed an array of factors, including spatial characteristics, vehicle characteristics, environmental conditions, geometric features, crash attributes, traffic patterns, and driver traits. These factors were intricately connected with human factors, particularly driver behaviors and characteristics, playing a pivotal role in work-zone crashes tied to geometric and non-geometric attributes. A significant contribution to flexible modeling is the Conway\u0026ndash;Maxwell\u0026ndash;Poisson regression, which offers a robust framework for handling data characterized by over\u0026ndash;, under\u0026ndash;, or equi\u0026ndash;dispersion. In a comparative study [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e], the model was evaluated using both simulated and real accident data from Canada and South Korea. Their findings demonstrated the model\u0026rsquo;s capability to provide reliable estimates across various dispersion scenarios, although limitations were noted for small sample sizes or datasets with low mean crash frequencies. Nevertheless, the CMP model\u0026rsquo;s computational burden was found to be manageable, and its predictive performance was comparable to or better than competing models such as the Poisson\u0026ndash;Gamma, making it a viable option for applied crash data analysis in transportation safety research. A study [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]applied the Poisson, CMP, and Zero\u0026ndash;Inflated Conway-Maxwell Poisson (\u003cem\u003eZICMP\u003c/em\u003e) models to analyze road crash fatalities in Thailand, where the data showed a rare pattern of under\u0026ndash;dispersion with excessive zero counts. Their study indicated that although the ZICMP model marginally outperformed the standard CMP model, the necessity for zero\u0026ndash;inflation modeling depends on the specific research question rather than solely on the data structure. Road and environmental conditions, as well as seasonal patterns (such as during the Songkran festival), were identified as significant contributors to fatal crashes.\u003c/p\u003e\u003cp\u003eAnother study [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e] conducted a comparative analysis using Poisson, Poisson\u0026ndash;Lognormal, and Negative Binomial regression models to examine road crash data from North Central Nigeria, evaluating the contributions of key factors Driver Error, Faulty Vehicle, and Road Condition to fatalities. Their analysis determined that the Poisson-Lognormal model offered the best statistical fit based on Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), and pseudo R-squared values. The study found that driver error was the most significant predictor of fatalities, contributing to a 2.17% increase in fatality risk per unit increase. Faulty vehicles were also substantial, whereas road condition did not exhibit a statistically meaningful effect.\u003c/p\u003e\u003cp\u003eSimilarly, studies by [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e] and [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e] further validated the usefulness of Poisson\u0026ndash;family models in regions with high accident data rates. Moreover, [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e] examined accident data in a subregion of Ghana, confirming the presence of over\u0026ndash;dispersion in fatality counts and advocating for the use of over\u0026ndash;dispersion\u0026ndash;tolerant models.\u003c/p\u003e"},{"header":"3. Trends in Road crash Fatalities in Albania","content":"\u003cp\u003eRoad crash fatalities result from a multifaceted interplay of human, environmental, vehicular, and institutional factors. Understanding these causes is essential for designing effective interventions aimed at reducing mortality on the roads. Although accidents may occur due to unforeseeable events, the literature consistently identifies a set of recurring\u0026mdash;and often preventable causes of fatal outcomes in road accidents [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. While road accidents can arise from multiple causes, human behavior consistently emerges as the most influential, responsible for over half of collisions on its own. When coupled with factors like road conditions or vehicle issues, human error is involved in the vast majority of cases, reaching nearly 93% [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Speeding continues to be a primary cause of fatal crashes, as excessive velocity reduces a driver\u0026rsquo;s ability to respond to hazards and increases the severity of impacts. The World Health Organization (2023) reports that speeding contributes to approximately 30% of road traffic deaths globally [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. In Albania, speed\u0026ndash;related incidents are particularly prevalent on rural and regional roads, where enforcement is inconsistent and signage is often inadequate. Driving under the influence of alcohol or drugs is another major factor. Substance impairment diminishes cognitive and motor functions, heightening the risk of fatal outcomes. The European Transport Safety Council (2023) highlights a troubling pattern of alcohol-related road deaths in Southeast Europe, including Albania, where young male drivers are disproportionately involved in fatal incidents. Additionally, distracted driving stemming from mobile phone use, in-vehicle technologies, or fatigue has become an increasingly prominent risk factor, especially in urban environments where multitasking is common [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eInstitutional and systemic limitations significantly contribute to the persistence of elevated road traffic fatality rates. In Albania, although legal frameworks exist to regulate driver behaviour and enhance traffic safety, enforcement remains inconsistent, particularly in rural and peri-urban areas where oversight is weaker [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]. Weak institutional coordination, limited capacity for traffic law enforcement, and insufficient investment in road safety infrastructure further undermine efforts to mitigate risk. Moreover, deficiencies in driver education and public awareness campaigns reduce the effectiveness of existing policies. Of critical concern is the inadequacy of post-crash emergency response systems. Delays in medical intervention following traffic collisions often due to logistical constraints or poor coordination between emergency services frequently lead to fatalities that might otherwise have been preventable [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e]. These institutional shortcomings underscore the need for a more integrated, well\u0026ndash;funded, and professionally managed road safety system. Over the recent years, Albania has witnessed fluctuating yet persistently concerning trends in road crash fatalities. While certain years have reflected modest declines in fatality rates, the broader trajectory highlights the ongoing difficulty in achieving sustained and meaningful reductions. This challenge is driven by a complex combination of factors, including inadequate and underdeveloped infrastructure, rapid urban expansion, increased motorization, and widespread risky road user behaviours. Together, these elements continue to undermine national road safety efforts and pose significant barriers to progress. Nevertheless, some infrastructure-focused initiatives have shown measurable results. For instance, a World Bank-supported program implemented through the International Road Assessment Programme (IRAP) led to a 23% reduction in road deaths on primary and secondary road corridors in Albania saving an estimated 68 lives annually by improving road design and safety features [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]. This demonstrates that, despite structural and behavioural challenges, targeted engineering interventions can yield meaningful progress toward fatality reduction goals. Failure to use protective equipment such as seatbelts and helmets further exacerbates fatality risks. According to WHO (2023), proper seatbelt use can reduce the risk of death by 45\u0026ndash;50% for front-seat occupants and up to 75% for those in the rear. Despite awareness campaigns in Albania, usage remains inconsistent, particularly among motorcyclists and drivers in rural areas [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eNational statistics indicate that fatality rates declined slightly between 2018 and 2020, likely reflecting the effects of early enforcement initiatives and public safety campaigns. However, this progress was reversed in 2023 and 2024, with a notable increase in accident frequency. This recent rise appears to correlate with Albania\u0026rsquo;s growing vehicle fleet and the resurgence of high-risk driving behaviours. According to WHO (2023), Albania\u0026rsquo;s road traffic death rate stood at approximately 10.8 deaths per 100,000 population in 2021[\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. INSTAT (2024) reported that 86.4% of road accidents in 2023 were attributed to driver violations, with drivers aged 25\u0026ndash;34 comprising the most affected demographic group. Despite some recent improvements in pedestrian safety evidenced by a reduced share of pedestrian-involved accidents in 2024 the broader picture remains dominated by human factors and increased motorization. Disaggregated data further reveal shifting patterns in the distribution of serious injuries across user groups. Between 2018 and 2024, pedestrians accounted for 25.3% of serious injuries, followed closely by passengers (25.2%) and motorcyclists (23.7%) [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e].\u003c/p\u003e\u003cp\u003ePrevious empirical studies underscore the direct correlation between vehicle volume and fatality rates [\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. Environmental and/or road factors have been consistently identified as statistically significant contributors to road crash fatalities in Albania and beyond [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e, \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e, \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e]. Furthermore, variables such as time of day, seasonal trends, road type, and intersection design are increasingly acknowledged by both global and national studies as influential determinants of crash severity [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. A nuanced understanding of these environmental and roadway dimensions is therefore essential for the development of data-driven safety interventions and infrastructure planning. Between 2018 and 2024, Albania recorded a total of 16,192 road traffic accidents, involving an estimated 24,020 individuals. These incidents resulted in 2,199 fatalities and 19,622 injuries. However, it is important to note that official figures may underestimate the true burden due to underreporting, especially in cases where individuals fled the scene or the incident was not formally documented [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. A comprehensive summary of fatality characteristics by key variables is provided in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eSummary of fatality characteristics by key variables, 2018\u0026ndash;2024.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"4\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVariables\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNo of Fatalities (%)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eVariables\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNo of Fatalities (%)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cem\u003eDrivers experience\u003c/em\u003e\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cem\u003eSeasons of accidents\u003c/em\u003e\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eWithout License\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e105 (7.78%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eAutumn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e338 (25.1%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e0\u0026ndash;3 years\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e238 (17.64%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSpring\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e308 (22.8%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e3\u0026ndash;6 years\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e247 (18.31%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSummer\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e356 (26.4%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e6\u0026ndash;9 years\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e219 (16.23%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eWinter\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e347 (25.7%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAbove 9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e505 (37.43%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003eTime of accidents\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMissing\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e35 (2.59%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e00:00\u0026ndash;06:00\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e149 (11.0%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eSex of drivers\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e06:00\u0026ndash;08:00\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e70 (5.2%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1047 (77.6%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e08:00\u0026ndash;12:00\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e192 (14.2%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFemale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e267 (19.8%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e12:00\u0026ndash;14:00\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e123 (9.1% )\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMissing\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e35 (2.6%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e14:00\u0026ndash;17:00\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e221 (16.4%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eThe age group of drivers\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e17:00\u0026ndash;19:00\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e266 (19.7%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e0\u0026ndash;24\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e246 (18.2%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e19:00\u0026ndash;00:00\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e328 (24.3%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e25\u0026ndash;34\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e510 (37.8%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003eRoad geometry\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e35\u0026ndash;44\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e221 (16.4%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eStraight road\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e968 (71.8%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e45\u0026ndash;59\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e180 (13.3%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eCurved\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e216 (16.0%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAbove 59\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e157 (11.6%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eT-Intersection\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e104 (7.7%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMissing\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e35 (2.6%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRoundabout\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e32 (2.4%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eDay of the week of accidents\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eOther\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e29 (2.1%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMonday\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e199 (14.8%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003eRoad type\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTuesday\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e208 (15.4%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eHighway\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e80 (5.9%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eWednesday\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e181 (13.4%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eInterurban Road\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e613 (45.4%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eThursday\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e182 (13.5%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eUrban Road\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e446 (33.1%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFriday\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e208 (15.4%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eLocal Road\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e210 (15.6%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSaturday\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e182 (13.5%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSunday\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e189 (14.0%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e"},{"header":"4. Modelling Road Crash Fatalities Data","content":"\u003cp\u003eOver the years, numerous data-related and methodological issues have been highlighted in the literature concerning accident frequency analysis. A core objective of model-building techniques in this context is to develop an accurate and parsimonious representation of the relationship between accident outcomes\u0026mdash;typically measured as count data and a set of explanatory variables [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e]. As emphasized in recent studies, crash prediction plays a fundamental role in the broader road safety management process, facilitating proactive interventions and infrastructure planning [\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. Given that the dependent variable in this study namely, the number of road crash fatalities is discrete and demonstrates under\u0026ndash;dispersion (the variance exceeds the mean), traditional Poisson regression may not suffice. To address this, two more flexible count regression models were applied: the Poisson-Lognormal (\u003cem\u003ePL\u003c/em\u003e) and the Conway\u0026ndash;Maxwell\u0026ndash;Poisson (\u003cem\u003eCMP\u003c/em\u003e) models. These models allow for a more accurate estimation of significant risk factors by accounting for unobserved heterogeneity and the potential for both over\u0026ndash;dispersion and under\u0026ndash;dispersion in the data.\u003c/p\u003e\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e4.1 Poisson-Lognormal Regression Model\u003c/h2\u003e\u003cp\u003eThe \u003cem\u003ePL\u003c/em\u003e model extends the standard Poisson framework by incorporating a log\u0026ndash;normally distributed random effect to account for unobserved heterogeneity. This approach is effective in modelling over\u0026ndash;dispersed count data and estimating how each covariate influences the expected number of fatalities. This model assumes a lognormal distribution for the error term, offering enhanced flexibility in modelling variance heterogeneity and multilevel data structures [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. The model specification is:\u003c/p\u003e\u003cp\u003e\u003cem\u003eY\u003c/em\u003e\u003csub\u003ei\u003c/sub\u003e ~ Poisson (\u003cem\u003eλ\u003c/em\u003e\u003csub\u003ei\u003c/sub\u003e), log(\u003cem\u003eλ\u003c/em\u003e\u003csub\u003ei\u003c/sub\u003e)\u0026thinsp;=\u0026thinsp;\u003cem\u003eX\u003c/em\u003e\u003csub\u003ei\u003c/sub\u003e\u003csup\u003eT\u003c/sup\u003e\u003cem\u003eβ\u003c/em\u003e+ \u003cem\u003eℇ\u003c/em\u003e\u003csub\u003ei\u003c/sub\u003e, \u003cem\u003eℇ\u003c/em\u003e\u003csub\u003ei\u003c/sub\u003e ~ \u003cem\u003eN\u003c/em\u003e(0, σ\u003csup\u003e2\u003c/sup\u003e)\u003c/p\u003e\u003cp\u003ewhere\u003cem\u003eY\u003c/em\u003e\u003csub\u003ei\u003c/sub\u003e denotes the fatality count for accident i, \u003cem\u003eX\u003c/em\u003e\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e is the vector of covariates, \u003cem\u003eβ\u003c/em\u003e the regression coefficients, and \u003cem\u003eℇ\u003c/em\u003e\u003csub\u003ei\u003c/sub\u003e​ the normally distributed random effect.\u003c/p\u003e\u003cp\u003eHowever, a significant limitation of the Poisson-lognormal model is the lack of a closed-form expression for its likelihood function, complicating parameter estimation and computational implementation [\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. Despite this, it remains a valuable tool when over\u0026ndash;dispersion or complex random effects are present. A few researchers have suggested the use of the Poisson-lognormal regression for modelling accident data [\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e].\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003e4.2 Conway-Maxwell Poisson Regression Model\u003c/h2\u003e\u003cp\u003eStandard Poisson regression assumes equi\u0026ndash;dispersion, where the mean equals the variance a condition rarely met in real-world crash data. The Conway\u0026ndash;Maxwell\u0026ndash;Poisson (\u003cem\u003eCMP\u003c/em\u003e) distribution generalizes the Poisson to accommodate both over\u0026ndash;dispersion and under\u0026ndash;dispersion, making it well-suited for traffic crash modelling [\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e]. The CMP\u0026ndash;Poisson regression framework offers greater flexibility by incorporating a dispersion parameter that adjusts variance independently from the mean. This model is particularly useful when accident data exhibit under\u0026ndash;dispersion, a scenario where traditional Poisson models perform poorly [\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e]. It generalizes the Poisson distribution with an additional dispersion parameter\u003cem\u003eν\u003c/em\u003e, allowing for flexible modelling of both over\u0026ndash;dispersion and under dispersion:\u003c/p\u003e\u003cp\u003e\u003cem\u003eY\u003c/em\u003e\u003csub\u003ei\u003c/sub\u003e ~ CMP(\u003cem\u003eλ\u003c/em\u003e\u003csub\u003ei\u003c/sub\u003e,\u003cem\u003eν\u003c/em\u003e), log(\u003cem\u003eλ\u003c/em\u003e\u003csub\u003ei\u003c/sub\u003e)\u0026thinsp;=\u0026thinsp;\u003cem\u003eX\u003c/em\u003e\u003csub\u003ei\u003c/sub\u003e\u003csup\u003eT\u003c/sup\u003e\u003cem\u003eβ\u003c/em\u003e\u003c/p\u003e\u003cp\u003eNevertheless, estimating CMP regression models can be computationally intensive, and small sample sizes or low mean counts can adversely affect parameter stability [\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e]. Such challenges have limited widespread adoption despite the theoretical advantages. Some researchers have suggested the Conway\u0026ndash;Maxwell\u0026ndash;Poissonmethod for modelling accident data [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e]\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003e4.3Assessment of Model Fit and Selection Criteria\u003c/h2\u003e\u003cp\u003eTo assess model adequacy and comparative performance, we employed three standard statistical metrics: the Akaike Information Criterion (\u003cem\u003eAIC\u003c/em\u003e), the Bayesian Information Criterion (\u003cem\u003eBIC\u003c/em\u003e), and the log-likelihood value. These measures provide a consistent framework for evaluating the trade-off between model fit and complexity.\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eThe Akaike Information Criterion, proposed by [\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e],estimates the relative quality of models by penalizing the number of parameters while prioritizing goodness-of-fit.It is defined as:\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eAIC = \u0026ndash;Log\u003cem\u003eL\u003c/em\u003e\u0026thinsp;+\u0026thinsp;2k\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eWhere Log\u003cem\u003eL\u003c/em\u003e is the maximized log-likelihood and \u003cem\u003ek\u003c/em\u003e is the number of model parameters.\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eThe Bayesian Information Criterion, introduced by [\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e], imposes a stronger penalty for model complexity based on sample size \u003cem\u003en\u003c/em\u003e:\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eBIC = \u0026ndash; Log\u003cem\u003eL\u003c/em\u003e\u0026thinsp;+\u0026thinsp;k Log\u003cem\u003en\u003c/em\u003e\u003c/p\u003e\u003cp\u003eThis renders the BIC criterion more conservative than AIC when selecting among competing models, as it imposes a stronger penalty for model complexity. While AIC tends to prioritize predictive accuracy, BIC emphasizes parsimony, especially as sample size increases [\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e].\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eThe log-likelihood measures the probability of the observed data given the model parameters. A higher log-likelihood value indicates a better fit [\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e]. Formally, it is expressed as:\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eLogL(θ∣y\u003csub\u003ei\u003c/sub\u003e) = \u003cem\u003e\u0026sum;\u003c/em\u003e​log f(y\u003csub\u003ei\u003c/sub\u003e∣θ)\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eWhere y\u003csub\u003ei\u003c/sub\u003e are observed values, and f(y\u003csub\u003ei\u003c/sub\u003e​∣θ) is the likelihood function given parameters θ.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e4.4Computational Environment\u003c/h2\u003e\u003cp\u003eAll statistical analyses were performed using the R statistical software environment [\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e]. A variety of specialized R packages were employed to model fatal traffic accidents using count data frameworks. The standard Poisson regression model was estimated via the \u003cspan fontcategory=\"NonProportional\" class=\"\" name=\"Emphasis\"\u003eglm()\u003c/span\u003e function. To address over-dispersion and accommodate non\u0026ndash;equidispersed count outcomes, we implemented the Conway\u0026ndash;Maxwell\u0026ndash;Poisson (CMP-Poisson) model using the \u003cspan fontcategory=\"NonProportional\" class=\"\" name=\"Emphasis\"\u003eglmmTMB\u003c/span\u003e package [\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e]. For the Poisson-lognormal model, we applied a generalized linear mixed model with random intercepts using the same framework, allowing us to model latent heterogeneity between groups. Model selection and comparative performance were assessed using the log-likelihood value, Akaike Information Criterion (\u003cem\u003eAIC\u003c/em\u003e), and Bayesian Information Criterion (\u003cem\u003eBIC\u003c/em\u003e). These statistics were calculated using functions from the \u003cspan fontcategory=\"NonProportional\" class=\"\" name=\"Emphasis\"\u003eperformance\u003c/span\u003e and \u003cspan fontcategory=\"NonProportional\" class=\"\" name=\"Emphasis\"\u003einsight\u003c/span\u003e packages [\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e]. Data management and visualization were conducted using \u003cspan fontcategory=\"NonProportional\" class=\"\" name=\"Emphasis\"\u003edplyr\u003c/span\u003e and \u003cspan fontcategory=\"NonProportional\" class=\"\" name=\"Emphasis\"\u003eggplot2\u003c/span\u003e [\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e], ensuring a reproducible and transparent workflow.\u003c/p\u003e\u003c/div\u003e"},{"header":"5. Estimated Model Results","content":"\u003cp\u003eThis study employs a cross-sectional design, analyzing all road traffic accident cases and their associated fatalities over eight years. Each observation in the dataset represents a single traffic accident, with the primary outcome variable being the count of fatalities per accident.\u003c/p\u003e\u003cp\u003eIndependent variables are grouped into three main categories: driver,time, and road factors.\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003e\u003cem\u003eDriver factors\u003c/em\u003e consisted of driver experience (categorized as without license, 0\u0026ndash;3 years, 3\u0026ndash;6 years, 6\u0026ndash;9 years, and above 9 years), driver age groups (0\u0026ndash;24, 25\u0026ndash;34, 35\u0026ndash;44, 45\u0026ndash;59, and above 59), and driver sex (male, female).\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cem\u003eTime factors\u003c/em\u003e included time of accident (divided into time intervals ranging from 00:00 to 24:00 hours), day of the week, and season of the accident (spring, summer, autumn, and winter).\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cem\u003eRoad factors\u003c/em\u003e comprised road type (highway, interurban roadway, urban roadway, and local roads) and road geometry (straight road, curved, roundabout, T-intersection, and other).\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eThe dependent variable in this study was defined as the number of fatalities recorded within each observational group. To adjust for differing levels of exposure across groups, the total number of road traffic accidents was incorporated as an offset term in the regression models, using its natural logarithm. This approach allowed for modelling fatality risk relative to accident frequency. The explanatory variables were selected based on their established relevance in the road safety literature and the availability of corresponding data from the Albanian Institute of Statistics (\u003cem\u003eINSTAT\u003c/em\u003e). These covariates were included in the modelling framework to assess their contribution to variations in fatality counts per accident. A summary of the main findings is provided in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eCoefficients(p-value) from the Poisson, PL,and CMP regression models.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"13\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c11\" colnum=\"11\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c12\" colnum=\"12\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c13\" colnum=\"13\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVariables\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003ePoisson\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003ePL\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003eCMP\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e\u003cp\u003e\u003cem\u003eDrivers' experience category\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"6\" nameend=\"c10\" namest=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c13\" namest=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIntercept\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-3.3844 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-3.3845 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-3.3814 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eWithout License\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e+\u0026thinsp;0.1063 (0.36416 )\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e+\u0026thinsp;0.1060 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e+\u0026thinsp;0.0069 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e0\u0026ndash;3 years\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e+\u0026thinsp;0.2744 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e+\u0026thinsp;0.2744 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e+\u0026thinsp;0.2747 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e3\u0026ndash;6 years\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.1769 (0.0514)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.1770 (0.0513 )\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.1777 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e6\u0026ndash;9 years\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e+\u0026thinsp;0.0792 (0.3971)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e+\u0026thinsp;0.0793 (0.3970)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e+\u0026thinsp;0.0791 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAbove 9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e+\u0026thinsp;0.0961 (0.2212 )\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e+\u0026thinsp;0.0962 (0.2213)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e+\u0026thinsp;0.0961 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMissing\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.6374 (0.0042)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.6374 (0.3642)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.6559 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e\u003cp\u003e\u003cem\u003eSex of drivers category\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFemale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e+\u0026thinsp;1.3083 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e+\u0026thinsp;1.3081 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e+\u0026thinsp;0.3516 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e+\u0026thinsp;1.7007 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e+\u0026thinsp;1.7006 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e+\u0026thinsp;1.6994 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMissing\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e+\u0026thinsp;0.3963 (0.0274)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e+\u0026thinsp;0.3962 (0.0275)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e+\u0026thinsp;0.3739 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e\u003cp\u003e\u003cem\u003eThe age group of the driver category\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"6\" nameend=\"c12\" namest=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c13\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cem\u003e0\u0026ndash;24\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e+\u0026thinsp;0.3543 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e+\u0026thinsp;0.3544 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e+\u0026thinsp;0.3521 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e25\u0026ndash;34\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e+\u0026thinsp;0.3851 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e+\u0026thinsp;0.3850 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e+\u0026thinsp;0.3824 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e35\u0026ndash;44\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.2084 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.2084 (0.0335)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.2124 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e45\u0026ndash;59\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.1097 (0.2364)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.1099 (0.2364)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.1140 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAbove 59\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e+\u0026thinsp;0.2049 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e+\u0026thinsp;0.2049 (0.0448)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e+\u0026thinsp;0.2022 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMissing\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.5575 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.5576 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.5634 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"7\" nameend=\"c7\" namest=\"c1\"\u003e\u003cp\u003e\u003cem\u003eDay of the week of the accident category\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c10\" namest=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c13\" namest=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMonday\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.0572 (0.5639 )\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.0582 (0.5639)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.0575 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTuesday\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e+\u0026thinsp;0.0335 (0.7320)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e+\u0026thinsp;0.0337 (0.7320)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e+\u0026thinsp;0.0313 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eWednesday\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.1040 (0.3060)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.1044 (0.3060)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.1055 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eThursday\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.1161 (0.2523)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.1160 (0.2523)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.1194 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFriday\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e+\u0026thinsp;0.2255 (0.0202)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e+\u0026thinsp;0.2256 (0.0202)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e+\u0026thinsp;0.2250 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSaturday\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.0964 (0.3422)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.0965 (0.3422)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.0987 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSunday\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.0397 (0.6911)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.0399 (0.6911)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.0426 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e\u003cp\u003e\u003cem\u003eSeasons of accidents category\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c9\" namest=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c13\" namest=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAutumn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e+\u0026thinsp;0.2947 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e+\u0026thinsp;0.2948 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e+\u0026thinsp;0.2950 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSpring\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e+\u0026thinsp;0.0430 (0.5845)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e+\u0026thinsp;0.0431 (0.5845)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e+\u0026thinsp;0.0431 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSummer\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.0414 (0.5809)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.0419 (0.5809)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.0416 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eWinter\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e+\u0026thinsp;0.0716 (0.3465)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e+\u0026thinsp;0.0719(0.3465)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e+\u0026thinsp;0.0712 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"8\" nameend=\"c8\" namest=\"c1\"\u003e\u003cp\u003e\u003cem\u003eTime of accidents category\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c13\" namest=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e00:00\u0026ndash;06:00\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.3293 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.3294 (0.0019 )\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.3258 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e06:00\u0026ndash;08:00\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.7385 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.7386 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.7358 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e08:00\u0026ndash;12:00\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.9838 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.9838 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.9805 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e12:00\u0026ndash;14:00\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.9652 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.9653 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.9622 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e14:00\u0026ndash;17:00\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.7248 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.7249 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.7213 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e17:00\u0026ndash;19:00\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.3553 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.3555 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.3521(\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e19:00\u0026ndash;00:00\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.4401 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.4405 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.4378 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e\u003cp\u003e\u003cem\u003eRoad geometry category\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStraight road\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.2596 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.2597 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.2596 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eT-Intersection\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.8992 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.8993 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.9000 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRoundabout\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.1136 (0.5485)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.1135 (0.5486)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.1150 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOther\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e+\u0026thinsp;0.3857 (0.0511)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e+\u0026thinsp;0.3858 (0.0513)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e+\u0026thinsp;0.3876 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u003cp\u003e\u003cem\u003eRoad type category\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHighway\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.4626 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.4627 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.4647 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eInterurban Road\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.3442 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.3443 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.3465 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUrban Road\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-1.2375 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-1.2375 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-1.2401 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLocal Road\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e-0.5056 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e\u003cp\u003e-0.5055 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c13\" namest=\"c11\"\u003e\u003cp\u003e-0.5087 (\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eEstimated parameters (p-value), Log-Likelihood value, AIC, and BIC criteria for the advanced regression model.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"4\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003ePoisson\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003ePL\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eCMP\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eΒ\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-3.3844 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-3.3845 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-3.3814 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eΘ\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u0026asymp;\u0026thinsp;0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.120\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAIC\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e381.26\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e383.30\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e137.3\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBIC\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e458\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e212\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLog-Likelihood Value\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-190.6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-148.6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-25.6\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eTo evaluate the adequacy of the fitted regression models, three widely used model selection criteria were employed: the Akaike Information Criterion (AIC), the Bayesian Information Criterion (BIC), and the log-likelihood value. These indicators provide a unified framework for comparing models with differing numbers of parameters by balancing model complexity with goodness\u0026ndash;of\u0026ndash;fit. As shown in Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, the standard Poisson regression model yielded an AIC of 381.3 and a log\u0026ndash;likelihood of \u003cb\u003e\u0026minus;\u0026thinsp;1\u003c/b\u003e48.6, indicating poor fit due to over\u0026ndash;dispersion. The Poisson-lognormal model, although slightly more flexible, failed to significantly improve the fit, as evidenced by negligible variance (\u0026asymp;\u0026thinsp;0) in the random intercept. In contrast, the standardCOM-Poisson model outperformed all others, with an AIC of 137.3, BIC of 212.0, and log-likelihood of \u0026minus;\u0026thinsp;25.6. Moreover, its estimated dispersion parameter (θ\u0026thinsp;=\u0026thinsp;0.0019) confirmed its capacity to address severe over\u0026ndash;dispersion in the data. These findings support the use of the CMP\u0026ndash;Poisson model as the most parsimonious and best\u0026ndash;fitting model for analyzing over\u0026ndash;dispersed fatality count data in road traffic research.\u003c/p\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e5.1 Discussion\u003c/h2\u003e\u003cp\u003eThis study examined fatal road traffic accidents in Albania between 2018 and 2024 using advanced count regression models to identify significant risk factors and evaluate model performance under various dispersion assumptions. Key demographic, temporal, and environmental variables were found to be significantly associated with fatal outcomes. Notably, over\u0026ndash;dispersion in the count data highlighted the necessity of employing flexible models such as the Conway\u0026ndash;Maxwell\u0026ndash;Poisson (CMP\u0026ndash;Poisson), which outperformed traditional alternatives in terms of fit and parameter stability [\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e]. \u003cem\u003eDriver Experience\u003c/em\u003e. Unlicensed drivers and observations with missing license status were linked to a markedly higher risk of fatality (CMP: β = \u0026minus;\u0026thinsp;2.0356 and \u0026minus;\u0026thinsp;4.2947, respectively; p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). In contrast, drivers with more than nine years of experience demonstrated a protective effect (β = +1.2491, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001), aligning with international evidence associating experience with improved hazard response [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e].\u003cem\u003eDriver Age and Sex\u003c/em\u003e. The age group 25\u0026ndash;34 showed the highest association with fatal crashes (β = +1.2686, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001), suggesting increased exposure or risk-taking tendencies. Male drivers were significantly more likely to be involved in fatal incidents (β = +2.7039, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001), reflecting global gender-based disparities in driving behaviour and crash involvement [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e, \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e]. \u003cem\u003eTemporal Patterns\u003c/em\u003e. Fatality rates tended to rise during evening hours (19:00\u0026ndash;00:00) and in the summer and autumn seasons, periods commonly associated with greater traffic density, fatigue, and travel activity. Early morning hours and midweek days were comparatively safer, suggesting time-specific patterns that could inform enforcement efforts [\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e].\u003c/p\u003e\u003cp\u003e\u003cem\u003eRoad Infrastructure.\u003c/em\u003e Higher fatality risks were observed on straight and interurban roads (β = +2.5534 and +\u0026thinsp;1.6233), while roundabouts and T-intersections exhibited strong protective effects (β = \u0026minus;\u0026thinsp;4.3028 and \u0026minus;\u0026thinsp;2.0496). These findings reinforce the established role of traffic\u0026ndash;calming infrastructure in reducing crash severity [\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e]. The CMP-Poisson model effectively addressed the observed over\u0026ndash;dispersion and provided more reliable estimates than the standard Poisson or Poisson-lognormal models. Its use in this context underscores the value of adopting flexible modelling strategies in traffic safety research when equi\u0026ndash;dispersion assumptions are violated. Overall, the results highlight the need for targeted interventions addressing unlicensed drivers, young male demographics, high\u0026ndash;risk periods, and specific roadway types.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e5.2 Limitations and Future Research\u003c/h2\u003e\u003cp\u003eThis study is subject to several limitations. First, the use of aggregated group-level data may mask important within-group variations, limiting the ability to detect individual-level risk patterns. Additionally, the absence of critical variables such as vehicle type, alcohol or drug involvement, and enforcement intensity constrains the comprehensiveness of the analysis.\u003c/p\u003e\u003cp\u003eFuture research should consider multilevel modelling frameworks that account for regional and individual-level variability, allowing for a more nuanced understanding of crash determinants. Incorporating longitudinal data would also enable the evaluation of temporal trends and the effectiveness of safety policies over time. Moreover, integrating additional variables such as weather conditions, vehicle specifications, and emergency response times could enhance model accuracy.\u003c/p\u003e\u003cp\u003eFurther studies should also examine the spatial and temporal distribution of fatal crashes using geospatial methods and explore the impact of recent infrastructure and legislative interventions in Albania. Access to more detailed, individual-level crash records would be essential for developing targeted, evidence-based road safety strategies.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003e5.3 Conclusions\u003c/h2\u003e\u003cp\u003eThis study examined fatal road traffic accidents in Albania between 2018 and 2024 using advanced count regression models to identify significant risk factors and evaluate model performance under various dispersion assumptions. Key demographic, temporal, and environmental variables were found to be significantly associated with fatal outcomes. Notably, over\u0026ndash;dispersion in the count data highlighted the necessity of employing flexible models such as the Conway\u0026ndash;Maxwell\u0026ndash;Poisson (CMP\u0026ndash;Poisson), which outperformed traditional alternatives in terms of fit and parameter stability [\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e]. Our findings regarding the influence of driver characteristics are consistent with previous research. Unlicensed and inexperienced drivers were associated with higher fatality risks, while those with more than nine years of experience showed a protective effect. This aligns with international evidence linking driving experience to improved hazard perception and reduced crash involvement [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. Similarly, our results identified young drivers, particularly those aged 25\u0026ndash;34, and male drivers as the most at-risk groups. These outcomes correspond with global studies that highlight young males as more prone to risk-taking behaviours and greater exposure to fatal crashes [\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eTemporal patterns observed in our study also resonate with earlier findings. We found that fatality risks increase during evening hours (19:00\u0026ndash;00:00) and in summer and autumn. These results echo prior studies emphasizing that night-time driving and seasonal variations (linked to higher traffic volumes, fatigue, or risky behaviour) significantly contribute to accident severity [\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e]. Such similarities reinforce the importance of targeted interventions during high-risk times. Road infrastructure factors emerged as significant determinants of fatal accidents. Straight and interurban roads showed higher fatality risks, whereas roundabouts and T-intersections provided protective effects. These results are in agreement with studies that demonstrate the traffic-calming role of roundabouts and improved intersection design in reducing crash severity [\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e]. Our findings add further evidence from Albania, underscoring the role of infrastructure interventions in mitigating risks.\u003c/p\u003e\u003cp\u003eFrom a methodological perspective, our study confirms the suitability of flexible regression models for analyzing road crash fatality data. The CMP regression model, which performed better than the standard Poisson and Poisson-lognormal models, effectively addressed the issue of over-dispersion in the dataset. This supports previous applications of the CMP model in traffic safety research, where it has been shown to yield more reliable estimates under similar data conditions [\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e]. The consistency of our findings with prior studies highlights both the robustness of the identified risk factors and the applicability of advanced statistical methods in capturing the complexity of accident data. At the same time, our analysis provides new evidence specific to Albania, where empirical studies remain limited, thereby contributing to a deeper understanding of risk determinants in developing contexts.\u003c/p\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for Publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003ch2\u003eFunding\u003c/h2\u003e\n\u003cp\u003eThis research received no funding. Hence the corresponding author hereby request 100% APC waiver for the publication of this article.\u003c/p\u003e\n\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\n\u003cp\u003eRaimonda Dervishi, , Data collection, Conceptualization, Model formulation, Methodology, Writing Original draft preparation, Analysis, Investigation.. Fabiana \u0026Ccedil;ullhaj, Literature review, Analysis, Investigation, Model formulation, Methodology, writing introduction, Editing Agbata, Benedict Celestine Model formulation, Methodology, Conceptualization, Methodology, Writing Original draft preparation, Analysis. Sander Kovaci, Editing, Literature review, writing introduction. Writing Original draft preparation.\u003c/p\u003e\n\u003ch2\u003eAcknowledgements\u003c/h2\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003ch2\u003eData Availability\u003c/h2\u003e\n\u003cp\u003eTable 2 and 3 of this publication present the data used in this investigation. Upon reasonable request, the associated author will provide the simulation\u0026apos;s source code.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eWorld Health Organization (2018). \u003cem\u003eGlobal status report on road safety 2018\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.who.int/publications/i/item/9789241565684\u003c/span\u003e\u003cspan address=\"https://www.who.int/publications/i/item/9789241565684\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen, S. (2019). The global macroeconomic burden of road injuries: Estimates and projections for 166 countries. \u003cem\u003eThe Lancet Planetary Health\u003c/em\u003e, \u003cem\u003e3\u003c/em\u003e(9), e390\u0026ndash;e398.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWorld Health Organization (2013). \u003cem\u003eProjections of mortality and causes of death, 2015 and 2030\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.who.int/healthinfo/\u003c/span\u003e\u003cspan address=\"http://www.who.int/healthinfo/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJoshi, A. K., Joshi, C., Singh, M., \u0026amp; Singh, V. (2014). Road traffic accidents in hilly regions of Northern India: What has to be done? \u003cem\u003eWorld Journal of Emergency Medicine\u003c/em\u003e, \u003cem\u003e5\u003c/em\u003e(2), 112\u0026ndash;115.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNyamuame, G. Y., Aglina, M. K., Akple, M. S., Philip, A., \u0026amp; Klomegah, W. (2015). Analysis of road traffic accidents trend in Ghana: Causing factors and preventive measures. \u003cem\u003eInternational Journal of Engineering Sciences \u0026amp; Management Research\u003c/em\u003e, 127\u0026ndash;132. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.ijesmr.com\u003c/span\u003e\u003cspan address=\"http://www.ijesmr.com\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAhmed, S., Hossain, M. A., \u0026amp; Sabuj, S. R. (2023). A study on road accident prediction and contributing factors using explainable machine learning models: Analysis and performance. \u003cem\u003eTransportation Research Interdisciplinary Perspectives\u003c/em\u003e, \u003cem\u003e19\u003c/em\u003e, 100814.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eArcher, J. (2005). \u003cem\u003eIndicators for Traffic Safety Assessment and Prediction and Their Application in Micro-Simulation Modelling: A Study of Urban and Suburban Intersections\u003c/em\u003e. Royal Institute of Technology.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEuropean Commission, Directorate-General for Mobility and Transport (2024). \u003cem\u003eTransport in the European Union \u0026ndash; Current trends and issues\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://data.europa.eu/doi/10.2832/131741\u003c/span\u003e\u003cspan address=\"https://data.europa.eu/doi/10.2832/131741\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDervishi, R., Kasharaj, J., \u0026amp; Kondi, I. (2021). Estimating of road accidents during the COVID-19 pandemic year in Albania. \u003cem\u003eInternational Journal of Ecosystems and Ecology Science\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(3), 489\u0026ndash;494. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.31407/ijees11.318\u003c/span\u003e\u003cspan address=\"10.31407/ijees11.318\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDervishi, R., Kovaci, S., Agbata, B. C., Gurmu, E. D., \u0026amp; Emadifar, H. (2025). Novel application of time-weighted statistical techniques for trend analysis in roadway accident fatalities: A case study from Albania. \u003cem\u003eEdelweiss Applied Science and Technology\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e(5), 2535\u0026ndash;2548. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.55214/25768484.v9i5.7516\u003c/span\u003e\u003cspan address=\"10.55214/25768484.v9i5.7516\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eUnited Nations Economic Commission for Europe (2018). \u003cem\u003eRoad safety performance review: Albania\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://digitallibrary.un.org/record/3813635?ln=en\u003c/span\u003e\u003cspan address=\"https://digitallibrary.un.org/record/3813635?ln=en\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSohail, A., Cheema, M. A., Ali, M. E., Toosi, A. N., \u0026amp; Rakha, H. A. (2023). Data-driven approaches for road safety: A comprehensive systematic literature review. \u003cem\u003eSafety Science\u003c/em\u003e, \u003cem\u003e158\u003c/em\u003e, 105949. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ssci.2022.105949\u003c/span\u003e\u003cspan address=\"10.1016/j.ssci.2022.105949\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eUlu, M., \u0026amp; \u0026amp;T\u0026uuml;rkan, Y. S. (2024). Bibliometric analysis of traffic accident prediction studies from 2003 to 2023: Trends, patterns and future directions. \u003cem\u003ePromet \u0026ndash; Traffic \u0026amp; Transportation\u003c/em\u003e, \u003cem\u003e36\u003c/em\u003e(5), 833\u0026ndash;851. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.7307/ptt.v36i5.576\u003c/span\u003e\u003cspan address=\"10.7307/ptt.v36i5.576\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZheng, L., Sayed, T., \u0026amp; Mannering, F. (2020). Modeling traffic conflicts for use in road safety analysis: A review of analytic methods and future directions. \u003cem\u003eAnalytic Methods in Accident Research\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.amar.2020.100142\u003c/span\u003e\u003cspan address=\"10.1016/j.amar.2020.100142\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePark, S. (2012). Analysis of injury severity in traffic crashes: A case study of Korean expressways. \u003cem\u003eKSCE Journal of Civil Engineering\u003c/em\u003e, \u003cem\u003e16\u003c/em\u003e(7), 1280\u0026ndash;1288. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s12205-012-1527-3\u003c/span\u003e\u003cspan address=\"10.1007/s12205-012-1527-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKovaci, S., Meco, Z. M., Agbata, B. C., Malotaj, M., \u0026amp; Dervishi, R. (2025). A data-driven study of road traffic accidents in Albania: Nonlinear regression applications in the transportation system. \u003cem\u003eEdelweiss Applied Science and Technology\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e(5), 1545\u0026ndash;1556. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://dx.doi.org/10.55214/25768484.v9i5.7215\u003c/span\u003e\u003cspan address=\"10.55214/25768484.v9i5.7215\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMinistry of Infrastructure and Energy (2020). \u003cem\u003eSectoral strategy of transport \u0026amp; action plan 2016\u0026ndash;2020: 3rd monitoring report\u003c/em\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWali, B., Khattak, A. J., \u0026amp; Karnowski, T. (2020). The relationship between driving volatility in time to collision and crash-injury severity in a naturalistic driving environment. \u003cem\u003eAnalytic Methods in Accident Research\u003c/em\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAmirifar, H., Ghahfarkhi, M. R., \u0026amp; \u0026amp;Chaleshtari, A. H. (2018). A study of factors affecting road accidents in Isfahan Province. In \u003cem\u003eThe Second International Conference on New Horizons in Engineering Sciences\u003c/em\u003e, Iran.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSingh, D., Singh, S. P., Kumaran, M., \u0026amp; Goel, S. (2015). Epidemiology of road traffic accident deaths in children in Chandigarh zone of North West India. \u003cem\u003eEgyptian Journal of Forensic Sciences\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e, 255\u0026ndash;260. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ejfs.2015.01.008\u003c/span\u003e\u003cspan address=\"10.1016/j.ejfs.2015.01.008\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRasool, F. A. A., Alekri, F. A., Nabi, H. A., Naiser, M. J., Shamlooh, N. M., Alnashaba, S. A., Alherz, Z. A., et al. (2015). Prevalence and behavioral risk factors associated with road traffic accidents among medical students of Arabian Gulf University in Bahrain. \u003cem\u003eInternational Journal of Medical Science and Public Health\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e, 933\u0026ndash;938. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5455/ijmsph.2015.14022015189\u003c/span\u003e\u003cspan address=\"10.5455/ijmsph.2015.14022015189\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZagorac, S., Markovic-Denic, L., \u0026amp; Nikolic, V. (2025). Trends in road traffic crash fatalities in Belgrade: A twelve-year retrospective analysis (2010\u0026ndash;2021). \u003cem\u003eJournal of Prevention\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10935-025-00846-x\u003c/span\u003e\u003cspan address=\"10.1007/s10935-025-00846-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDervishi, R., \u0026amp; Cenaj, E. (2021). Traffic accident characteristics of Albania and their prediction using Bayesian approach. \u003cem\u003eInternational Scientific Journal Trans \u0026amp; MOTAUTO World\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e(4), 151\u0026ndash;153.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRa\u0026ccedil;o, E., Beliu, E., Haxhi, K., \u0026amp; \u0026amp;Za\u0026ccedil;aj, O. (2025). Statistical modeling of weather impact on road accidents: A nonlinear distributed lag analysis in Tirana, Albania. \u003cem\u003eWSEAS Transactions on Environment and Development\u003c/em\u003e, \u003cem\u003e21\u003c/em\u003e, 196\u0026ndash;205. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.37394/232015.2025.21.17\u003c/span\u003e\u003cspan address=\"10.37394/232015.2025.21.17\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eInstitute of Statistics Republic of Albania (INSTAT). \u003cem\u003eStatistics of road accidents\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.instat.gov.al/\u003c/span\u003e\u003cspan address=\"https://www.instat.gov.al/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAbdulhafedh, A. (2017). Road accident prediction models: Different statistical modeling approaches. \u003cem\u003eJournal of Transportation Technologies\u003c/em\u003e, \u003cem\u003e7\u003c/em\u003e(2), 190\u0026ndash;205. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.4236/jtts.2017.72014\u003c/span\u003e\u003cspan address=\"10.4236/jtts.2017.72014\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAmerican Association of State Highway and Transportation Officials (AASHTO) (2010). \u003cem\u003eHighway Safety Manual\u003c/em\u003e. Washington, D.C.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFrempong, N. K., Yirenkyi, P. A., Nakua, E., \u0026amp; Dontwi, I. K. (2016). Modeling Road accident Fatalities in a Subregion of Ghana. \u003cem\u003eAmerican Journal of Scientific and Industrial Research\u003c/em\u003e, \u003cem\u003e7\u003c/em\u003e(4), 88\u0026ndash;94. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5251/ajsir.2016.7.4.88.94\u003c/span\u003e\u003cspan address=\"10.5251/ajsir.2016.7.4.88.94\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLord, D., Guikema, S. D., \u0026amp; Geedipally, S. R. (2008). Application of the Conway-Maxwell-Poisson generalized linear model for analyzing motor vehicle crashes. \u003cem\u003eAccident Analysis \u0026amp; Prevention\u003c/em\u003e, \u003cem\u003e40\u003c/em\u003e(3), 1123\u0026ndash;1134.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAbdel-Aty, M., \u0026amp; Radwan, A. E. (2000). Modeling traffic accident occurrence and involvement. \u003cem\u003eAccident Analysis \u0026amp; Prevention\u003c/em\u003e, \u003cem\u003e32\u003c/em\u003e(5), 633\u0026ndash;642. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0001-4575(99)00094-9\u003c/span\u003e\u003cspan address=\"10.1016/S0001-4575(99)00094-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJovanis, P., \u0026amp; Chang, H. (1986). Modeling the relationship of accidents to miles traveled. \u003cem\u003eTransportation Research Record\u003c/em\u003e, \u003cem\u003e1068\u003c/em\u003e, 42\u0026ndash;51.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAl-Ghamdi, A. S. (1994). Comparison of Accident Rates Using the Likelihood Ratio Testing Technique. \u003cem\u003eTransportation Research Record\u003c/em\u003e, \u003cem\u003e1401\u003c/em\u003e, 50\u0026ndash;54.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAl-Ghamdi, A. S. (2002). Using logistic regression to estimate the influence of accident factors on accident severity. \u003cem\u003eAccident Analysis \u0026amp; Prevention\u003c/em\u003e, \u003cem\u003e34\u003c/em\u003e(6), 729\u0026ndash;741.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEluru, N., Bhat, C. R., \u0026amp; Hensher, D. A. (2008). A mixed generalized ordered response model for examining pedestrian and bicyclist injury severity level in traffic crashes. \u003cem\u003eAccident Analysis \u0026amp; Prevention\u003c/em\u003e, \u003cem\u003e40\u003c/em\u003e(3), 1033\u0026ndash;1054.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eIslam, M. (2024). Assessing the effects of geometry and non-geometry related factors in work-zone crashes. \u003cem\u003eTraffic Injury Prevention\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/15389588.2024.2321914\u003c/span\u003e\u003cspan address=\"10.1080/15389588.2024.2321914\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHermans, E., Brijs, T., Wets, G., \u0026amp; Vanhoof, K. (2006). Explanatory modeling of traffic crash data using the COM\u0026ndash;Poisson model. \u003cem\u003eTransportation Research Board 85th Annual Meeting\u003c/em\u003e, Paper No. 06-1034.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSimmachan, T., Wongsai, N., Wongsai, S., \u0026amp; Lerdsuwansri, R. (2022). Modeling road accident fatalities with underdispersion and zero-inflated counts. \u003cem\u003ePlos One\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e(11), e0269022. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0269022\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0269022\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eUsman, A., \u0026amp; Alhassan, S. (2024). Fit a Poisson regression, negative binomial regression and Poisson log-normal regression models: An application to road accident data in North Central Nigeria. \u003cem\u003eInternational Journal of Modeling and Applied Science Research\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e(9). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://cambridgeresearchpub.com/ijmasr/article/view/412\u003c/span\u003e\u003cspan address=\"https://cambridgeresearchpub.com/ijmasr/article/view/412\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eH\u0026oslash;ye, A. (2016). The effects of speed on accidents and accident severity: A review. \u003cem\u003eAccident Analysis \u0026amp; Prevention\u003c/em\u003e, \u003cem\u003e95\u003c/em\u003e, 162\u0026ndash;170. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.aap.2016.07.006\u003c/span\u003e\u003cspan address=\"10.1016/j.aap.2016.07.006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWorld Health Organization (WHO). (2023). \u003cem\u003eGlobal status report on road safety 2023\u003c/em\u003e. World Health Organization. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.who.int/publications/i/item/9789240072250\u003c/span\u003e\u003cspan address=\"https://www.who.int/publications/i/item/9789240072250\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEuropean Transport Safety Council (2023). \u003cem\u003ePIN Flash Report 43: Reducing deaths in single-vehicle collisions\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://etsc.eu/pin\u003c/span\u003e\u003cspan address=\"https://etsc.eu/pin\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eInternational Road Assessment Programme (iRAP) (2024). \u003cem\u003eAlbania cuts road deaths by 23% with innovative road maintenance programme\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://irap.org/2024/02/albania-cuts-road-deaths-by-23-with-innovative-road-maintenance-programme/\u003c/span\u003e\u003cspan address=\"https://irap.org/2024/02/albania-cuts-road-deaths-by-23-with-innovative-road-maintenance-programme/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWorld Health Organization. (2015). \u003cem\u003eGlobal status report on road safety 2015\u003c/em\u003e. WHO. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.who.int/publications/i/item/9789241565066\u003c/span\u003e\u003cspan address=\"https://www.who.int/publications/i/item/9789241565066\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCenaj, E., \u0026amp; Dervishi, R. (2024). Road accident fatalities forecasting models using Smeed\u0026rsquo;s regression analysis: A case study. \u003cem\u003eEuropean Journal of Engineering and Technology Research\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e(6), 20\u0026ndash;24. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.24018/ejeng.2024.9.6.3213\u003c/span\u003e\u003cspan address=\"10.24018/ejeng.2024.9.6.3213\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhou, M., \u0026amp; Sisiopiku, V. P. (1997). Relationship between volume-to-capacity ratios and accident rates. \u003cem\u003eTransportation Research Record\u003c/em\u003e, \u003cem\u003e1581\u003c/em\u003e(1), 47\u0026ndash;52. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3141/1581-07\u003c/span\u003e\u003cspan address=\"10.3141/1581-07\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKassa, S. M., Lachore, T. S., Fetene, A. T., Sewnet, H., \u0026amp; \u0026amp;Demlew, M. B. (2023). Developing a correlation model between the frequency of road traffic accidents and road geometric design elements. \u003cem\u003ePublic Infrastructure Studies\u003c/em\u003e, \u003cem\u003e70\u003c/em\u003e(4). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.31075/PIS.70.04.01\u003c/span\u003e\u003cspan address=\"10.31075/PIS.70.04.01\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBasu, S., \u0026amp; Saha, P. (2017). Regression models of highway traffic crashes: A review of recent research and future research needs. \u003cem\u003eProcedia Engineering\u003c/em\u003e, \u003cem\u003e187\u003c/em\u003e, 59\u0026ndash;66. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.proeng.2017.04.351\u003c/span\u003e\u003cspan address=\"10.1016/j.proeng.2017.04.351\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSellers, K. F., \u0026amp; Premeaux, B. (2021). Conway\u0026ndash;Maxwell\u0026ndash;Poisson regression models for dispersed count data. \u003cem\u003eWiley Interdisciplinary Reviews: Computational Statistics\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(3), e1533. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/wics.1533\u003c/span\u003e\u003cspan address=\"10.1002/wics.1533\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAbdella, G. M., Kim, J., Al-Khalifa, K. N., \u0026amp; Hamouda, A. M. (2019). Penalized Conway-Maxwell-Poisson regression for modelling dispersed discrete data: The case study of motor vehicle crash frequency. \u003cem\u003eSafety Science\u003c/em\u003e, \u003cem\u003e120\u003c/em\u003e, 157\u0026ndash;163. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ssci.2019.06.036\u003c/span\u003e\u003cspan address=\"10.1016/j.ssci.2019.06.036\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi, X., Lord, D., Zhang, Y., \u0026amp; Xie, Y. (2009). Predicting motor vehicle crashes using support vector machine models. \u003cem\u003eAccident Analysis \u0026amp; Prevention\u003c/em\u003e, \u003cem\u003e40\u003c/em\u003e(5), 1611\u0026ndash;1618. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.aap.2009.07.009\u003c/span\u003e\u003cspan address=\"10.1016/j.aap.2009.07.009\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLord, D. (2006). Modeling motor vehicle crashes using Poisson-gamma models: Examining the effects of low sample mean values and small sample size on the estimation of the fixed dispersion parameter. \u003cem\u003eAccident Analysis \u0026amp; Prevention\u003c/em\u003e, \u003cem\u003e38\u003c/em\u003e(4), 751\u0026ndash;766. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.aap.2005.11.008\u003c/span\u003e\u003cspan address=\"10.1016/j.aap.2005.11.008\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMitra, S., \u0026amp; Washington, S. (2007). On the nature of over-dispersion in motor vehicle crash prediction models. \u003cem\u003eAccident Analysis \u0026amp; Prevention\u003c/em\u003e, \u003cem\u003e39\u003c/em\u003e(3), 459\u0026ndash;468. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.aap.2006.08.002\u003c/span\u003e\u003cspan address=\"10.1016/j.aap.2006.08.002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKhan, N. M., Soobhug, A. D., \u0026amp; Jannoo, Z. (2020). Modeling road traffic accidents in Mauritius using clustered longitudinal COM-Poisson with gamma random effects. \u003cem\u003eCommunications in Statistics: Case Studies, Data Analysis and Applications, 7\u003c/em\u003e(2), 113\u0026ndash;127. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/23737484.2020.1842269\u003c/span\u003e\u003cspan address=\"10.1080/23737484.2020.1842269\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAbdulla, R., Qader, B., \u0026amp; Sdiq, K. (2023). Traffic Accident Traits and Driver Characteristics Implication on Road Accidents using Descriptive Analysis: A Cross Sectional Study in Sulaymaniyah, Iraq. \u003cem\u003eEngineering Technology and Applied Science Research\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(2), 10372\u0026ndash;10376. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://dx.doi.org/10.48084/etasr.5669\u003c/span\u003e\u003cspan address=\"10.48084/etasr.5669\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFowode, K. V., Nwaogazie, I. L., \u0026amp; Anyanwu, B. O. (2023). Prevalence and Factors Causing Road Traffic Accidents in Lagos State, Nigeria. \u003cem\u003eAsian Journal of Advanced Research and Reports\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e(6), 36\u0026ndash;44. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.9734/ajarr/2023/v17i6488\u003c/span\u003e\u003cspan address=\"10.9734/ajarr/2023/v17i6488\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAkaike, H. (1974). A new look at the statistical model identification. \u003cem\u003eIEEE Transactions on Automatic Control\u003c/em\u003e, \u003cem\u003e19\u003c/em\u003e(6), 716\u0026ndash;723. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/TAC.1974.1100705\u003c/span\u003e\u003cspan address=\"10.1109/TAC.1974.1100705\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSchwarz, G. (1978). Estimating the dimension of a model. \u003cem\u003eThe Annals of Statistics\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e(2), 461\u0026ndash;464. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1214/aos/1176344136\u003c/span\u003e\u003cspan address=\"10.1214/aos/1176344136\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBurnham, K. P., \u0026amp; Anderson, D. R. (2002). \u003cem\u003eModel Selection and Multimodel Inference: A Practical Information-Theoretic Approach\u003c/em\u003e (2nd ed.). Springer.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eR Core Team. (2024). \u003cem\u003eR: A Language and Environment for Statistical Computing\u003c/em\u003e. R Foundation for Statistical Computing. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.R-project.org/\u003c/span\u003e\u003cspan address=\"https://www.R-project.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBrooks, M. E., Kristensen, K., van Benthem, K. J., Magnusson, A., Berg, C. W., Nielsen, A., Skaug, H. J., Maechler, M., \u0026amp; Bolker, B. M. (2017). glmmTMB balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling. \u003cem\u003eThe R Journal\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e(2), 378\u0026ndash;400. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.32614/RJ-2017-066\u003c/span\u003e\u003cspan address=\"10.32614/RJ-2017-066\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eL\u0026uuml;decke, D., Ben-Shachar, M., Patil, I., Waggoner, P., \u0026amp; Makowski, D. (2021). performance: An R package for assessment, comparison and testing of statistical models. \u003cem\u003eJournal of Open Source Software\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e(60), 3139. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.21105/joss.03139\u003c/span\u003e\u003cspan address=\"10.21105/joss.03139\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWickham, H. (2016). \u003cem\u003eggplot2: Elegant graphics for data analysis\u003c/em\u003e. Springer. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://ggplot2.tidyverse.org\u003c/span\u003e\u003cspan address=\"https://ggplot2.tidyverse.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShmueli, G., Minka, T. P., Kadane, J. B., Borle, S., \u0026amp; Boatwright, P. (2005). A useful distribution for fitting discrete data: Revival of the Conway-Maxwell-Poisson distribution. \u003cem\u003eJournal of the Royal Statistical Society: Series C (Applied Statistics)\u003c/em\u003e, \u003cem\u003e54\u003c/em\u003e(1), 127\u0026ndash;142. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1467-9876.2005.00474.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1467-9876.2005.00474.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSellers, K. F., \u0026amp; Shmueli, G. (2010). A flexible regression model for count data. \u003cem\u003eThe Annals of Applied Statistics\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e(2), 943\u0026ndash;961. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1214/10-AOAS325\u003c/span\u003e\u003cspan address=\"10.1214/10-AOAS325\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eElvik, R. (2010). The stability of long-term trends in the real risk of traffic injury for different road user groups. \u003cem\u003eAccident Analysis \u0026amp; Prevention\u003c/em\u003e, \u003cem\u003e42\u003c/em\u003e(2), 523\u0026ndash;532. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.aap.2009.09.001\u003c/span\u003e\u003cspan address=\"10.1016/j.aap.2009.09.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMoran, C., Bennett, J. M., \u0026amp; Prabhakharan, P. (2020). The relationship between cognitive function and hazard perception in younger drivers. \u003cem\u003eTransportation Research Part F: Traffic Psychology and Behaviour\u003c/em\u003e, \u003cem\u003e74\u003c/em\u003e, 104\u0026ndash;119.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAckaah, W., \u0026amp; Salifu, M. (2011). Crash prediction model for two-lane rural highways in the Ashanti Region of Ghana. \u003cem\u003eIATSS Research\u003c/em\u003e, \u003cem\u003e35\u003c/em\u003e(1), 34\u0026ndash;40. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.iatssr.2011.02.001\u003c/span\u003e\u003cspan address=\"10.1016/j.iatssr.2011.02.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFederal Highway Administration (2017). \u003cem\u003eRoundabouts: An informational guide\u003c/em\u003e (2nd ed., FHWA-SA-17-067). U.S. Department of Transportation. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://safety.fhwa.dot.gov/intersection/roundabouts/\u003c/span\u003e\u003cspan address=\"https://safety.fhwa.dot.gov/intersection/roundabouts/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHilbe, J. M. (2011). \u003cem\u003eNegative binomial regression\u003c/em\u003e (2nd ed.). Cambridge University Press. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1017/CBO9780511973420\u003c/span\u003e\u003cspan address=\"10.1017/CBO9780511973420\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Road crash fatalities (RCF), Statistical modelling, Driver behaviours, CMP regression, Over–dispersed data","lastPublishedDoi":"10.21203/rs.3.rs-7860402/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7860402/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe global escalation in road crash has led to a substantial public health burden, with annual fatalities reaching millions worldwide and causing extensive economic and social disruption. While existing research has largely focused on predicting the likelihood of road crash fatalities as a classification problem, relatively few studies have investigated the underlying relationships among the complex factors contributing to such fatalities. This study aims to identify the significant determinants of road crash fatalities in Albania by employing advanced modelling approaches. The analysis is based on official statistics from the Albanian Institute of Statistics (\u003cem\u003eINSTAT\u003c/em\u003e) covering seven years from 2018 to 2024. During this period, a total of 13,203 individuals were involved in road traffic accidents, with 1,349 fatalities recorded. Although the overall number of accidents has decreased, the fatality rate has shown a concerning upward trend.To better understand the dynamics of road crash fatalities (\u003cem\u003eRAF\u003c/em\u003es), this study examines the impact of driver behaviours, time factors, and road characteristics. Two regression approaches Poisson\u0026ndash;Lognormal (\u003cem\u003ePL\u003c/em\u003e) and Conway\u0026ndash;Maxwell\u0026ndash;Poisson (\u003cem\u003eCMP\u003c/em\u003e) regression models are applied to model the number of fatalities concerning these key factors. These models are particularly suited to handling the over\u0026ndash;dispersed nature of the data, in contrast to the traditional Poisson regression model, which assumes equi\u0026ndash;dispersion. The CMP-Poisson model outperformed both, offering better model fit and more robust estimation under dispersion, validating its application in traffic fatality research with count data.By quantifying the effects of these risk factors, this study provides valuable insights for the development of targeted interventions and policy strategies aimed at reducing road crash fatalities in Albania.\u003c/p\u003e","manuscriptTitle":"Modelling Road Crash Fatalities in Albania: A Cross-Sectional Study of Risk Factors Using Advanced Count Regression","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-11-28 06:56:02","doi":"10.21203/rs.3.rs-7860402/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"00b68676-d0b3-4e4f-8a5f-8ae6bfc732f5","owner":[],"postedDate":"November 28th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-11-28T06:56:02+00:00","versionOfRecord":[],"versionCreatedAt":"2025-11-28 06:56:02","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7860402","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7860402","identity":"rs-7860402","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}