Identification and validation of a logistic regression model for predicting serious injuries associated with motor vehicle crashes

2011 ◽  
Vol 43 (1) ◽  
pp. 112-122 ◽  
Author(s):  
Douglas W. Kononen ◽  
Carol A.C. Flannagan ◽  
Stewart C. Wang
Keyword(s):  
Logistic Regression ◽  
Regression Model ◽  
Logistic Regression Model ◽  
Motor Vehicle ◽  
Motor Vehicle Crashes ◽  
Vehicle Crashes ◽  
Serious Injuries

Risky Driving: Relationship Between Cellular Phone and Safety Belt Use

2003 ◽  
Vol 1843 (1) ◽  
pp. 20-23 ◽  
Author(s):  
David W. Eby ◽  
Lidia P. Kostyniuk ◽  
Jonathon M. Vivoda
Keyword(s):  
Logistic Regression ◽  
Regression Model ◽  
Logistic Regression Model ◽  
Motor Vehicle ◽  
Cellular Phone ◽  
Motor Vehicle Crash ◽  
Public Health Issue ◽  
Cellular Phones ◽  
Safety Belt ◽  
Phone User

The main purpose of the study was to explore the relationship between cellular phone and safety belt use. Rates of safety belt use of drivers using and drivers not using handheld cellular phones were compared. All data for safety belt and handheld cellular phone use were collected through direct observation while vehicles were stopped at intersections and freeway exit ramps in Michigan. Data were weighted to be representative of drivers during daylight hours in Michigan. Analyses included statistical comparisons of safety belt use rates and a logistic regression model to determine the effects of handheld cellular phone use on safety belt use. The study found that safety belt use for drivers using a handheld cellular phone was significantly lower than for drivers not using cellular phones. This same significant relationship was found within nearly all demographic categories analyzed. The logistic regression model showed that the odds of a handheld cellular phone user not using a safety belt were 1.77 times that of a driver not using a cellular phone. These results stress the importance of the public health issue posed by cellular phone use; not only are those who are conversing on cellular phones potentially more likely to be in a motor vehicle crash, they are also more likely to sustain greater injury because of the lack of safety belt use.

scholarly journals Modeling and Analyzing the Factors Affecting the Mortality of Pedestrians in Urban Accidents

2018 ◽  
Vol 14 (4) ◽  
pp. 10-21
Author(s):  
Ahmet Tortum ◽  
Alireza Motamadnia
Keyword(s):  
Logistic Regression ◽  
Regression Model ◽  
Logistic Regression Model ◽  
Motor Vehicle ◽  
Binary Logistic Regression ◽  
Education Level ◽  
Road Surface ◽  
Binary Logistic Regression Model ◽  
Factors Affecting ◽  
Surface Conditions

Abstract The nature of urban and rural accidents has been different from each other in some of the factors and even the severity of damage rate, mayhem, and death. In this research, using statistical methods and binary logistic regression model, we have addressed to analyze important parameters such as age, gender, education level, the color of the pedestrian dress, season of accident, time of accident, the speed of the vehicle colliding with pedestrians and road surface conditions at the time of accident on the way of death (at the scene of the incident or in the hospital) pedestrians who have been traumatized. After the creation of the binary logistic regression model, it was determined that only the parameters of speed and the accident time have been significant in the level less than 5%. And other parameters such as age, gender, the season of accident occurrence, the color of the pedestrian dress, road surface conditions and education level had no significant effect in terms of statistical on the incidence of mortality arising from a pedestrian accident with the motor vehicle. The results revealed that by adopting decisions related to the traffic calming, attention to passages lighting and brightness the mortality rate of a pedestrian due to the urban accidents can be reduced.

Motor Vehicle Crashes Involving Railroad Crossing Signal Masts and Guardrails

2019 ◽  
Vol 2673 (12) ◽  
pp. 191-203
Author(s):  
Justin D. Cyr ◽  
John W. Shaw ◽  
Peter T. Savolainen
Keyword(s):  
United States ◽  
Regression Models ◽  
Injury Severity ◽  
Motor Vehicle ◽  
Motor Vehicle Crashes ◽  
Future Research ◽  
Vehicle Crashes ◽  
Principal Element ◽  
Crash Rates ◽  
Serious Injuries

The installation of “active” warning devices, such as crossing signals and gates, is a principal element of state and federal railroad crossing safety programs, and their effectiveness in preventing vehicle–train collision is well documented. While most active crossings are occupied by trains only a few minutes each day, non-crashworthy items are frequently present at active crossings including crossing signal masts, crossing gate mechanisms, cantilever supports, and signal controller bungalows. Various agencies utilize longitudinal guardrail systems to protect errant motorists from striking non-frangible crossing signal hardware. This study sought to determine whether such guardrails influence the prevalence and severity of motor vehicle strikes involving crossing hardware. A review of 10 years of police crash-report narratives in Iowa, United States (U.S.) indicates an average of 15.6 crashes per year at 1,853 crossings involving signal masts, signal guardrails, or related infrastructure. Over this period, one fatality and five serious injuries occurred as a result of these crashes. The crash-injury severity distributions appeared similar in crashes with and without guardrail, and regression models showed that guardrail presence did not affect the rate of crashes involving railroad apparatus. Although the use of longitudinal guardrail systems did not appear to have a strong effect on crash rates or severity at active grade crossings, other methods for reducing crash prevalence and severity could be explored in future research, such as the development of crashworthy signal assemblies for flashing-signal-only crossings and the redesign of gate and cantilever supports to increase separation from the traveled way.

scholarly journals A Predictive Model to Analyze the Factors Affecting the Presence of Traumatic Brain Injury in the Elderly Occupants of Motor Vehicle Crashes Based on Korean In-Depth Accident Study (KIDAS) Database

2021 ◽  
Vol 18 (8) ◽  
pp. 3975
Author(s):  
Hee Young Lee ◽  
Hyun Youk ◽  
Oh Hyun Kim ◽  
Chan Young Kang ◽  
Joon Seok Kong ◽  
...  
Keyword(s):  
Traumatic Brain Injury ◽  
Logistic Regression ◽  
Regression Analysis ◽  
Brain Injury ◽  
Predictive Model ◽  
Motor Vehicle ◽  
The Elderly ◽  
Motor Vehicle Crashes ◽  
Vehicle Crashes ◽  
Factors Affecting

Traumatic brain injury (TBI), also known as intracranial injury, occurs when an external force injures the brain. This study aimed to analyze the factors affecting the presence of TBI in the elderly occupants of motor vehicle crashes. We defined elderly occupants as those more than 55 years old. Damage to the vehicle was presented using the Collision Deformation Classification (CDC) code by evaluation of photos of the damaged vehicle, and a trauma score was used for evaluation of the severity of the patient’s injury. A logistic regression model was used to identify factors affecting TBI in elderly occupants and a predictive model was constructed. We performed this study retrospectively and gathered all the data under the Korean In-Depth Accident Study (KIDAS) investigation system. Among 3697 patients who visited the emergency room in the regional emergency medical center due to motor vehicle crashes from 2011 to 2018, we analyzed the data of 822 elderly occupants, which were divided into two groups: the TBI patients (N = 357) and the non-TBI patients (N = 465). According to multiple logistic regression analysis, the probabilities of TBI in the elderly caused by rear-end (OR = 1.833) and multiple collisions (OR = 1.897) were higher than in frontal collision. Furthermore, the probability of TBI in the elderly was 1.677 times higher in those with unfastened seatbelts compared to those with fastened seatbelts (OR = 1.677). This study was meaningful in that it incorporated several indicators that affected the occurrence of the TBI in the elderly occupants. In addition, it was performed to determine the probability of TBI according to sex, vehicle type, seating position, seatbelt status, collision type, and crush extent using logistic regression analysis. In order to derive more precise predictive models, it would be needed to analyze more factors for vehicle damage, environment, and occupant injury in future studies.

scholarly journals Factors Contributing to Police Attendance at Motor Vehicle Crash Scenes

2014 ◽  
Author(s):  
Richard Tay ◽  
Lina Kattan ◽  
Yuan Bai
Keyword(s):  
Logistic Regression ◽  
Regression Model ◽  
Logistic Regression Model ◽  
Motor Vehicle ◽  
Motor Vehicle Crash ◽  
Crash Data ◽  
Police Services ◽  
Vehicle Crash ◽  
Secondary Crashes ◽  
The City

Police attendance at a motor vehicle crash scene is important for investigating the causes of crashes, reducing secondary crashes, managing traffic, and reducing congestion. However, very little research has been conducted to examine the factors contributing to the likelihood of police attendance. This study hypothesizes that the policies of the police services concerned, convenience and comfort, and expectations of injuries or driver violations will increase the likelihood of police attendance at a crash scene. This conceptual framework is supported by the results from fitting a logistic regression model to crash data from the City of Calgary in Alberta, Canada.

scholarly journals Determinants of emergency department disposition of patients with traumatic brain injury in Uganda: results from a registry

2018 ◽  
Vol 3 (1) ◽  
pp. e000253 ◽  
Author(s):  
Amber Mehmood ◽  
Nukhba Zia ◽  
Olive Kobusingye ◽  
Rukia H Namaganda ◽  
Hussein Ssenyonjo ◽  
...  
Keyword(s):  
Emergency Department ◽  
Logistic Regression ◽  
Hospital Admission ◽  
Regression Model ◽  
Logistic Regression Model ◽  
Injury Severity ◽  
Multinomial Logistic Regression ◽  
Multinomial Logistic Regression Model ◽  
Trauma Score ◽  
Serious Injuries

BackgroundTraumatic brain injuries (TBIs) are a common cause of emergency department (ED) visits and hospital admissions in Kampala, Uganda. The objective of this study was to assess determinants of ED discharge disposition based on patient demographic and injury characteristics. Four ED outcomes were considered: discharge home, hospital admission, death, and others.MethodsThis prospective study was conducted at Mulago National Referral Hospital, Kampala, Uganda, from May 2016 to July 2017. Patients of all age groups presenting with TBI were included. Patient demographics, external causes of injury, TBI characteristics, and disposition from EDs were noted. Injury severity was estimated using the Glasgow Coma Scale (GCS), Kampala Trauma Score (KTS), and the Revised Trauma Score (RTS). A multinomial logistic regression model was used to estimate conditional ORs of hospital admission, death, and other dispositions compared with the reference category “discharged home”.ResultsA total of 3944 patients were included in the study with a male versus female ratio of 5.5:1 and a mean age of 28.5 years (SD=14.2). Patients had closed head injuries in 62.9% of cases. The leading causes of TBIs were road traffic crashes (58.8%) and intentional injuries (28.7%). There was no significant difference between the four discharge categories with respect to age, sex, mode of arrival, cause of TBI, place of injury, type of head injury, transport time, and RTS (p>0.05). There were statistically significant differences between the four discharge categories for a number of serious injuries, GCS on arrival, change in GCS, and KTS. In a multinomial logistic regression model, change in GCS, area of residence, number of serious injuries, and KTS were significant predictors of ED disposition.DiscussionThis study provides evidence that ED disposition of patients with TBI is differentially affected by injury characteristics and is largely dependent on injury severity and change in GCS during ED stay.Level of evidenceLevel II.

Sign in / Sign up

Export Citation Format

Share Document