Investigating the Relation between Functional Road Class and Hospital Diagnosed Injury Severity for Crashes in Maryland

The true toll of Motor Vehicle Crashes (MVC) lies in mortality and injuries that force lifelong changes, impacting mental health, occupation, means of attaining livelihood for self and family, and long-term chronic suffering. However, limited emphasis has been placed on morbidity as compared to mortality from MVC. This paper presents results from modeling the association between injury severities, Blood Alcohol Content (BAC), and roadway classes using data from 2016 - 2021 obtained from the R Adams Cowley Shock Trauma Center (STC) registry linked with police crash reports and the Functional Road Class (FRC) data at the patient level for Maryland. The record-level linkages between the STC and police data were produced by a linear combination of edit distance scores between common variables in either dataset, while geospatial proximity linkage was used to combine FRC data with incident locations. Ordered logistic regression model with partial proportional odds was used for determining impact of exogenous variables on injury severities. The model incorporates BAC obtained from blood screens at the time of arrival to the hospital. Moreover, the analytics presented use globally accepted consensus derived injury severities based on hospital diagnoses. The three elements taken together are unique to this study and are not replicated in existing literature, although they have been studied individually. Results show that facilities from major arterials to local roads have a higher risk for more severe injuries and longer hospital stay compared to facilities like interstates and freeways. Moreover, the risk of injuries that can cause permanent lifestyle changes due to disabilities is also higher on those roads.