Truck drivers had the highest number of occupational injuries and illnesses causing time away from work during 1992-97 and the third highest rates. Except for back injuries, almost no information is available on risk factors for such injuries. Neither the role of personal factors nor that of motor carrier (trucking firm) operating characteristics and vehicle features is understood. This study brings together experts from truck transportation, industrial relations, occupational medicine/epidemiology and biostatistics to link and analyze data sets that have not been combined previously to investigate factors associated with truck driver injury and outcome. The study will: 1) calculate incidence rates by motor carrier operating characteristics, fleet size and truck configuration; 2) estimate medical care use in different medical care settings by social-demographic, truck firm and specific medical diagnosis and model such use with logistic or Poisson regression; 3) model outcome measured by lost work-time using logistic, Poisson and Cox models; and 4) calculate the predictive value of the resulting models by appropriate methodology. The compensation information will be obtained from a major industrialized State's workers' compensation bureau. The data covers about 60-65% of the state workers, including all workers in firms with 2-500 workers, with the balance of the workforce covered by self-insured employers. Data from the workers' compensation agency include social-demographic, accident characteristics, medical care use and compensation for lost work-time. The study population consists of about 12,000 workers who had work-related injuries or illnesses occurring in 1996-98. Follow-up will be available through mid-2001. Motor carrier operating characteristics and vehicle features will be obtained from federal or state regulatory agencies. Truck crash information will also be obtained from federal or state sources. Incidence rates are calculated by linking compensation and truck firm data. Logistic, Poisson and Cox models are used to test hypotheses and identify significant factors in incidence, medical care use and outcome. Predictive value is evaluated by ROC curves, the Bayes Information Criteria or other recommended approaches.
