Application and interpretation of statistical evaluation of relationships is a necessary element in biomedical research. Statistical analyses rely on P value to demonstrate relationships. The traditional level of significance, P<0.05, can be negatively impacted by small sample size, bias, and random error, and has evolved to include interpretation of statistical trends, correction factors for multiple analyses, and acceptance of statistical significance for P>0.05 for complex relationships such as effect modification.

Objective This study assesses relationships between the Framingham Cardiovascular Disease Risk (CVD risk) Score and prevalence of US Department of Transportation (DOT)-reportable crashes in commercial motor vehicle (CMV) drivers, after controlling for potential confounders. Methods Data were analyzed from CMV drivers (N=797) in a large cross-sectional study. CVD risk was calculated for each driver. Adjusted odds ratios (OR) and 95% Confidence Intervals (95% CI) between CVD risk and DOT-reportable crashes were calculated. Results Drivers in the two highest CVD risk groups had significantly higher likelihood of crash (OR=2.08, 95% CI=1.20-3.63 and OR=1.99, 95% CI=1.05-3.77, respectively) after adjusting for confounders. There was a significant trend of increasing prevalence of crashes with an increasing CVD risk score (p=0.0298). Conclusion Drivers with high CVD risk had a higher likelihood of a crash after controlling for confounders.