James W Price1. 1. St. Mary's Occupational Medicine Clinic, Evansville; and Marian University College of Osteopathic Medicine, Indianapolis, IN.
Abstract
INTRODUCTION: If employee drug use plays a significant part in the incidence of workplace accidents, one would anticipate the positivity rates of postaccident drug tests to be higher than the positivity rates for random tests. Past studies examined the difference of a dichotomous outcome between 2 groups. Dichotomous (positive or negative) categorization may have been a source of systemic error, which minimized the difference between random and postaccident groups. METHODS: This is a study exploring the difference of urine drug concentrations between random and postaccident groups. The population consists of employees from various industries in Southern Indiana, Western Kentucky, and Eastern Illinois. Normalization of drug excretion to urinary creatinine concentration was carried out. Preliminary assumption testing was done. Logarithmic transformation was completed. One-way between-groups analysis of variance was performed to investigate random and postaccident drug test differences in urine drug concentrations. RESULTS: There was a statistically significant difference between test groups on the combined dependent variables. The only dependent variables to reach statistical significance, to an alpha level of 0.025, were urine opiate and urine amphetamine concentrations. Higher mean concentration of both drugs was found in the postaccident group. However, the proportion of variances explained by the reason for testing was quite small. CONCLUSIONS: The results of this study are consistent with studies using dichotomous dependent variables, which demonstrated a significantly higher prevalence for positive postaccident urine opiate tests. The study also identified a similar result for postaccident urine amphetamine tests which was not previously known.
INTRODUCTION: If employee drug use plays a significant part in the incidence of workplace accidents, one would anticipate the positivity rates of postaccident drug tests to be higher than the positivity rates for random tests. Past studies examined the difference of a dichotomous outcome between 2 groups. Dichotomous (positive or negative) categorization may have been a source of systemic error, which minimized the difference between random and postaccident groups. METHODS: This is a study exploring the difference of urine drug concentrations between random and postaccident groups. The population consists of employees from various industries in Southern Indiana, Western Kentucky, and Eastern Illinois. Normalization of drug excretion to urinary creatinine concentration was carried out. Preliminary assumption testing was done. Logarithmic transformation was completed. One-way between-groups analysis of variance was performed to investigate random and postaccident drug test differences in urine drug concentrations. RESULTS: There was a statistically significant difference between test groups on the combined dependent variables. The only dependent variables to reach statistical significance, to an alpha level of 0.025, were urine opiate and urine amphetamine concentrations. Higher mean concentration of both drugs was found in the postaccident group. However, the proportion of variances explained by the reason for testing was quite small. CONCLUSIONS: The results of this study are consistent with studies using dichotomous dependent variables, which demonstrated a significantly higher prevalence for positive postaccident urine opiate tests. The study also identified a similar result for postaccident urine amphetamine tests which was not previously known.