Creatinine normalization of workplace urine drug tests: does it make a difference?

Background: This study examines the effect of creatinine normalization on urine drug concentrations of 5 substances (amphetamines, cocaine, marijuana, opiates, and phencyclidine) and how this affects the proportion of reported positives.

Methods: The Wilcoxon matched-pairs signed-ranks test was used to compare the mean prenormalization urinary drug concentration with the mean postnormalization urinary drug concentration. Frequency analysis was performed on dichotomous drug test results and the information was used to complete McNemar testing for each drug to determine the difference of proportions for prenormalization positive drug tests to postnormalization positive drug test.

Results: Each drug tested (N = 4460) was found to have a statistically significant increase in mean urinary drug concentration after creatinine normalization with effect sizes ranging from small to medium with cocaine having the largest effect size (r = 0.229) and phencyclidine having the lowest effect size (r = 0.121). The differences in proportion of dichotomous results between study and control groups for drugs tested were compared with the McNemar test. Each drug had a statistically significant (P = 0.0010) increase of positive drug tests.

Conclusions: This result indicates that specimen dilution does affect the number of laboratory-positive results confirmed.