Dilution of urine drug tests: is it random?

The purpose of this study was to determine whether the act of in vivo urine dilution is a random phenomenon.

The sample population was divided into 4 groups for each of the drugs tested. The groups are precreatinine normalization laboratory positive, precreatinine normalization laboratory negative, and postcreatinine normalization laboratory drug positive, and postcreatinine normalization laboratory drug negative. The nonparametric Kruskal-Wallis 1-way ANOVA by ranks with post hoc Mann-Whitney U testing Bonferroni adjustment was used to compare the mean urine creatinine concentration of each group.

The Kruskal-Wallis test revealed that each drug had a statistically significant difference in group comparisons. Post hoc testing found that cocaine demonstrated no statistically significant difference between the prenormalization drug positive and negative groups. Amphetamines, marijuana, and opiates did confirm statistically significant differences between the prenormalization drug positive and negative groups. The examination of phencyclidine was limited by the lack of a prenormalization drug positive group; therefore, no definitive findings can be presented. Statistically significant differences were found between the postnormalization drug positive and negative groups for each drug tested.

This study supports the theory that urine dilution is not a random happening and it may adversely affect the integrity of the drug-testing process.