Rates and types of urine drug screen false negative results compared with confirmatory toxicology testing in major trauma patients.

BACKGROUND

Trauma centers are required to screen patients for alcohol use, and if necessary, intervene and refer to treatment (SBIRT). Similar screening for illicit drug use is recommended but not required. Urine drug screening (UDS) underestimates problematic substance use. This study aimed to estimate the types and rates of UDS false negatives (FN) compared to comprehensive testing by liquid chromatography-mass spectrometry (LC-MS) in trauma patients.

METHODS

We performed a prospective cohort study of deidentified urine samples from adult trauma and burn activation patients. Both UDS and LC-MS comprehensive testing of >200 analytes were performed by a reference laboratory on all samples. Iatrogenic medications were excluded from the FN count. Crosstab analyses were conducted for UDS versus LC-MS outcomes to establish FN types and rates. We dichotomized the results by creating an "intentionality" variable (intentional injuries by self/others versus accidental injuries). A series of crosstabs with odds ratios considered intentionality by substance class and demographics. Statistically significant variables by Chi-Square were assessed by logistic regression.

RESULTS

Psychoactive FN were detected in 56/100 urine samples analyzed; the most frequent included anticonvulsants (primarily gabapentin, N = 13), opioid agonists (N = 12), antihistamines (primarily diphenhydramine, N = 10), and phenethylamines (primarily bupropion, N = 5). Nonpsychoactive FN were detected in 70/100 samples; the most common were nicotine (N = 33), caffeine (N = 23), acetaminophen (N = 22), and antidepressants (N = 12). Of substance classes included in the UDS and also tested by LC-MS, FN occurred for opiates (3%), amphetamines (5%) and opioids (25%). Polypharmacy was associated with fall injuries in elderly patients. Cocaine (p = 0.015) and cannabinoids (p = 0.002) were significantly associated with intentionality.

CONCLUSIONS

Our results indicate that FN for potentially important psychoactive and nonpsychoactive substances are common when toxicologic testing is limited to routine UDS in trauma patients. We recommend expanding SBIRT in this patient population to include misuse of tobacco products, prescription analgesics, and over-the-counter antihistamines.