Decision rule and utility of routine urine toxicology screening of trauma patients.

The objective of this study was to determine the impact of urine drug screening of major trauma victims on patient care and derive a decision rule for selective screening. Retrospective chart review of 170 trauma patients at a Level I Trauma Center, certified by the American College of Surgeons, was undertaken. The decision rule was developed by Classification and Regression Tree (CART) analysis to maximize sensitivity, with secondary attention to specificity. Eighty-nine percent of trauma patients were screened, while 26.0% had positive tests for illicit drugs. Serum ethanol was positive in 31.2%, over the legal limit of 0.08 g/dl. Both a legally intoxicated ethanol level and positive illicit drug screen were found in 11.0%. Additionally, 42.5% of patients with a positive illicit drug screen were also intoxicated (blood alcohol level above legal limit). Conversely, 35.4% of legally intoxicated patients also had positive illicit screens. Drug treatment referral occurred in 17.5% of positive drug screens. For urgent surgery, median time to drug screen result was 117 min, while median time to operation was 110 min. Of operative patients, 57% had the drug screen result recorded on the chart at any time, but only 14.3% of illicit screens were noted in the anaesthesia record. For all patients with and without operations, 71.1% had the result noted on the chart. We derived a 'low risk rule' to identify most patients with positive illicit drug screens (95% sensitivity, 55% specificity, 66% positive and 93% negative predictive values; accuracy 74%), while limiting the number of unnecessary tests. The rule avoids screening 48% of patients, missing only 5% of true positives. It is concluded that urine screening for illicit drugs in trauma patients can be performed selectively according to a decision rule based on demographics, mechanism of injury and time of presentation. This rule, which captures most positive screens while eliminating screening in low risk patients, could result in significant cost savings. Only prospective validation of these rules in patient populations of other trauma centres will offer confidence that the decision points are valid. Urine drug screening infrequently affected patient management or resulted in drug treatment referral in our sample. We call for increased vigilance in recording results and referring patients for treatment.