Can clinicians accurately predict which patients are misusing their medications?

Abstract

Identifying patients on chronic opioid therapy who may be abusing or misusing their medications remains difficult despite the numerous risk assessment tools available. In this study, clinicians were asked to identify patients who they thought were at risk for medication misuse and those that weren't based on risk assessment methods used in their practice. In addition, a third group was identified as a random sample where no risk assessments were performed. Urine drug testing results were then compared with these assessments. For the purposes of this study, urine drug testing results were considered abnormal if illicits were present or if the prescribed medications were absent from the urine sample. A total of 549 samples were received from 62 clinicians in 50 practices. Clinician assessments of patients who were thought to potentially be misusing their medications (N= 173) were correct 72% of the time. In patients they thought were not at risk for misuse of medications (N=204) they were only correct in 40% of patients, with 60% of this group, who were thought to be compliant, having urine drug tests showing illicits, missing prescribed medication or both. The third random group (N=172) showed 61% of samples were normal and 30% abnormal. Overall, the ability of clinicians to correctly categorize patients into the two groups was only slightly better than by chance alone. Accuracy was greater in the group thought to be misusing, but in the group thought to be compliant, clinicians missed 60% of patients who had an abnormal urine drug test. This study reinforces that all patients on chronic opioid therapy need to have urine drug monitoring in order to identify those patients who may be misusing their medications as risk assessment methods utilized in practice settings may miss a significant portion of these patients.