Objectives: Risk assessment and stratification has become an important aspect of the prescribing of opioids to patients with chronic pain. There is little empirical data available on the sensitivity and specificity of commonly used risk assessment tools. This paper describes 2 studies that compare the prediction capabilities of various risk assessment tools.
Methods: The first study presents data on patients at a pain practice whose treatment with opioids was stopped due to their engaging in aberrant drug-related behavior. Patients were assessed with the Screener and Opioid Assessment for Patients with Pain-Revised (SOAPP-R), the Pain Medication Questionnaire, the Opioid Risk Tool, and a clinical interview. A second study compared the risk assessment measures, SOAPP-R, Pain Medication Questionnaire, Opioid Risk Tool, and a clinical interview. Data were gathered on whether patients had engaged in aberrant drug-related behavior at 6-month follow-up.
Results: Significant differences in the measures were found. Accuracy did not appear to be a function of the type of aberrant drug-related behavior that the patient engaged in for any of the measures. The clinical interview showed the best sensitivity of the 4 risk measures in predicting risk. The SOAPP-R showed the best sensitivity of the self-report measures. However, the SOAPP-R appears to overrate risk.
Discussion: Overall, these studies indicate that not all risk assessment tools are equal in their ability to accurately predict future aberrant drug-related behavior. It may be that written risk assessment tools that use more subtle items are better suited to certain patient populations.