Objective: To evaluate potential alternatives to the numeric rating scale (NRS) for routine pain screening. Design: Cross-sectional. Setting: Nineteen Veterans Affairs outpatient clinics in Southern California at two hospitals and six community sites. Patients: Five hundred twenty-eight veterans from primary care, cardiology, and oncology clinics sampled in proportion to the total number of visits made to each clinic during the previous year. Methods: Veterans were approached following clinic visits to complete researcher-administered surveys about their clinic experience. Using the Brief Pain Inventory (BPI) interference scale of ≥5 as a reference standard for important unrelieved pain, we evaluated potential alternative pain screening items and item combinations by analyzing sensitivity and specificity, area under the receiver operating curve (AUC), and likelihood ratios. Results: Of the veterans, 43.6% had unrelieved pain as measured by the reference standard. Approximately half had painful musculoskeletal diagnoses and one-third had comorbid mental health or substance use disorders. The fifth vital sign detected pain less accurately than did an NRS with a 1-week timeframe and an item assessing pain-related bother over the past week. AUCs were 0.79, 0.86, and 0.86, respectively. A sequential approach combining the pain-related bother and NRS with a 1-week timeframe items had good discriminatory ability. Conclusions: Alternative single or combined pain screening strategies assessing pain-related bother may improve routine pain detection.
- Ambulatory Care