Background: Self-reported pain scores are used widely in clinical and research settings, yet little is known about their interpretability in children. In this prospective, observational study we evaluated the relationship between 0 to 10 numerical rating scale (NRS) pain scores and other self-reported, clinically meaningful outcomes, including perceived need for medicine (PNM), pain relief (PR), and perceived satisfaction (PS) with treatment in children postoperatively.
Methods: This study included children ages 7 to 16 years undergoing surgery associated with postoperative pain. One to 4 observations were recorded in each child within the first 24 hours postoperatively. At each assessment, children rated their pain with the NRS, stated their PNM, and rated their satisfaction with pain management. Assessments were repeated within 1 to 2 hours, and children additionally rated their PR as the same, better, or worse in comparison with the earlier assessment. Receiver operator characteristic curves were developed to examine potential NRS cut-points for PNM and PS, and the minimum clinically significant difference (MCSD) in pain score associated with PR was calculated.
Results: Three hundred ninety-seven observations (including 189 pairs) were recorded in 113 children. NRS scores associated with PNM were significantly higher than "no need" (median 6 vs. 3; P < 0.001). NRS scores >4 had good sensitivity (0.81) and specificity (0.70) to discriminate PNM, but with a large number of false positives and negatives (e.g., 42% of children with scores >4 did not need analgesia). The MCSD in NRS scores was -1 (95% confidence interval [CI] -0.5 to 1) or +1 (CI 0.5 to 2.7) in relation to feel "a little better" or "worse," respectively (P < 0.001 vs. the same). NRS scores >6 had a sensitivity of 0.82 and specificity of 0.76 in discriminating dissatisfaction with treatment, yet 46% and 24% of children with scores >6, respectively, were somewhat to very satisfied with their analgesia.
Conclusions: This study provides important information regarding the clinical interpretation of NRS pain scores in children. Data further support the NRS as a valid measure of pain intensity in relation to the child's PNM, PR, and PS in the acute postoperative setting. However, the variability in scores in relation to other clinically meaningful outcomes suggests that application of cut-points for individual treatment decisions is inappropriate.