Abstract
OBJECTIVE: To determine the best regional pain score cutoff value that corresponds with patient-reported improvement in lumbosacral radiculopathy (LSR). DESIGN: Retrospective pooled data analysis from 3 randomized, controlled, multicenter trials using similar outcome assessments. All participants were exposed to interventions (epidural injections). SETTING: Military medical centers (6 U.S.A., 1 Germany) and large tertiary care hospitals (4 urban, 1 Veterans Affairs) between 2008 and 2014. SUBJECTS: A total of 352 active duty military personnel and civilians ≥ 18 years of age with LSR. METHODS: Receiver operating characteristics (ROC) with area under the curve (AUC) were calculated for 1-month outcomes for pain (numeric rating scale) using absolute and relative change in regional pain scores (back, leg) to predict clinical improvement (global perceived effect). RESULTS: Leg pain demonstrated greater predictive ability to identify clinical improvement compared to back pain for both absolute (ROC AUC [95% confidence interval (CI)] 0.855 [0.813, 0.896] vs. 0.753 [0.702, 0.805]; P < 0.001) and relative (AUC [95% CI]; 0.867 [0.826, 0.909] vs. 0.780 [0.729, 0.831]; P = 0.002) reduction in reported pain. Clinical improvement was best identified using a leg pain reduction threshold of ≥ 1.75 points (absolute) and ≥ 23.5% (relative). CONCLUSIONS: Region-specific pain cutoff ratings predicted clinical improvement for patients with LSR. Cutoff points using newly identified, smaller reductions of 1.75 points and 23.5% more accurately predicted clinical improvement for LSR than conventionally used cutoffs (2 points and 30%). LSR patients report meaningful clinical improvement with smaller reductions in pain compared to other chronic pain diagnoses, suggesting LSR patients may have different expectations.