Defining clinically meaningful pain reduction after spinal injections using an 11-point numerical rating scale and patient-reported relative reduction of pain

Background We aimed to establish thresholds for assessing clinically meaningful pain improvement after spinal injections using numerical rating scale (NRS) and patient-reported relative reduction (PRR) of pain. Methods This single-center retrospective study included 1,828 patients who received spinal injections at our spinal intervention center and completed a 2-week follow-up telephone interview in 2017. NRS absolute and relative reductions were calculated using the NRS scores before injection and at follow-up. The PRR and 5-point global perceived effect (GPE; 1, completely recovered; 2, much improved; 3, slightly improved; 4, no change; and 5, worsened) were recorded. The mean ± standard deviation (SD) of NRS absolute reduction, NRS relative reduction, and PRR in the GPE 2 and 3 subgroups were calculated. Receiver-operating-characteristic curve analysis was used to determine the cutoffs for NRS absolute and relative reduction and PRR in predicting the substantial clinically important difference (SCID, GPE 1–2) and minimal clinically important difference (MCID, GPE 1–3). Results The mean ± SD of NRS absolute and relative reduction, and PRR in the GPE 2 and GPE 3 were 4.5 ± 1.8, 65.4% ± 21.1%, and 67.7% ± 14.8% and 2.4 ± 1.4, 36.8% ± 22.5%, and 35.7% ± 16.1%, respectively. The cutoffs for predicting SCID and MCID were NRS absolute reduction of ≥3 and ≥1, NRS relative reduction of ≥50% and ≥10%, and PRR of ≥50% and ≥20%, respectively. Conclusions The NRS absolute and relative reduction and PRR thresholds were established to differentiate SCID and MCID in terms of pain improvement after spinal injections.