Initial estimates of the minimal clinically important difference for the Neuropathic Pain Symptom Inventory: a...

The Neuropathic Pain Symptom Inventory (NPSI) is a commonly used assessment in neuropathic pain (NP) trials, yet a Minimal Clinically Important Difference (MCID) has not been established. An MCID would enhance the interpretability of NPSI scores, guiding clinicians and researchers in assessing clinically important improvements in NP symptoms. The aim of this study was to calculate an MCID from the available scientific research that used the NPSI.

We conducted a systematic review and meta-analysis of NP trials reporting the NPSI. Four distributional approaches were applied to estimate the MCID: 1) meta-regression on the set of standard deviation (SD) of change scores, 2) meta-regression on the set of baseline SD scores, 3) simple aggregation on the set of SD of change scores, and 4) simple aggregation on the set of baseline SD scores. Only treatment arms within Randomized Controlled Trials (RCTs) were examined for MCID estimation. Control arms were examined separately in a sensitivity analysis using the simple aggregation method for both SD of change and baseline SD sets. Bias for each included study was assessed using the Cochrane tool for quality assessment of randomized controlled trials.

323 trials were examined, 12 were selected for inclusion with a total of 17 treatment arms. The calculated MCID estimates for the NPSI total score (range 1-100) were 6.21 for the SD of change meta-regression and 7.1 for baseline SD meta regression. The MCID values aggregated using simple aggregation methods were 7.95 using pooled SD of change scores, 7.8 using pooled baseline SD. Control arms had a MCID of 8.04 for SD of Change and 8.71 for Baseline SD.

This study provides preliminary MCID estimates for the NPSI. Limitations include limited data for NP subtypes, highlighting the need for additional anchor-based and etiology-specific MCID research to refine these estimates. These findings can aid in future NP trial design and the interpretation of results.

https://www.crd.york.ac.uk/PROSPERO/view/CRD42025649343, PROSEPRO CRD42025649343.