ISSLS prize in clinical science 2020: the reliability and interpretability of score change in lumbar spine research.

Abstract

A statistically significant score change of a PROM (Patient-Reported Outcome Measure) can be questioned if it does not exceed the clinically Minimal Important Change (MIC) or the SDC (Smallest Detectable Change) of the particular measure. The aim of the study was to define the SDC of three common PROMs in degenerative lumbar spine surgery: Numeric Rating Scale (NRS_BACK/LEG), Oswestry Disability Index (ODI) and Euroqol-5-Dimensions (EQ-5D_INDEX) and to compare them to their MICs. The transition questions Global Assessment (GA_BACK/LEG) were also explored. Reliability analyses were performed on a test-retest population of 182 symptomatically stable patients, with similar characteristics as the Swespine registry population, who underwent surgery for degenerative lumbar spine conditions 2017-2018. The MIC values were based on the entire registry (n = 98,732) using the ROC curve method. The ICC for absolute agreement was calculated in a two-way random-effects single measures model. For categorical variables, weighted kappa and exact agreement were computed. For the NRS, the SDC exceeded the MIC (NRS_BACK:3.6 and 2.7; NRS_LEG: 3.7 and 3.2, respectively), while they were of an equal size of 18 for the ODI. The gap between the two estimates was remarkable in the EQ-5D_INDEX, where SDC was 0.49 and MIC was 0.10. The GA_BACK/LEG showed an excellent agreement between the test and the retest occasion. For the tested PROM scores, the changes must be considerable in order to distinguish a true change from random error in degenerative lumbar spine surgery research. These slides can be retrieved under Electronic Supplementary Material.