Values derived from patient reported outcomes in spine surgery: a systematic review of the minimal clinically important difference, substantial clinical benefit, and patient acceptable symptom state.

Abstract

While patient reported outcome measures (PROMs) define value in spine surgery, several values such as minimal clinically important difference (MCID), substantial clinical benefit (SCB), and patient acceptable symptom state (PASS) help guide the interpretation of PROMs and identify thresholds of clinical significance. Significant variation exists in reported values and their calculation, so the primary objective of this study was to systematically review the spine surgery literature for metrics of clinical significance derived from PROMs. We conducted a query of PubMed/MEDLINE and Scopus databases from inception to January 1, 2023, for studies that derived quantitative metrics (e.g., SCB, MCID, PASS) from PROMs in the setting of spine surgery with minimum 1-year follow-up. Details regarding the specific PROMs were collected including which PROM was measured, whether anchor- or distribution-based methods were utilized, the specific calculations, and the recommended value for a given PROM based on all evaluated calculations. Thirty-seven studies of 21,780 patients were included. The most commonly evaluated PROM-derived value was the MCID (n = 28), followed by PASS (n = 6) and SCB (n = 4). Twenty-one studies only utilized anchor-based calculations, 15 utilized both anchor-based and distribution-based methods, and one only utilized distribution-based calculations. The most commonly evaluated legacy PROMs were the Oswestry Disability Index (ODI) (N = 11, MCID range 4-20) and visual analog scale back pain (N = 5, MCID range 0.5-4.6). All 10 studies that derived SCB or PASS utilized the receiver operating characteristic methods. Among the six studies deriving a PASS value, four only evaluated ODI, identifying PASS ranging from 5 to 22. While calculated measures of clinical significance such as MCID, PASS, and SCB exist, significant heterogeneity exists in the current literature. Current shortcomings include a wide variability of reported value thresholds across the literature, and limited applicability to more heterogenous patient populations than the targeted cohorts included in published investigations. Continued investigations that apply these methods to heterogenous, large-scale populations can help increase generalizability and validity of these measures. III.