Ten-Year Heterogeneity of Minimal Important Change and Patient Acceptable Symptom State After Lumbar Fusions

Abstract

<p>STUDY DESIGN: Cohort study. OBJECTIVE: To evaluate heterogeneity (fluctuation) in minimal important change (MIC) and patient-acceptable symptom state (PASS) for patient-reported outcomes (PROMs) through 10 years after lumbar fusion. SUMMARY OF BACKGROUND DATA: PROMs have become key determinants in spine surgery outcomes studies. MIC and PASS were established to aid PROM interpretations. However, their long-term stability has not yet been reported. METHODS: A consecutive series of elective lumbar fusions were followed up using the Oswestry Disability Index (ODI) and Visual Analogue Scale (VAS) for pain. Improvement was rated by a 4-point Likert scale into "improved" or "nonimproved." Satisfaction-to-treatment was rated by the patients' willingness to undergo surgery again. Receiver operating characteristics (ROC) curve analysis estimated MIC (95% confidence interval, CI) as the PROM change that best predicted improvement at distinct time-points. PASS (CI) was estimated as the lowest PROM score at which the patients were still satisfied. Heterogeneity across thresholds was evaluated using the DeLong algorithm. RESULTS: MIC for ODI represented heterogeneity across 10 years, ranging from -21 (-24 to -16) at two years to -8 (-7 to -4) at five years, P<0.001. The areas under the ROC curves (AUCs) (0.79 to 0.85) indicated acceptable to excellent discrimination. Heterogeneity was not significant in the MICs for the pain scores. At one year, MIC for back pain was -24 (-38 to -15), AUC 0.77, and for leg pain, it was -26 (-44 to -8), AUC 0.78. No significant heterogeneity was observed in 10-year PASS scores. At 1-year, PASS for ODI was 22 (15 to 29), AUC 0.85. Similarly, the one-year PASS for back pain was 38 (20 to 56), AUC 0.81, and for leg pain, it was 49 (26 to 72), AUC 0.81. CONCLUSIONS: MIC for ODI fluctuated over 10 years after lumbar fusions. PASS values for all PROMs seemed most stable over time. Caution is needed when generic MIC values are used in long-term studies. LEVEL OF EVIDENCE: Level III.</p>