A validation defense of the 10-item PROMIS Global Health scale in lumbar decompression surgery.

Abstract

To date, the 10-item Patient-Reported Outcome Measurement Information System Global Health (PROMIS GH) scale has not been validated for use in lumbar decompression surgery. Therefore, this study was performed to validate the scale and identify minimum clinically important differences (MCIDs) in PROMIS GH global physical health (GPH) and global mental health (GMH) domain scores in this patient population. A prospectively collected quality improvement registry was retrospectively reviewed. PROMIS GH scores were obtained at baseline and 12 months postoperatively from patients undergoing lumbar decompression surgery between 2016 and 2023. Other validated, spine-specific patient-reported outcome measures (PROMs) assessing quality of life were also collected, including the Oswestry Disability Index (ODI), EuroQol visual analog scale (EQ-VAS), EQ-5D-3L (3-level version of the 5-dimension EuroQol health survey), and numeric rating scale for back pain (NRS-BP) and leg pain (NRS-LP). Pearson correlation coefficients assessed the relationship between PROMIS GH and other PROMs at baseline (r0) and 12 months (r12), as well as changes from baseline to 12 months (rΔ12). Cronbach's alpha was used to evaluate the internal consistency of GPH and GMH at each time point (α0, α12, and αΔ12). MCIDs were calculated for GPH and GMH using four established anchor-based methods, with North American Spine Society patient satisfaction index scores as the anchor. A total of 1134 patients completed baseline and 12-month PROMIS GH questionnaires. GPH scores demonstrated strong correlations with ODI (r0 = -0.72, r12 = -0.78, rΔ12 = -0.68), EQ-5D-3L (r0 = 0.67, r12 = 0.74, rΔ12 = 0.60), and EQ-VAS (r0 = 0.55, r12 = 0.64, rΔ12 = 0.53) and moderate correlations with NRS-BP (r0 = -0.45, r12 = -0.67, rΔ12 = -0.47) and NRS-LP (r0 = -0.47, r12 = -0.56, rΔ12 = -0.47) scores. GMH scores had moderate to weak correlations with ODI (r0 = -0.53, r12 = -0.62, rΔ12 = -0.51), EQ-5D-3L (r0 = 0.55, r12 = 0.67, rΔ12 = 0.52), EQ-VAS (r0 = 0.50, r12 = 0.58, rΔ12 = 0.46), NRS-BP (r0 = -0.26, r12 = -0.49, rΔ12 = -0.30), and NRS-LP (r0 = -0.29, r12 = -0.43, rΔ12 = -0.33) scores. Good to questionable internal consistency was observed in GPH (α0 = 0.66, α12 = 0.78, αΔ12 = 0.63) and GMH (α0 = 0.75, α12 = 0.85, αΔ12 = 0.74). Based on 12-month score changes, MCID thresholds ranged from 5.04 to 12.77 for GPH and from 4.49 to 9.79 for GMH. The receiver operating characteristic (ROC) curve analysis was deemed most appropriate for calculating MCIDs. The GPH and GMH components of PROMIS GH have strong validity and reliability, with moderate to strong correlations to established PROMs and acceptable internal consistency. Based on the ROC approach, MCID thresholds were 11.05 for GPH and 4.90 for GMH. The current findings define and defend clinically significant thresholds for score improvement in PROMIS GH and thereby bolster its clinical utility for patients undergoing lumbar decompression surgery.