P62. Responsiveness of outcome measures in nonsurgical patients with lumbar spinal stenosis: a secondary analysis from a randomized controlled trial

Abstract

<h3>BACKGROUND CONTEXT</h3> Lumbar spinal stenosis (LSS) is a prevalent condition associated with limited walking capacity, reduced physical function and increased disability. While many patients with LSS potentially benefit from nonsurgical interventions, it is critical to have sensitive assessment tools to detect changes in these domains over time to guide clinical decisions. However, limited information is available about the responsiveness of performance-based tests and patient-reported outcomes commonly used to assess nonsurgical patients with LSS. <h3>PURPOSE</h3> To assess the responsiveness of the Self-Paced Walking Test (SPWT), Swiss Spinal Stenosis (SSS), and Oswestry Disability Index (ODI) in patients with LSS not surgically treated and derive their Minimal Clinically Important Difference (MCID) values. <h3>STUDY DESIGN/SETTING</h3> Secondary analysis from a randomized clinical trial of nonsurgical interventions for patients with LSS. <h3>PATIENT SAMPLE</h3> A total of 180 patients with complete data for the SPWT, SSS, and ODI at baseline, 2- and 6-month followups were analyzed. <h3>OUTCOME MEASURES</h3> Responsiveness analyzes included the SPWT, SSS, ODI and Patient Global Index of Change (PGIC). <h3>METHODS</h3> The SPWT, SSS, and ODI responsiveness were investigated by distribution- and anchor-based methods at 2- and 6-month followup. The distribution-based responsiveness was assessed using standardized effect sizes, and standardized response means. The anchor-based responsiveness used the PGIC as the external anchor to distinguish responders and non-responders. A 7-point scale was used at 2 and 6 months, ranging from ‘very much worse' to ‘very much improved.' First, Spearman correlation coefficients (Rho) were calculated between the mean change in each outcome measure and the PGIC scores. Next, based on the anchor, we defined the ‘minimal improvement' subgroup, including patients rating their overall status at least as ‘minimal improved,' and the ‘moderate improvement' subgroup, including patients rating theirs overall status at least as ‘much improved.' Then, we calculated Areas Under the Curve and MCIDs for each outcome in both subgroups. <h3>RESULTS</h3> The following values represent indices of responsiveness at 2 and 6 months for each outcome measure: (a) standard effect sizes were 0.48 and 0.50 for SPWT, -0.42 and -0.36 for SSS, -0.29 and -0.25 for ODI; (b) standard response means were 0.44 and 0.37 for SPWT, -0.45 and -0.34 for SSS, -0.33 and -0.27 for ODI; (c) Spearman's correlation coefficients between the PGIC and outcome measures were 0.44 and 0.39 for SPWT, -0.53 and -0.55 for SSS, -0.46 and -0.54 for ODI; (d) in the ‘minimal improvement' subgroup, MCIDs were 376 and 319 meters for SPWT, -5.3 and -5.8 points for SSS, -9.3 and -10.8 points for ODI; (e) AUCs were 0.76 and 0.68 for SPWT, 0.78 and 0.76 for SSS, 0.76 and 0.76 for ODI; (f) in the ‘moderate improvement' subgroup, MCIDs were 344.2 and 538.2 meters for SPWT, -5.5 and -7.5 points for SSS, -9.1 and -13.6 points for ODI; (g) AUCs were 0.71 and 0.74 for SPWT, 0.77 and 0.82 for SSS, 0.73 and 0.81 for ODI. <h3>CONCLUSIONS</h3> The SPWT, SSS, and ODI are similarly responsive outcome measures to assess and monitor changes in walking capacity, physical function, and disability over time in nonsurgical patients with LSS. Based on this finding, clinicians may use the MCIDs to assess patients' clinical progress, treatment effectiveness, and support shared decision-making. Also, researchers may use the MCIDs and effect sizes as references to inform future studies in the field. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.