Validation and Generalizability of Preoperative PROMIS Scores to Predict Postoperative Success in Foot and Ankle Patients

Abstract

Background: A recent publication reported preoperative Patient-Reported Outcomes Measurement Instrumentation System (PROMIS) scores to be highly predictive in identifying patients who would and would not benefit from foot and ankle surgery. Their applicability to other patient populations is unknown. The aim of this study was to assess the validation and generalizability of previously published preoperative PROMIS physical function (PF) and pain interference (PI) threshold t scores as predictors of postoperative clinically meaningful improvement in foot and ankle patients from a geographically unique patient population. Methods: Prospective PROMIS PF and PI scores of consecutive patient visits to a tertiary foot and ankle clinic were obtained between January 2014 and November 2016. Patients undergoing elective foot and ankle surgery were identified and PROMIS values obtained at initial and follow-up visits (average, 7.9 months). Analysis of variance was used to assess differences in PROMIS scores before and after surgery. The distributive method was used to estimate a minimal clinically important difference (MCID). Receiver operating characteristic curve analysis was used to determine thresholds for achieving and failing to achieve MCID. To assess the validity and generalizability of these threshold values, they were compared with previously published threshold values for accuracy using likelihood ratios and pre- and posttest probabilities, and the percentages of patients identified as achieving and failing to achieve MCID were evaluated using χ2 analysis. Results: There were significant improvements in PF ( P < .001) and PI ( P < .001) after surgery. The area under the curve for PF (0.77) was significant ( P < .01), and the thresholds for achieving MCID and not achieving MCID were similar to those in the prior study. A significant proportion of patients (88.9%) identified as not likely to achieve MCID failed to achieve MCID ( P = .03). A significant proportion of patients (84.2%) identified as likely to achieve MCID did achieve MCID ( P < .01). The area under the curve for PROMIS PI was not significant. Conclusions: PROMIS PF threshold scores from published data were successful in classifying patients from a different patient and geographic population who would improve with surgery. If functional improvement is the goal, these thresholds could be used to help identify patients who will benefit from surgery and, most important, those who will not, adding value to foot and ankle health care. Level of evidence: Level II, Prospective Comparative Study