Reliability, Validity and Efficiency of an Item Response Theory-based Balance Confidence Patient Reported Outcome Measure

Abstract

<h3>Research Objectives</h3> To calibrate items from the original 16-item Activities-Specific Balance Confidence (ABC) Scale to create an item response theory (IRT)-based item bank and scoring metric and assess its reliability, validity, and efficiency using computerized adaptive test (CAT) and short form (SF) administration modes. <h3>Design</h3> Retrospective cohort; data from patients assessed for balance impairments. <h3>Setting</h3> Outpatient rehabilitation. <h3>Participants</h3> Cohort of 20,354 adult patients [60% female; mean age (SD) 66 (16)] treated for orthopedic and neurological conditions in 822 clinics (44 states; USA) who completed the ABC Scale at intake. <h3>Interventions</h3> Not applicable. <h3>Main Outcome Measures</h3> An IRT-based patient-reported outcome measure (PROM) was developed based on the original 16 ABC items, examining assumptions of local independence, unidimensionality, item fit, and absence of differential item functioning (DIF). CAT and SF T-scores (mean=50, SD=10) were assessed for reliability, validity, and administrative efficiency. Known-groups validity, floor and ceiling effects and responsiveness were evaluated. <h3>Results</h3> All original 16 items were retained in the model, supporting the ABC Scale's unidimensionality and item-fit. A short form was created selecting 6 items that provided the most statistical information at targeted score levels. Reliability estimates were >= 0.95 for both administration modes. Scores discriminated between patient groups in clinically expected ways. Negligible floor and ceiling effects and no item DIF were observed. Change score effect size was medium (0.58). Simulated CAT scores were generated using an average of 4.7 items (median=4, range 4 to 10) and correlated highly with scores derived from the full bank (r=0.99). <h3>Conclusions</h3> The IRT-based balance confidence PROM was reliable, valid, moderately responsive to change, and efficient, with excellent score coverage. The presentation will address optimizing the use of the CAT, SF, and the full ABC Scale, depending on clinical and research purposes. <h3>Author(s) Disclosures</h3> Dr. Kallen consults for Net Health Systems, Inc., the company that owns the Focus on Therapeutic Outcomes (FOTO) Patient Outcomes system that gathers and manages the data analyzed in this manuscript. Dr. Myers is the developer of the ABC Scale. The authors have no other conflicts to declare.