Construct Validity and Item Response Theory Analysis of the PROMIS-29 v2.0 in Recipients of Lumbar Spine Surgery.
STUDY DESIGN: Observational cohort design involving measurement property assessment. OBJECTIVES: The aim of this study was to assess construct validity through hypothesis testing and to examine reliability and discrimination of the PROMIS-29 v2.0 using item response theory (IRT) analyses. SUMMARY OF BACKGROUND DATA: Patient-Reported Outcomes Measurement Information System (PROMIS) 29.0 version 2.0 includes 28 questions for the seven domains of Physical function, Anxiety, Depression, Fatigue, Sleep disturbance, Social role, and Pain interference, and one item related to pain intensity. To date, the tool has not been tested for construct validity for selected concepts, nor has it been evaluated using IRT in a population of spine surgery recipients. METHODS: Using the Quality Outcomes Dataset lumbar registry, we evaluated the construct validity of the PROMIS-29 v2.0 against pain intensity measures for back and leg, the Oswestry Disability Index, the EQ5D 3L-visual analog scale (quality of life) and proxy measures of activities, mobility and self-care. IRT was used to test PROMIS-29 v2.0 assumptions and fit. RESULTS: The sample included 652 surgery recipients (mean age = 60.1, SD = 14.0) who had a high level of baseline disability. Hypothesis testing confirmed direction and magnitude of correlation between the PROMIS and legacy measures in 10 of the 12 hypotheses. IRT identified three misfit items, but otherwise adequate scale reliability and unidimensionality. CONCLUSION: The PROMIS-29 v2.0 measures several different constructs pertinent to a patient's health and recovery during spine surgery. We feel that the PROMIS-29 v2.0 tool is a useful and effective outcome measure for populations who receive spine surgery.Level of Evidence: 4.
Cook, CE; Rethorn, ZD; Chiarotto, A; Garcia, AN; Gottfried, O
Volume / Issue
Start / End Page
Electronic International Standard Serial Number (EISSN)
Digital Object Identifier (DOI)