OBJECTIVES:
To compare measurement properties of Patient-Reported Outcomes Measurement Information System (PROMIS) (Physical Function (PF) and Pain Interference (PI)) computerized adaptive testing (CAT) to traditional Short Musculoskeletal Function Assessment (SMFA) (dysfunction index (DI) and bother index (BI)). To explore factors associated with PROMIS scores.
METHODS:
Design: Cross-sectional study.
Setting:
Level-I Trauma Center.
Patient Selection Criteria:
isolated upper/lower extremity fracture patients were recruited from the orthopaedic trauma outpatient clinic (October 1, 2021 to January 1, 2023).
Outcome measures:
Correlations (Pearson), reliability (Standard error (SE)(T score), efficiency (amount of information per item(1 – SE2/ Nitems)) and floor/ceiling effects were assessed. An r>0.7 represented high correlation, SE≤2.2 represented sufficient reliability. Factors associated with worse PROMIS scores were also identified.
RESULTS:
In total, 202 patients completed PROMs at median 98 days follow-up. Correlations between PROMIS-PF and SMFA-DI, and PROMIS-PI and SMFA-BI were -0.84 and 0.65. Reliability was very high for both instruments (mean SE 2.0(PROMIS-PF), SE 2.1(PROMIS-PI) and SE 1.2(SMFA-DI), SE 1.8(SMFA-BI)). Relative efficiency for PROMIS-PF versus SMFA-DI, and PROMIS-PI versus SMFA-BI was 7.8(SD2.5), and 4.1(SD1.7), respectively. Neither PROMIS nor SMFA exhibited floor/ceiling effects. In the multivariable regression analyses, elevated levels of depression, among other factors, showed an (independent) association with worse PROMIS-PF and PROMIS-PI scores.
CONCLUSIONS:
PROMIS-PF and PROMIS-PI CATs showed a (high and moderate) correlation with SMFA and hence measure a comparable construct of physical function and discomfort. As CATs are much more efficient to administer, they present a compelling alternative to SMFA for evaluating impact of fracture treatment. The relation between symptoms of depression and PROMIS scores emphasizes the importance of psychosocial aspects of health in orthopaedic trauma patients.
LEVEL OF EVIDENCE:
Diagnostic Level II. See Instructions for Authors for a complete description of levels of evidence.