DOI: 10.1177/24730114261454983 ISSN: 2473-0114

PROMIS Computerized Adaptive Testing Demonstrates Strong Convergent Validity and Lower Patient Burden Compared With Legacy Instruments in Foot and Ankle Fracture Care

Michiel A. J. Luijten, Lotte Haverman, Martijn Poeze, Diederik O. Verbeek
Show PDF Cite

Background:

Despite its advantages, Patient-Reported Outcomes Measurement Information System (PROMIS) remains underused in patients recovering from foot and ankle fractures, and it is unclear whether condition- and region-specific PROs, such as the Foot and Ankle Outcome Score (FAOS), offer additional clinical insight over PROMIS in this cohort. Therefore, PROMIS Computerized Adaptive Test (CAT) measurement properties were evaluated and compared with region-specific and lower-extremity instruments. Also, variables associated with poorer outcomes were identified.

Methods:

In this prospective, cross-sectional study, patients with foot or ankle fractures completed questionnaires. PROMIS Mobility (MOB) and Pain Interference (PI) CATs were compared with FAOS, Lower Extremity Functional Scale (LEFS), and Short Musculoskeletal Function Assessment (SMFA). Convergent validity (correlations), reliability (SE, Cronbach α), efficiency (number of items/completion time), and floor/ceiling effects of each instrument were assessed. Multivariable linear regression was used to evaluate factors associated with PROMIS outcomes.

Results:

Seventy-three patients were included (mean age 50 [SD 17] years, 60% female). PROMIS-MOB showed particularly high correlations with domains focused on lower extremity functioning (FAOS-Sport, r  = 0.74; FAOS-ADL, r  = 0.76; LEFS, r  = 0.86; and SMFA-dysfunction index, r  = −0.83). PROMIS-PI showed particularly high correlations with pain-focused domains (FAOS-Pain, r  = −0.78; SMFA-bother index, r  = 0.72). PROMIS CATs and (most) other legacy instruments showed excellent reliability (SE = 2.1, α = 0.81-0.96). PROMIS CATs required fewer items (mean 6 and 4 items, respectively, vs 20-46) and less time to complete (mean 56 and 36 seconds, respectively, vs 118-295 seconds; P  < .001). No floor or ceiling effects were observed for PROMIS CATs; a limited proportion of patients reached score extremes on some FAOS subscales (up to 12%), remaining below the 15% threshold. Greater pain intensity and depressive symptoms were independently associated with worse mobility and greater pain interference.

Conclusion:

Condition- and region-specific instruments may provide limited information beyond that obtained with PROMIS CATs. PROMIS (MOB, PI) CATs captured comparable constructs of functional recovery and pain interference, as demonstrated by high correlations with commonly used instruments, while substantially reducing patient burden. The relation between depressive symptoms and worse outcomes underscores the significance of psychological factors in recovery in patients with foot and ankle fractures.

Levels of Evidence:

Level II, diagnostic study.

More from our Archive