Responsiveness of the PROMIS physical function measure in orthopaedic trauma patients

Abstract

Objectives: To compare the responsiveness of the Patient-Reported Outcomes Measurement Information System Physical Function (PROMIS-PF) to the 36-Item Short Form Survey Physical Component Score (SF36-PCS) in orthopaedic trauma patients from pre-injury to one year recovery.

Design and setting: Prospective cohort study at a Level 1 trauma centre.

Participants: Patients over the age of 18 with orthopaedic trauma injuries to the pelvis, lower extremity or upper extremity between 2017 and 2018.

Main outcomes measurements: The PROMIS-PF and SF36-PCS assessments were conducted at baseline, 3 months, 6 months and 12 months. Responsiveness of each measure was assessed between time points by calculating the standardized response mean (SRM), the proportions of patients exceeding minimal clinically important difference (MCID), and the floor and ceiling effects.

Results: Sixty-eight patients with completed assessments at every timepoint were included: mean age 44.7 years, 39 were male and mean Injury Severity Score (ISS) was 7.4 (range: 4-16). Mean time of completion for the SF-36 at all the time points was 5.6 min vs 1.7 min for the PROMIS-PF (p<0.01). The SRM was comparable between measures at all the time points. Although a greater proportion of patients achieved MCID for SF36-PCS between all the time points, this only approached statistical significance between the 6- and 12-month assessments (47.1% vs 33.8%; p = 0.15). There was a significant ceiling effect demonstrated with the PROMIS-PF at baseline and 12-month assessments, with 34 (50.0%) patients and 7 (10.3%) patients achieving the maximum scores at each time point, respectively.

Discussion and conclusions: PROMIS-PF has a more favourable responder burden based on lower time to completion and comparable responsiveness to the SF-36 PCS. However, there are limitations in responsiveness with the PROMIS-PF in patients who are higher functioning as demonstrated by the ceiling effects in patients at baseline pre-injury and at 12 months post-injury timepoints.

Keywords: Ceiling Effects; Fracture; Injury; Minimal Clinically Important Difference; Orthopaedic; PROMIS; Patient Reported Outcome Measure; Physical Component Score; Physical Function; Responder Burden; Responsiveness; SF-36; Standardized Response Mean; Trauma.