AAHS Main Site  | Past & Future Meetings  
American Association for Hand Surgery
Meeting Home

Back to 2017 Annual Meeting Program

Efficacy of PROMIS Pain Scores and Likert Pain Scores to Assess Function
Matthew J. St.John, MS, BA; Warren C. Hammert, MD; DDS, BS,
Department of Orthopaedics, University of Rochester, Rochester, NY

Introduction: Patient Reported Outcome Measurement Information System (PROMIS), developed by the NIH, utilizes a health domain related to pain. We evaluated this domain and its association with self-reported function (as determined by PROMIS physical function CAT), and secondarily its association to a Likert 0-10 pain score.

Methods: Adult patients presenting to upper extremity clinic from February to December 2015 completed PROMIS physical function, pain interference, and Likert 0-10 pain scale questionnaires. PROMIS models were completed electronically in their computer adaptive form. Mean population scoring on each module is defined at 50 (SD 10, range 0-100) with larger scores indicating greater amounts of each health element (100=maximal function and most pain). Patients were also asked to rate their pain on a 0-10 scale with 0 being no pain and 10 being worst pain possible. They repeated this same procedure at each subsequent appointment. These data were collected prospectively as routine clinical care and were extracted from the electronic health record for cross-sectional evaluation. Univariate descriptive analyses explored each module's scores in the cohort. Bivariate Pearson correlation analysis defined the directional relationship between modules and Likert pain scores. Fisher transformation was used to correlate significance between correlations.

Results: 10,174 first visit, 4,847 second visit, 2,445 third visit, 1,289 fourth visit, and 684 fifth visit patient's data were recorded. Likert pain scores were slightly-moderately negatively correlated to PROMIS function at each visit (rp1=-0.32; rp2=-0.38; rp3=-0.40; rp4=-0.41; rp5= -0.42, p<0.01). PROMIS pain scores were moderately-highly negatively correlated to PROMIS function at each visit (rp1=-0.64; rp2=-0.67; rp3=-0.70; rp4=-0.73; rp5= -0.69, p<0.01). Likert pain scores were also moderately correlated to PROMIS pain scores (rp1=0.52; rp2=0.57; rp3=0.59; rp4=0.60; rp5= 0.59, p<0.01). Likert pain scores were significantly less correlated then PROMIS pain scores through time (p<0.01) relative to self-reported function (Figure 1).

Conclusions: Both PROMIS pain and Likert pain scores had statistically significant correlations to self-reported physical function for each office visit. PROMIS pain had a significantly stronger correlation to physical function than Likert pain scores. Likert pain scores consistently had only a slight-moderate correlation, while PROMIS pain consistently had moderate-high correlation to self-reported function. PROMIS pain and Likert pain scale scores are only moderately correlated, and are less correlated than PROMIS pain and self-reported function.

Back to 2017 Annual Meeting Program