Assessing Correlations Between Radiographic Classification and Patient-Reported Outcome Measures for Symptomatic...

Clinicians routinely use the Eaton-Glickel radiographic classification to stage severity of thumb carpometacarpal osteoarthritis (CMC OA). Our purpose was to evaluate correlations between baseline radiographic stages and prospectively collected patient-reported outcome measures (PROMs) for CMC OA.

This is a retrospective review of prospectively collected data from a larger observational study of patients aged 35-85 years with newly symptomatic thumb CMC OA. Patients with bilateral presentation complete PROMs for each side. We identified patients with adequate radiographs available for review and who had completed baseline PROMs. PROMs were collected electronically and included the Brief Michigan Hand Questionnaire and visual analog/numerical rating scales regarding pain as well as Patient-Reported Outcomes Measurement Information System Pain Interference v1.1 and Global Health modules. We recorded participants' demographics, medical and surgical history, prior treatments, and PROM scores at initial presentation. A medical student was first taught by a fellowship-trained hand surgeon on Eaton-Glickel radiographic staging. They each then independently staged the first 40 series of radiographs. After establishing adequate reliability, the student staged the remaining available radiographic series. We used Cohen's Kappa analyses to determine intrarater reliability and interrater reliability. We used Spearman's rho to assess correlations between PROM scores and radiographic stages.

Adequate radiographs and baseline PROMs were retrievable for 85 patients (64 unilateral, 21 bilateral) for a total of 106 series of radiographs. The student's intrarater reliability was moderate, whereas the surgeon's was substantial. Their interrater reliabilities were moderate (all P < .05). We found no significant correlations between any PROMs and radiographic staging.

Our study adds to growing evidence that radiographic severity for thumb CMC OA does not correlate with validated PROMs. It is becoming clearer that objective measures of CMC OA severity do not fully capture the way patients experience this condition.

Therapeutic IV.