Development and Assessment of the Curtis Thumb Carpometacarpal Osteoarthritis Radiographic Classification

Given the availability of modern digital imaging and measurement technologies, we hypothesized we could improve precision and reliability for radiographic classification of thumb carpometacarpal osteoarthritis (CMC OA).

We developed our Curtis CMC classification with literature reviews and input from several hand surgeons at our center. We designed the scheme to allow embedding of the Eaton-Glickel system to facilitate cross-use and comparisons. We applied specific digital imaging measurements to radiographic criteria as well as discrete cutoffs between stages. We conducted a retrospective review of consecutive patients aged 35-85 in our CMC registry. Eligible patients presented with symptomatic thumb CMC OA with no prior nonsurgical care within the past year and no history of CMC surgery. We identified those with adequate digital radiographs within 1 year of baseline presentation that were available for review. A board-certified orthopedic hand surgeon and an instructed medical student independently classified 40 series of radiographs using the Curtis and Eaton-Glickel classifications. One week later, they independently repeated the classifications to determine intra-/interrater reliabilities.

For the Eaton-Glickel classification intrarater reliability assessments, there was moderate agreement for the medical student and substantial agreement for the hand surgeon. For the Curtis classification intrarater assessments there was substantial agreement for student and surgeon. Regarding interrater reliability between the student and surgeon, the Eaton-Glickel classification had moderate agreement for the initial and 1-week evaluations. The Curtis classification had almost perfect interrater agreement for the initial determinations and substantial agreement for the 1-week measures.

The Curtis CMC classification has potential to improve rater reliability for grading OA radiographic severity. Its use of precise digital imaging technology measurements could aid research and care paradigms. Future studies are needed for independent verification of our findings and to evaluate these measures in relation to clinical outcomes.

Diagnostic II.