Identification of Radiographic Foot Osteoarthritis: Sensitivity of Views and Features Using The La Trobe Radiographic Atlas.

OBJECTIVE To compare the sensitivity of alternative case-finding approaches for the identification of foot osteoarthritis (OA) based on the La Trobe radiographic atlas. METHODS Cross-sectional study of 533 adults aged ≥50 years with foot pain in the past year. Weightbearing dorso-plantar (DP) and lateral x-rays were taken of both feet. The La Trobe radiographic atlas was used to document the presence of osteophytes (OP) and joint space narrowing (JSN). Prevalence of OA in each joint was documented using both views and features in combination (as recommended in the original atlas), and by using a single view (DP or lateral only) and a single feature (OP or JSN only). RESULTS Compared to the recommended case definition based on OP and JSN using both views, a DP only view identified between 15 and 77% of OA cases, while a lateral only view identified between 28 and 97% of OA cases. Compared to the recommended case definition of using both features, using only OP identified between 46 and 94% of OA cases, while using only JSN identified between 19 and 76% of OA cases. CONCLUSION Applying the La Trobe radiographic atlas but using only one x-ray view (DP or lateral) or one feature (OP or JSN) in isolation misses a substantial number of OA cases, and the sensitivity of these approaches varies considerably between different foot joints. These findings indicate that, where possible, the atlas should be administered according to the original description to avoid under-ascertainment of radiographic foot OA.