Discussion
The aim of this study was to assess the use of radiologic imaging in patients with RA compared with subjects without RA. We found that patients with RA, especially women and those with positive RF, undergo more radiologic procedures, than do patients without RA. The rate of radiologic procedures has not changed in recent years. While biennial hand radiographs are obtained more often in patients with RF, we found that clinicians are not consistently following suggestions regarding obtaining baseline and follow-up radiographs.
Conventional radiography has traditionally been used to document characteristic changes of RA including bone erosions and joint space narrowing. Indeed, the latter 2 radiologic findings help quantify joint damage progression in scoring systems such as the Sharp score and its modified versions. This tool is commonly utilized in clinical trials for assessment of efficacy of therapeutic interventions. In addition to hand radiographs, patients with RA are most likely to undergo imaging of the lower limbs as well as spine and pelvis.
Usage of advanced imaging (CT/MRI) of upper and lower extremity is higher in patients with RA than those without RA and has increased in recent decades. The frequency of bone radionucleotide tests was also higher in patients with RA. However, among female patients, mammograms were performed less frequently among patients with RA than among those without RA. Perhaps patients with RA receive more specialist care and less preventive services. This population also had DXA scans more often than non-RA patients, probably because of awareness that their disease and corticosteroid use increase the risk for osteopenia and osteoporosis. Patients with RF had more radiologic studies than did RF individuals. The presence of RF correlates with disease severity and may account for the reason that patients with RF had greater numbers of imaging studies as a reflection of more aggressive disease and need for close monitoring.
Female patients with RA underwent more radiologic procedures. Hand/wrist radiographs were obtained more often in women, which may reflect disease severity and that women have more joint surgeries than do men. As well, mammograms and bone density scans were included in the overall analysis; both are obtained more frequently in women.
The American College of Rheumatology recommends obtaining hand and feet radiographs at the initial diagnosis of RA. They and others have suggested periodic reevaluation thereafter. We found that 4 years after the initial diagnosis only about 35% of patients had regular radiographs at 2-year intervals. In this subgroup of patients, RF positivity was also a predictor of regular hand/wrist x-rays at 2-year interval until 6 years from initial diagnosis. After that time, no significant difference was noted compared with patients who were RF.
Strengths of this study include the virtually complete ascertainment of all radiologic procedures performed on this population over a very long period. Limitations include the fact that indications for radiologic imaging could not be adequately assessed. As well, radiographs done prior to incidence of RA were not captured; however, the 2 cohorts were matched on date of RA incidence so that the follow-up periods were comparable. Although we studied procedures with ionizing radiation, we did not attempt to formally assess overall radiation exposure or disease progression in the hand radiographs for technical reasons. Because not all ultrasound procedures done in the context of joint evaluations were consistently recorded until recently, we do not have full information about ultrasound use.
A judgment about whether routine hand/wrist (as well as ankle/foot) radiographs are needed at regular intervals, such as every 2 years, was not directly assessed in this study and will depend on the information they provide. Although rates of progression were not studied, in the current era where radiographic progression does not occur in a large percentage of patients, it may be that frequent radiographs are unnecessary and may be regarded as "overused." Certainly in this cohort, these radiographs were not routinely obtained at this interval over the generally long disease duration. A better understanding of the profile of the patient at risk who would benefit from more frequent assessment of radiographic progression will help to target those patients for whom these examinations are beneficial. In the future, in many cases, the need for radiographs may be superseded by periodic low-cost ultrasonographic joint evaluation. In the meantime, patients who have stable hand radiographs after 2 years may not need further routine radiographs of the hands.