Glucocorticoid use in RA differs greatly, driven by provider preference
Glucocorticoid treatment for rheumatoid arthritis varies greatly among rheumatologists, with provider preference being the greatest predictor of its long-term use, according to data published in Arthritis Care & Research.
“Even though glucocorticoids are recommended for short-term use in rheumatoid arthritis while other medications are being started, many patients remain on glucocorticoids long-term,” Michael D. George, MD, MSCE, of the University of Pennsylvania, told Healio Rheumatology. “Using glucocorticoids for a long time may lead to health risks such as infections, bone loss, and other side effects, especially if used at higher doses. For some patients, continued glucocorticoid use may be needed to help control their disease, but for others there may be safer options.”
“In this study, we wanted to understand how much glucocorticoid use is driven not by how sick someone is, but rather by the rheumatologist they happen to see,” he added. “In other words, we were looking to see whether some rheumatologists are much more likely to prescribe glucocorticoids than others. This is important because when there are big differences between what different providers do, this suggests possible quality of care issues and an opportunity to better standardized and improve clinical practice.”
To analyze variations in glucocorticoid use in RA among rheumatologists, as well as the predictors of its long-term use among patients, George and colleagues conducted a retrospective cohort study of Medicare claims data from 2006-2015. Using this data, the researchers created two cohorts. The first cohort included 1,272,644 annual observations of 385,597 patients with RA who were new or prevalent DMARD users, and was used to calculate provider preference for glucocorticoids among 6,875 rheumatologists. Preference was defined based on frequency of use compared with other providers.
The second cohort featured 155,539 patients with RA who had been on stable DMARD therapy for at least 9 months. Here, the researchers examined whether provider preference could predict treatment with glucocorticoid doses of at least 5 mg per day, 6 to 9 months after initiation.
According to the researchers, provider preference was highly variable, with rheumatologists at the lowest end prescribing glucocorticoids 33% less often than expected, and the upper quartiles using glucocorticoids 31% more often than expected. In addition, provider preference was strongly associated with glucocorticoid use at 6 to 9 months, with a predicted probability of 22% (95% CI, 21.7-22.7) among patients with providers in the highest preference quintile, compared with 11% (95% CI, 10.2-10.9) for those seeing a provider in lowest quintile.
“These results suggest that we need better evidence on the long-term risks and benefits of glucocorticoids to help guide clinical decision making and better standardize clinical practice,” George said. “Ultimately, we need to understand which treatment algorithms lead to the best outcomes for patients and find ways to promote these algorithms in the clinic.”