Understanding the factors that affect the appropriateness of rheumatology referrals.

BACKGROUND: Reducing inappropriate referrals to specialists is a challenge for the healthcare system as it seeks to transition from volume to value-based healthcare. Given the projection of a severe shortage of rheumatologists in the near future, innovative strategies to decrease demand for rheumatology services may prove more fruitful than increasing the supply of rheumatologists. Efforts to increase appropriate utilization through reductions in capacity may have the unintended consequence of reducing appropriate care as well. This highlights the challenges in increasing the appropriate use of high cost services as the health system transitions to value based care. The objective of this study was to analyze factors affecting appropriateness of rheumatology services. METHODS: This was a cross-sectional study of patients receiving Rheumatology services between November 2013 and October 2019. We used a proxy for "appropriateness": whether or not there was any follow-up care after the first appointment. Results from regression analysis and physicians' chart reviews were compared using an inter-rater reliability measure (kappa). Data was drawn from the EHR 2013-2019. RESULTS: We found that inappropriate referrals increased 14.3% when a new rheumatologist was hired, which increased to 14.8% after wash-out period of 6 months; 15.7% after 12 months; 15.5% after 18 months and 16.7% after 18 months. Other factors influencing appropriateness of referrals included severity of disease, gender and insurance type, but not specialty of referring provider. CONCLUSIONS: Given the projection of a severe shortage of rheumatologists in the near future, innovative strategies to decrease demand for rheumatology services may prove more fruitful than increasing the supply of rheumatologists. Innovative strategies to decrease demand for rheumatology services may prove more fruitful than increasing the supply of rheumatologists. These findings may apply to other specialties as well. This study is relevant for health care systems that are implementing value-based payment models aimed at reducing inappropriate care.

Discrepancy between preference and actual adjuvant therapy for breast cancer.

PURPOSE: Pretreatment preferences for adjuvant therapy were examined and compared with actual treatment received. PATIENTS AND METHODS: Before definitive surgery, women with node-negative breast cancer were asked to indicate their preference for adjuvant therapy in response to 3 different clinical scenarios. The scenarios provided precise risk, benefit, and side effect information with low-, moderate-, or high-risk risk of death from breast cancer. Contingency table and Spearman rank correlation coefficients were used to examine associations. Kruskal-Wallis rank sum tests were used for group comparisons, with the Friedman rank sum test being used for correlated samples. RESULTS: A total of 75 women enrolled between February 2002 and April 2005; 24% were aged > 65 years. After definitive surgery, 21% of women had ductal carcinoma in situ, and 89% had receptor-positive disease. There was a significant correlation between risk of recurrence and aggressiveness of treatment preferred (P < .001). After surgical staging, the high-risk group received more aggressive treatment compared with the low-risk group (P = .004). In the 51 women with invasive receptor-positive tumors, there was a significant difference (P = .002) in aggressiveness of treatment received based on risk of recurrence. Only 45% of the women received what they had preferred for the level of their risk before surgery. Women were more likely to receive a less aggressive therapy than they preferred initially (P = .0002). CONCLUSION: This study is among the first to correlate pretreatment preference for therapy with the actual therapy received. Less than half of women received their indicated preference before definitive surgery, with most women receiving less aggressive therapy. Future studies will need to examine this discrepancy.