Healthcare-economics innovators looking for a model after which to fashion an episodic bundled-payment package for cancer care would do well to consider radiologist-led breast cancer screening.
In a study report published online Aug. 17 in the Journal of the American College of Radiology, researchers lay out a way such modeling could be done.
Lead author Danny Hughes, PhD, of the Harvey L. Neiman Health Policy Institute and colleagues from Weil Cornell, Tufts, Emory and George Mason University came up with their viable framework after examining a 5 percent national sample of retrospective Medicare claims data filed from 2009 to 2013.
To make sure these data were broadly representative across care settings, they used the same study design to pull in all mammographic claims data from a large private health system in the Northeast reimbursed by both CMS and private insurers from 2012 to 2014.
The team used these datasets to craft a detailed sample approach by which radiologists and other non-patient-facing physicians might participate in alternative payment models (APMs) as U.S. healthcare rushes along the “volume to value” road.
The authors note that, as these models have so far focused on treatment episodes and primary care providers, radiologists have found clarity hard to come by when they ask what roles are available for them.
“[I]f CMS is serious about transitioning 50 percent of fee-for-service payments to APMs by 2018, thoughtful consideration should be made about how radiologists and other specialists can participate,” the authors write. “Cancer screening episodes provide one such opportunity.”
Hence the authors’ framework (which would likely draw wide consideration beyond radiology should it become available as open-access content in the near future).
In their discussion, Hughes et al. stipulate that providers and payers would need to carefully examine their own data to calculate appropriate risk-adjusted bundle prices.
The authors also include in their report three notable insights that came to light as they analyzed the claims.
1. Service utilization is stable over time. The underlying patient demographics and distribution of care pathways were similar across three separate Medicare cohorts, resulting in less than a 1 percent difference in calculated bundle prices, the authors write.
“Although care must be made to ensure that appropriate prices are used for each service when calculating a bundled price in practice,” they add, “final bundle prices would likely not require substantive revisions over short periods of time.”
2. Including high-risk patients does not materially increase risk. The team found that patient demographics and mammography recall rates were similar in all screening patients as well as in high-risk excluded patient groups. Removing patients at high risk for breast cancer had almost no effect on calculated bundled prices, they observed.
“This suggests that radiology practices should be able to adopt bundles for all patients without incurring substantive risk, thus eliminating resource intensive efforts to identify and exclude high-risk patients from bundle participation.”
3. Bundle prices are sensitive to included services. Because all practices may not be able to offer the full scope of imaging services, the team provided bundle prices for different combinations of services.
“Increasing the scope of services offered in a bundle predictably increases the bundle price,” they write. “However, because additional services are distributed across all screening patients, the aggregated effect is mitigated. For example, including MRI in a mammography and ultrasound bundle increases the calculated episode price by only $3.38 (2 percent).”
Breast cancer screening “may provide a mechanism to expand the use of bundled payments in radiology and could serve as a framework for other episodic specialty bundles,” Hughes and colleagues conclude.
“Because screening bundles include costs for follow-up diagnostic imaging in addition to the initial screening mammographic examination, patient adherence to screening guidelines may improve, which may have profound effects on public health.”