Ishani Ganguli is a primary care physician, health services researcher and assistant professor of medicine at Harvard Medical School and Brigham and Women’s Hospital and a member of the 2021 Tradeoffs Research Council. She studies medical decision-making and delivery and payment innovation in ambulatory care.

As health systems increasingly buy up physician practices, researchers and policymakers are closely watching the consequences of this so-called vertical integration. The theoretical benefits of integration like better coordination between primary care doctors and specialists have yet to be widely supported by evidence. And two new studies in Health Affairs looking at how newly employed doctors change how they order tests add to growing evidence that this kind of consolidation increases health care spending. 

First, Christopher Whaley, Xiaoxi Zhao, Michael Richards and Cheryl Damberg used national Medicare billing data to track where physicians sent their patients for 10 common imaging and laboratory tests before and after health system integration, compared to those in practices that were always or never integrated. Post-integration, they found doctors ordered slightly more imaging and lab tests overall and shifted many of them to their affiliated hospital (which can tack on extra facility fees), leading to an estimated $73 million in additional Medicare spending.

Second, Gary Young, E. David Zepeda, Stephen Flaherty and Ngoc Thai looked at whether primary care physicians (PCPs) newly employed at a health system ordered more inappropriate MRIs for otherwise healthy adults with one of three common conditions with clear clinical guidelines on when MRIs are appropriate: uncomplicated lower back pain, knee pain or shoulder pain. Using claims data from payers across Massachusetts, the researchers found that after their doctor was employed by a health system, a given patient had a net 18% higher chance of getting an inappropriate MRI compared to patients seeing non-employed PCPs, and most of those MRIs were done at the health systems where the doctors were employed.

I’m not surprised PCPs ordered more tests at the health systems that employed them given the financial incentives for health systems as well as the practicality of clinical workflows and limitations of working across different electronic health record (EHR) systems. When I order a test through my health system’s EHR, it is automatically set up to be performed within the system. There is no direct way for me to order a test somewhere else or to see those results that doesn’t involve several phone calls and a fax machine. But I was struck by the finding that PCPs ordered more low-value MRIs post-employment, which could reflect the relative convenience in a health system of ordering MRIs or of referring patients to specialists who request them. 

Both sets of authors suggest Medicare shift to a site-neutral payment policy — where doctors get paid the same for a procedure whether it happens in or outside of a hospital — as one way to mitigate the financial impacts of vertical integration. While many argue that change would also reduce health systems’ incentives for buying up physician practices, it won’t necessarily help the many PCPs choosing to join health systems because the alternative – running a small business – is increasingly untenable.

Since it’s unlikely we can significantly reverse this integration trend, I’m still hopeful health systems can offer promised benefits like more efficient workflows and centralized services that help me take better care of my patients. I see examples of these benefits in my practice that are difficult to measure. But as these studies make clear, there are consequences to integration that require thoughtful policy solutions.