The United States is in the midst of what the Federal Trade Commission (FTC) calls a “dramatic restructuring” of how doctors practice medicine, with a marked shift away from the traditional model of physicians owning their own small practices toward one where they work for hospitals.
Between 2010 and 2016, the percentage of primary care physicians employed by a hospital or a large healthcare system rose to an estimated 44% from 28%, the FTC said in an April overview of its plan to study of the effects of consolidation of physician practices.
Alan Weil, the editor-in-chief of the journal Health Affairs, also sees a need for more research on the effects of this consolidation. There are potential opportunities in mergers and acquisitions such as improving coordination of care, a point emphasized by hospital associations. But research indicates these transactions also seem likely to raise costs, Weil says.
“We have lots of reasons to think that consolidation drives up prices,” Weil says. “It reduces competition. It gives the seller of the services more leverage when they’re trying to negotiate prices.”
There are two major trends in in health care consolidation. There are combinations of large organizations that provide similar services, so-called horizontal mergers, which tend to draw scrutiny from journalists and antitrust agencies alike, amid concerns about such mergers driving up the cost of medical care. But hospitals also are steadily acquiring smaller physician practices — so-called vertical consolidation.
Some of these acquisitions don’t meet the minimum size-of-transaction bar to warrant antitrust investigations, explain economists Cory Capps, David Dranove, and Christopher Ody in a 2017 Health Affairs article.
The May 2021 edition of Health Affairs includes two research articles that reveal new information about the effects of vertical consolidations. In both studies, researchers show that practice patterns changed when physicians moved from private practice to a larger hospital system.
“They look at actual behavioral choices made by individual clinicians who move from being on their own into a larger system, and you can really quite directly see the changes in what they do, and that that has consequences.”
Alan Weil, editor-in-chief of the journal Health Affairs
In one of the Health Affairs papers, “Hospital Employment Of Physicians In Massachusetts Is Associated With Inappropriate Diagnostic Imaging,” Gary Young, director of the Center for Health Policy and Healthcare Research at Boston’s Northeastern University, and his coauthors examine magnetic resonance imaging (MRI) referrals for three common conditions for which the American College of Radiology describes these tests as being “usually not appropriate.” The ACR views are widely viewed as recommendations against use of MRIs in the early stages of treating these conditions. The researchers show that patients whose doctors moved from private practice to hospital employment were more likely to get referred for MRIs.
In the other paper, “Higher Medicare Spending On Imaging And Lab Services After Primary Care Physician Group Vertical Integration,” Christopher Whaley of Rand Corp. and his coauthors find that vertical integration of physician group practices with hospitals or health systems increases the use of hospitals for common diagnostic imaging and laboratory tests and increases Medicare payments to hospitals.
Both papers advance the discussion of how joining hospital systems may alter the way physicians practice medicine, says Weil, who earlier served as the executive director of the National Academy for State Health Policy and as executive director of Colorado’s Department of Health Care Policy and Financing. The researchers use data that allows them to home in on how choices of referrals differed.
“They look at actual behavioral choices made by individual clinicians who move from being on their own into a larger system, and you can really quite directly see the changes in what they do, and that that has consequences,” Weil says.
Inappropriate MRIs
In the paper focused on MRIs, Young and his coauthors build on earlier work that finds a substantial percentage of imaging procedures to be unnecessary or inappropriate. Yet diagnostic imaging remains a key source of revenue for hospitals. “Imaging services have been estimated to account for more than 30 percent of hospitals’ profits, and MRI scans specifically constitute a large portion of the high-margin imaging services that hospitals deliver,” they write.
Young’s coauthors on this paper were E. David Zepeda, clinical associate professor of health law, policy, and management at Boston University’s School of Public Health; Stephen Flaherty, a data scientist at Harvard Pilgrim Health Care, in Boston, Massachusetts, and an assistant professor, Meehan School of Business, Stonehill College, in Easton, Massachusetts; and Ngoc Thai, a PhD student in population health at Bouve College of Health Sciences at Northeastern University.
The researchers followed data on patient care as physician employment changed. To do this, they used the Massachusetts All Payer Claims Database, drawing information about claims from 2009 to 2016. This dataset included patient-level information for diagnoses, procedures, and demographic characteristics. The claims also included information about which physicians provided the services involved in the claim. The researchers then used Medicare and insurer data to identify which physicians were in private practice at the time of the service or referral, and which were affiliated with hospitals.
From this data, Young and his co-authors identified 583 primary care physicians who transitioned from private practice to hospital employment and a comparison group including 3,102 physicians who were not employed by a hospital during that period. Each group’s demographic characteristics were similar for physicians and patients in terms of average years of experience and age, respectively.
Young and co-authors report that the patients whose doctors had shifted into hospital employment were more likely to get referred for MRIs and for “inappropriate” MRIs – those that fell outside of the guidelines set by the ACR. The average overall percent of office visits that generated a referral for an MRI jumped to 16.5% from 11.6% in the cohort of physicians after they shifted to hospital employment. In contrast, the comparison group showed an average referral rate of 11.3%.
The researchers also focused on three conditions for which the guidelines set by the American College of Radiology (ACR), intend to prevent inappropriate imaging use. They selected three common conditions for which ACR describes routine use of MRIs as being “usually not appropriate”: uncomplicated lower back pain, meaning cases where there does not appear to be another serious medical condition causing this discomfort; nontraumatic knee pain without joint effusion, or knee pain not known to be caused in injury and not showing an abnormal swelling; and nontraumatic shoulder pain without joint effusion.
In examining the claims data, they set a 30-day window for considering whether MRIs for these conditions should be judged inappropriate. If an MRI for one of the three conditions was scheduled less than a month after a doctor’s visit, the researchers judged that there had not been time for trying other options such as asking the patients to exercise at home. The percent of MRIs deemed to be inappropriate rose to 29.8% from 25.1% among the physicians who transitioned to hospital-based practice. The percentage for the control group was 24.9%
Young and his co-authors acknowledge several limitations with their study. They note the claims data do not contain complete information for patients’ health conditions and so cannot fully capture the reasons why patients were referred for an MRI. It’s also possible that patients were more likely to seek referrals from hospital-employed physicians, they write.
Part of the challenge in reducing inappropriate use of medical treatments and tests such as MRIs will be educating patients as well as doctors about the costs of low-value care, says Young, director of the Center for Health Policy and Healthcare Research at Northeastern University. The money spent on these tests diverts financial resources that could be used elsewhere in health care. But patients may not factor that into their personal decisions.
“In some cases, it’s patients putting the pressure on the clinician and the clinician is responding to that and thinking ‘Well you know what, I’m going to keep my patient happy. The patient isn’t paying very much for the procedure, the hospital’s happy about this if I do this. So, it’s a win-win-win, but not really because in the end, we’re all losing,” Young says.
Young’s study was funded by a grant from the National Institute for Health Care Management Foundation. NIHCM Foundation also provides funding for The Journalist’s Resource.
There have been efforts for many years to curb excess use of medical tests, especially in cases where a seemingly harmless test can set off a cascade of subsequent treatments that expose a patient to harm. Physicians may act on incidental findings from any imaging studies, including MRIs, write Amanda M. Hall, assistant professor of Memorial University of Newfoundland, St John’s, Newfoundland, and co-authors, in a February 2021 article in The BMJ, a U.K.-based medical journal, titled “Do not routinely offer imaging for uncomplicated low back pain.”
“For example, disc abnormalities such as bulging disc or degenerative disc disease are commonly seen on images but may not be the source of pain, as they are also seen on images in up to 97% of asymptomatic patients,” they write. “Incidental findings may lead to further investigations, specialist referral, and more intensive treatment such as surgery, which limits access to those services for the patients who are in genuine need of such care.”
Hall’s co-authors on this paper are Kris Aubrey-Bassler, director of the Primary Healthcare Research Unit at Memorial University of Newfoundland; Bradley Thorne, a patient advisor at Memorial University of Newfoundland; and Chris G. Maher, director of the Institute for Musculoskeletal Health, University of Sydney.
Inappropriate testing done for uncomplicated back pain has been a target of the Choosing Wisely campaign, created by the ABIM Foundation, a nonprofit charitable organization created by the American Board of Internal Medicine. Through this initiative, more than 70 medical specialty societies have published recommendations regarding tests and treatments that they say have been overused. These include a recommendation addressing back pain, developed with the American Academy of Family Physicians.
This recommendation notes that most people with lower-back pain will feel better in about a month, whether or not they have an imaging test. People may fare better by taking simple steps such as walking and using over-the-counter medicines, the recommendation says.
“Why waste money on tests when they don’t help your pain? And if the tests lead to surgery, the costs can be much higher,” the recommendation says.
A link between hospital affiliations and more tests ordered
In the second paper, Whaley and co-authors report a link between hospital affiliations for clinicians and a rise in the number of tests ordered for people enrolled in the federal Medicare program as well as in reimbursement for these tests. Medicare is the nation’s largest purchaser of health care, covering about 61 million Americans who are age 65 and older or have disabilities that qualify them for the program.
Whaley’s co-authors on this paper were Xiaoxi Zhao, a PhD student in the department of economics at Boston University; Michael Richards, an associate professor of economics at Baylor University; and Cheryl L. Damberg, the RAND Distinguished Chair in Health Care Payment Policy and a principal senior researcher at the RAND Corporation.
The researchers drew from a pool of data for Medicare claims paid between 2013 and 2016. They used algorithms developed by the RAND Center of Excellence on Health System Performance to examine how the acquisition of physician practices affected orders for tests and imaging.
The monthly number of diagnostic imaging tests per 1,000 people represented in the Medicare claims pool, performed in a hospital setting, increased by 26.3 tests per 1,000 after the integration. The number of these tests performed in other settings decreased by 24.8 per 1,000. Hospital-based laboratory tests increased by 44.5 per 1,000 people enrolled in traditional Medicare and non-hospital-based laboratory tests decreased by 36.0 per 1,000.
Spending rose, too. Average Medicare reimbursement rose by $6.38 for imaging tests and $0.57 for laboratory tests That translates to $40.2 million and $32.9 million increases in Medicare spending, respectively, for the entire study period.
Whaley and co-authors note limitations to their work. They focused only on people enrolled in traditional Medicare rather than in insurer-run plans that manage Medicare benefits, commonly known as Medicare Advantage plans. The traditional Medicare program runs on a model of the federal government setting fixed prices for services, while the Medicare Advantage plan may involve more complex price negotiations. More than a third of people enrolled in Medicare — about 23 million — are in insurer-run plans.
In the paper, the researchers note it would be difficult to argue that people in traditional Medicare benefit from the extra spending on diagnostic tests. Instead, the increased spending reflects the different rates Medicare sets for services depending on the setting in which they are provided, they write. Such rates reimburse hospitals more for these tests than they reimburse competing sites of care, such as stand-alone imaging centers and freestanding diagnostic laboratory companies.
The higher rate of payment may reflect newer hospital systems that make it easier to order tests, Whaley says. That could make the difference in cases where a physician is weighing whether a patient needs a test.
In a case where a doctor isn’t sure if a patient needs a test, the convenience of ordering one within a hospital system may be a deciding factor, Whaley says.
“And if you do join a large system, I imagine that how much revenue you bill is tracking pretty well, so, there’s also probably a financial incentive as well,” he says.
The research reported in that paper was supported through the RAND Center of Excellence on Health System Performance, which is funded through a cooperative agreement with the federal Agency for Healthcare Research and Quality. Additional support was provided by the National Institute on Aging.
For more help on reporting on the effects of hospital consolidation, see Hospital mergers and acquisitions of physician practices: Research illuminates what’s at stake for consumers and Covering hospital mergers and acquisitions of physician practices: 3 tips from experienced health care journalists.
Expert Commentary