Photo by Thinkstock
Photo by Thinkstock.

PCPs in the hospital

Innovators are taking steps to help banish care discontinuity once and for all, with some especially focusing on the role of primary care physicians (PCPs) in the hospital.

Discontinuity between inpatient and outpatient care has always been the Achilles heel of hospitalist practice, but perhaps not for much longer.

The problem has improved over the past decade, according to Robert M. Wachter, MD, FACP, professor and chair of the department of medicine at the University of California, San Francisco. He credits those improvements to changes ranging from new handoff processes to the digitization of the U.S. health care system.

“In general, I think it's getting better, in part because good hospitalist programs focus on it and have created systems that help improve it,” said Dr. Wachter, who coined the term “hospitalist” in 1996. “That's everything from medication reconciliation to postdischarge phone calls to patients.”

Now, researchers and innovators are taking further steps to help banish discontinuity once and for all, with some especially focusing on the role of primary care physicians (PCPs) in the hospital.

Hospitalist or PCP?

Over the past three decades, more and more PCPs have allowed hospitalists to assume the care of their hospitalized patients. This is because fee-for-service paid them so poorly that they didn't have time to go to the hospital anymore, explained Allan H. Goroll, MD, MACP, a physician at Massachusetts General Hospital and professor of medicine at Harvard Medical School in Boston.

“The reason why primary care doctors stopped going and serving as the attending was not because they didn't want to see their patients in the hospital; it was because it became logistically and financially impossible to do,” he said. The need to shorten length of stay and remain current in inpatient care measures and guidelines also contributed to the disappearance of PCPs from the wards, Dr. Goroll added. “So they were both not able and not welcome,” he said.

But recent studies and pilots that have looked at outcomes besides length of stay indicate that there might be some reason to consider bringing PCPs back to the hospital.

An analysis of more than 560,000 Medicare admissions found that patients who were cared for by their PCP in the hospital were 14% more likely to be discharged home, were 6% less likely to die within 30 days, and had 12% longer lengths of stay compared to those cared for by hospitalists, according to results published in November 2017 by JAMA Internal Medicine.

The mortality finding, while not entirely surprising, was striking to C. Seth Landefeld, MD, FACP, chair of the department of medicine at the University of Alabama at Birmingham. “One of the things that's so interesting is that this question hadn't really been looked at before. It's only now that we're really taking advantage of the Medicare data to be able to look at that,” he said.

The results deserve further evaluation and give grounds for testing the “reasonable” hypothesis that having continuity of care from a PCP through all care settings may improve such outcomes as discharge to home, short-term mortality, and possibly patient and family satisfaction, Dr. Landefeld said. Although hospitalists and PCPs may, to some degree, have differing domains of medical expertise, what PCPs bring to the table is knowledge of the context of the patient's life and illness, he said.

In the study, which looked at data from 2013, PCPs cared for only 14.2% of the hospitalized Medicare patients, compared to hospitalists, who cared for 59.7% of the cohort (other generalists handled the remaining 26.1%). Even 30 years ago, being cared for by your PCP in the hospital was not universal, “but it's certainly less common today than it was,” noted Dr. Landefeld.

He said the study provides a basis, particularly for integrated health systems, to ask if it makes sense to have a continuity-of-care system in place for patients with serious chronic illnesses, wherein a PCP and his or her team provides care over the course of the illness, including during hospitalization.

Reimagining inpatient care

Most PCPs are still paid on a fee-for-service basis. But in recent years, the Center for Medicare & Medicaid Innovation (CMMI) has rolled out its Comprehensive Primary Care Plus payment model at thousands of U.S. practices. “We're getting rid of high-volume, low-payment practice,” Dr. Goroll said. “As payment moves from retrospective fee-for-service in primary care to prospective risk-adjusted comprehensive payment, you're going to see a change in how physicians allocate their time.”

This payment reform gives PCPs the opportunity to return to the hospital, raising the question, “What should the primary care doctor's role be in the hospital now that they are freed up to come in?” said Dr. Goroll.

He noted that the division of general internal medicine and the hospitalist service at Mass General are in the process of jointly exploring design of a system that makes the PCP a collaborating member of the inpatient team.

In a model he described with a hospitalist colleague in a 2015 New England Journal of Medicine perspective, the PCP would perform an in-person consultation on admission, be available (in person or virtually) during hospitalization, and participate in design of the discharge plan at the end of the stay. The admission consultation would provide the hospitalist team with a distillation of essential information about the patient, including a focused history, any relevant psychosocial issues and patient preferences, and recommendations for scope and intensity of workup and management, Dr. Goroll said.

“That should save the hospitalist team precious hours and stress, facilitate delivery of personalized care, better support patients and their families, and make for a more efficient hospitalization, not to mention hopefully reduce the readmission rate,” he said. “The time commitment should be modest. The goal is to enhance, not duplicate or interfere with care; it should be welcome by all and enhance professional satisfaction.”

How to pay for the PCPs' inpatient work has yet to be determined. It could be paid for under traditional payment systems as a fee-for-service consultation or built into a risk-adjusted comprehensive payment model, Dr. Goroll said. “By the way, avoiding redundant and unnecessary testing and saving just one readmission could pay for a year's worth of payment to a PCP,” he added.

A separate initiative, the Comprehensive Care Physician (CCP) model, focuses on patients at high risk of hospitalization (i.e., those who've been hospitalized in the past year) by having their PCPs (called CCPs) care for them both in clinic and in the hospital. The CCPs have a small panel of patients, caring for patients in the hospital in the morning and seeing them in clinic in the afternoon, said David O. Meltzer, MD, PhD, FACP, principal investigator of the program, which is funded by a $6.1 million grant from CMMI.

“Even that small panel of 200 to 300 patients will produce enough patients in the hospital each morning that it makes sense for the doctor to reserve their mornings to go see those patients, whereas a typical primary care doctor can't do that,” he said.

The University of Chicago in Illinois has a group of five CCPs, who assume an attending role in the hospital and essentially serve as both hospitalists and primary care doctors, said Dr. Meltzer, professor of medicine and chief of the section of hospital medicine. “All of the elements of the doctor-patient relationship that are really critical here—the knowledge of the patient, the trust, the communication, the interpersonal relationship—are all enhanced by continuity,” he said.

Monday through Friday, the CCPs all come to the hospital, round on their own patients, and write orders. In the late morning, they attend multidisciplinary rounds, and one of the five (they rotate every fifth week) then stays in the hospital through the afternoons and over the weekend to care for his or her colleagues' patients while they go off to clinic, Dr. Meltzer explained.

CCP programs are now developing at about half a dozen locations around the country and at least one international site, said Dr. Meltzer. Vanderbilt University in Nashville, Kaiser Permanente's Mid-Atlantic region, and the University of Singapore are all emulating the model in various ways, he said.

Researchers evaluated the program, which began in November 2012, with a 2,000-person randomized controlled trial. They've gathered one year of follow-up data and are currently submitting abstracts to be presented this year at national conferences, said Dr. Meltzer, who is mum about the outcomes, which include both quality and cost of care. “We're enthusiastic about the program, and we're going to be presenting the results soon,” he said.

Potential challenges

Hospitalists agree that PCPs can provide valuable input on inpatient care, but while some support the principles behind the proposed new models, others have some reservations.

“I feel there are situations where it's really helpful to have people who know [patients] the best in the hospital, or at least in the loop, helping with difficult decisions like transitioning to hospice or maybe changing medications that need close management,” said Christine D. Jones, MD, MS, assistant professor of medicine and director of care transitions for hospital medicine at the University of Colorado in Denver.

Indeed, a major complaint among hospitalists is that the PCP is absent when a complex patient is hospitalized, said Dr. Goroll. “The hospitalists face very difficult situations when they get a complex patient and a very worried and involved family to deal with, and they're swamped,” he said.

However, challenges include the fact that PCPs may not be as able as hospitalists to deal with the sudden arrival of families who want clinical updates, said Jürgen L. Holleck, MD, ACP Member, a hospitalist at Veterans Affairs Connecticut Healthcare System in West Haven and assistant professor at Yale University School of Medicine in New Haven. “We can kind of adapt and respond to that. I think a PCP in that regard would be hard pressed to do that, even having the advantage of that longstanding relationship and, in many cases, implicit trust that they've built up over time,” he said.

Dr. Holleck also expressed concerns about PCPs' inpatient skills. “Inpatient care has gotten so specialized that I think many PCPs may actually not feel comfortable in the hospital doing inpatient medicine, especially when they're not doing it all the time....I think that it would actually be very challenging for most PCPs to end up back in the inpatient setting,” he said.

He noted that hospitalists can get things done much more efficiently in the hospital than PCPs because they know exactly whom to call, where to go, and how to order.

Dr. Wachter expressed similar doubts about returning to what was the standard approach until the mid-1990s. “It's not really going back to the future; it's reimagining the old model in an effort to try to create more continuity for those high-risk patients that really need it,” he said. “But understand that it's got to look very different from the old model for it to have any chance at succeeding.”

Challenges of a comprehensive-care model include finding doctors who are good at both inpatient and outpatient care and, of course, sorting out the economics, said Dr. Wachter. The average PCP has a panel of 2,000 patients, and to cut that down sufficiently to make it feasible for the PCP to substantially contribute to inpatient care would mean either having a concierge practice or a system in place that infuses money into the practice, he said. “If you can sort those things out, then I think this is a very attractive model, but it's not a slam dunk,” Dr. Wachter said.

Overhauling the model of care to improve continuity might be attractive, but it's tricky to accomplish logistically, he added. “Today's system of having a primary care doctor who is in the office all day focusing on doing the best they can for those patients and the hospitalists doing the same in the hospital, and then working really hard to build bridges between those two worlds, it's not obvious to me that's the wrong system,” Dr. Wachter said.

However, he added that he's looking forward to seeing published results from the CCP program. “If it really does work very, very well, then I think we'll take a good hard look at it and see whether it's worth it,” Dr. Wachter said.