A study of a three-year multi-practice patient-centered medical home pilot finds lackluster results. "We're now concerned that medical home transformations may not really achieve the goals set out for them at the rapid pace that people have hoped," says the study author.
Yet another report showing lackluster quality improvement in physician practices certified as patient centered medical homes—after three years of concerted effort—sparks the question: why is it so difficult to get better results?
The report, published in Wednesday's Journal of the American Medical Association, compared quality, utilization, and costs of care delivered to about 120,000 patients in 32 Pennsylvania practices. About half of the patients were treated by physicians in PCMHs certified or recognized by the National Committee for Quality Assurance; the others were treated by physicians in traditional practices.
The 2008–2011 report found that compared with traditional practices, recognized NCQA PCMHs did no better at controlling costs, or on 10 of 11 quality measures evaluated, such as cholesterol testing and cervical cancer screening, or in avoiding emergency room visits of patients who could have been seen in an ambulatory setting.
The only measure where some improvement was seen in the medical home groups was in nephropathy screening for kidney disease in patients with diabetes.
Though these pilot practices were able to achieve NCQA certification and adopted registries designed to identify patients in need of certain chronic disease services, that didn't result in fewer patients making trips to the emergency department or fewer patients getting admitted to the hospital.
More Time May Be Needed
"It turns out that taking the patient-centered medical home concept out into the real world is a whole 'nother kettle of fish," says Mark Friedberg, MD, a RAND Corporation scientist and practicing internist at Brigham and Women's Hospital. Friedberg is the study's lead author.
"When you take a system [primary care] that's already under stress and potentially apply even more stress by asking it to transform, it may not be able to respond as you hope, especially over the short time period that most pilot projects run, and this one ran much longer, three years."
He adds that results from this and several other trials show that changing physicians' behavior to do more for their patients outside the office, with reminders and communication and follow-up, "is just really hard to do. And we're now concerned that medical home transformations may not really achieve the goals set out for them at the rapid pace that people have hoped."
It may, in fact, take a lot longer than three years, he says.
In an accompanying editorial, Thomas Schwenk, MD, of the University of Nevada School of Medicine in Reno and Las Vegas, wrote that the PCMH model has been held up as "the foundation for all primary care delivery, including the solo and small group practices that dominate the primary care delivery system."
But their success has been demonstrated primarily in "highly integrated healthcare systems and single-payer community-based practices."
2 Key Lessons
Advocates for the medical home concept may be disappointed, he added, "but they should pay close attention to the study's lessons."
Friedberg says there are two key lessons from the Pennsylvania project:
"First, it's not a foregone conclusion that transforming and receiving recognition as a medical home will produce all the expected benefits of a medical home. And lesson number 2 is we still need to experiment with the model. It's not that we know what works and we just need to implement it at a bigger scale. This argues that we need to understand how to make this model effective. "
The concept of the patient-centered medical home involves making sure that patients are followed up. For example, patients with diabetes may be called at home if they haven't scheduled an appointment in six months, or invited to speak with a nurse on the phone.
Friedberg pointed to other research he and others conducted on a similar pilot model in Rhode Island, that showed "no significant improvements were found in any of the quality measures." That work was published in September.
NCQA Responds
The questionable usefulness of NCQA certification or recognition, which Friedberg says requires "a major effort and extra time, even the process of submitting the NCQA application, to receive recognition," prompted a NCQA response.
The NCQA said the pilots in the Friedberg study were rated on "outdated" NCQA standards for patient-centered medical homes that have since been revised. The organization said Friedberg's report "contradicts several others that have shown improvements in cost, quality, access, and patient experience." That evidence is why "most states and many private and commercial insurers support PCMHs with financial or technical support."
The NCQA intends to give an additional update on its requirements with new standards in March.
In the Pennsylvania pilot, commercial payers and Medicaid health plans paid the physician practices in the pilot about $92,000 per practice in bonuses, which they could have used to expand or support personnel or technology in an effort to improve care to patients beyond the office visit. They also received traditional fee for service payments.
There was no payment incentive for keeping patients out of the hospital, or for avoiding emergency department visits for medical complaints that could have been handled in an outpatient setting. That may be the next thing to come.
An Anthropologist's View
Ben Crabtree, an anthropologist who has teamed up with physician researchers under a Robert Wood Johnson grant to evaluate pilot models of innovative practices including PCMHs, says physician practices need to be totally redesigned with a care coordinator who is part of the team in order to make the concept work.
The main obstacle to improved performance, he says, is the apparent difficulty physicians have in redistributing workload in their practices. "They were socialized that the physician does everything, that they're the captain of their ship," Crabtree says.
"Medical school does that to them. But now there is the reluctance among some physicians to give up autonomy. They need to have a more distributed leadership style."
Care coordinators need to work with nurse practitioners, physicians, and others in the practice. And they need to have certain skill sets, formalized training programs for which do not yet exist.
"Most of the places that are really successful at this are training their own people, but that's not a very viable model for broad dissemination," Crabtree says.
Friedberg says that the researchers were "surprised" at the lackluster results from this three-year study because these practices that were patient-centered medical homes were large, and had been among the longest running PCMH model practices in the country.
In summary, Friedberg says that the study demonstrates loud and clear that doctors and policy makers "are still really new at this. It's not an area where we've had a lot of experience, and so we're definitely still in an experimental phase.
"We really won't know what works until we test it."