Primary Care as a Platform for Value

Quantifying the worth of primary care and its delivery are essential for raising its profile. New data and payment incentives may push us closer.

David M. Levine, MD, tells a story that pretty much captures the value that policy makers have placed on primary care in America. In essence, says Levine—a clinical investigator in the Division of General Internal Medicine at Brigham and Women’s Hospital and Harvard Medical School—it comes down to favoring physicians who treat with their hands over those who treat with their heads.

“I remember in medical school working with an amazing primary care doctor in St. Louis. He told me, ‘David, did you see that wart I just cut off her shoulder? I just made more doing that than I will all week.’”

As the U.S. health care system inches from fee for service toward value-based payment, primary care has what may be its best opportunity in decades to smooth out inequities in pay. But change will require evidence of value—not just that of primary care, but of the primary care provider as well.

Massive study quantifies value

Levine, Bruce Landon, MD, and Jeffrey Linder, MD, took on the challenge of showing what primary care brings to the table. In a study published earlier this year in JAMA Internal Medicine, the trio looked for associations between primary care delivery and receipt of high- or low-value care, access, and patient experience.

The 4 C’s of primary care

  • First-contact care
  • Comprehensive care
  • Continuous care
  • Coordinated care

In their study of the value of primary care, published in JAMA Internal Medicine earlier this year, David M. Levine, MD, and colleagues limited their definition of receipt of primary care to people who said they would visit their usual source of care for new health problems (first contact), preventive care (comprehensive care), an ongoing health problem (continuous care), and referrals to other health care professionals (coordinated care).

The researchers compared outcomes of two groups of patients. One group had a usual source of primary care—a source that also fulfilled what health services researcher Barbara Starfield identified as the “4 C’s” of primary care (see box). The other group also got care (as opposed to no care)—but from sources other than a primary care physician, like specialists or the emergency department. The study accounted for 70,000 adults between 2012 and 2014, and outcomes were propensity weighted to guard against clinical and sociodemographic confounders.

Levine’s team aggregated 45 clinical-quality and patient-experience measures into 12 composites (see “12 Ways to Determine Value”). Many of the findings were what you might expect: Those with a usual source of primary care got more preventive care and prescriptions and indicated an overall better patient experience. The other group got a similar amount of care, although on the whole, these patients received fewer high-value cancer screenings and diagnostic tests.

12 ways to determine value

In their JAMA Internal Medicine examination of associations between the delivery of primary care and the receipt of high- or low-value care, access, or a positive patient experience, David M. Levine, MD, and colleagues aggregated 38 clinical-quality and seven patient-experience measures into 12 composites. High-value composites were those for which the delivery of the service was likely to benefit the patient. Low-value composites defined services in which delivery would have been either inappropriate or of little to no benefit.

High-value composites

  1. Certain cancer screenings (cervical, breast, colorectal)
  2. Diagnostic and preventive testing
  3. Diabetes care
  4. Counseling (weight loss, exercise, smoking cessation)
  5. Medical care (12 treatments, e.g., anticoagulation for atrial fibrillation; controller medication for poorly controlled asthma or COPD)

Low-value composites

  1. Certain cancer screenings (cervical, colorectal, or prostate) in older adults
  2. Antibiotic use for flu or upper respiratory tract infection
  3. Imaging (e.g., MRI/CT for back pain or headache)
  4. Medical care (five treatments, e.g., opioids for headache or back pain; NSAIDs for hypertension, heart failure, or kidney disease)

Patient-experience composites

  1. Global rating of health care
  2. Communication with physician
  3. Access to care

Some other findings, though, were more of a surprise. Notably, receipt of low-value care—defined as certain cancer screenings, imaging procedures, and prescriptions—was similar in the two groups, and low-value antibiotic prescribing was actually higher among those with a usual source of primary care.

“Intuitively, it makes sense if you have a relationship with a primary care physician you’re probably going to get less low-value care, but we didn’t see that,” says Levine. “It may mean that primary care as it’s currently set up is not as well equipped as we would like it to be to defend against low-value care. There’s not the right set of incentives right now to combat all the low-value care that’s going on. It’s really a place where primary care can grow.”

But what of the physician?

In an invited commentary published alongside Levine’s study, Allan Goroll, MD, of Massachusetts General Hospital remarked that while the paper provides evidence of value of primary care, it doesn’t tell us about the value of the primary care physician. From the composites examined in the study, he wrote, it appears that PCPs “are spending most of their time arranging prevention, performing screening, and offering counseling while their physician-level diagnostic and management skills seem to go unused and might be atrophying.”

Shawn Martin, senior vice president of advocacy, practice advancement, and policy at the American Academy of Family Physicians has a quick response: The physician’s value lies in the ability to provide comprehensive care.

In some respects, the check-box, metrics-driven foundation for value-based payment devalues the primary care physician’s expertise. “We measure vaccine rates, we measure screening exams, we measure depression scores,” says Martin, “and very seldom do we create a system that holistically looks at the collective outcome of all of these things being done.”

The holistic approach provides no immediate gratification, nor is it flashy. It’s the antithesis to America’s love affair with superheroes and, as Martin wrote in a blog on the AAFP website, our health care system “has become entrenched in the hero mentality.”

“That’s not to take away from cardiology or oncology or orthopedic surgery,” Martin tells Managed Care. “We need that. It’s just that we as a health care system have almost said to ourselves, ‘It doesn’t matter what we do from Day 1 to Day 9, because on Day 10 we have all these amazing specialist services and high-end drugs that are going to save the day.’”

In their seminal meta-analysis in Milbank Quarterly in 2005, Starfield and colleagues examined primary care’s contribution to health status. They concluded that physician supply is associated with mortality rates, calculating that an increase of one primary care physician per 10,000 people would avert 127,000 U.S. deaths annually.

Dialing for dollars

As persuasive as that may sound, there is plenty of debate over the value of primary care. More specifically, expenditure on it.

Levine believes that underfunded primary care is a principal cause of poor health system performance. “We don’t spend very much in America compared to our contemporaries in the developed world,” he says. “We see that in microcosms everywhere. We see that in how much different physicians get reimbursed for different services. We see that in how much hospitals get reimbursed for different services. We see that in where trainees decide to practice.”

Martin agrees that expenditure is at issue, but views it through a different lens. Martin thinks that any definition of value based on per-capita spending thresholds misses the mark. A better metric, he believes, is “‘How long can we [avoid] what we know will be high per-capita spending?’ If we can keep people healthier longer, and if they’re 80 before we spend $13,000 per year versus 50, maybe that’s the value.”

As a policy matter, talk of investing in primary care doesn’t answer the more salient question: Where to invest the money? In a 2010 Health Affairs paper, a Rand Corp. team led by Mark Friedberg, MD, asked whether the biggest bang lies in increasing the number of primary care physicians, reshaping primary care as a set of functions, or focusing on health system orientation. Starfield’s study made a case for physician supply, while most quality measures today promote primary care functions. Levine seems to opt for health-system improvement.

“Primary care needs to be a captain in a value-based system,” says Levine. “It’s refreshing when a health system is put into a value-based arrangement, and they form an ACO, for example. And all of a sudden, ‘Wow, we need phenomenal primary care—otherwise, we are going to be in the red.’ We see example after example of that.”

CMS makes a move

CMS acknowledged the role of primary care in high-functioning systems last April when it proposed the Primary Care First payment model. Similar in some respects to the payment structure of the Oncology Care Model, PCF is a two-sided risk program that provides two payments to clinicians. One is a risk-adjusted, per-beneficiary, per-month payment for services during or outside of a visit, such as care-coordination activities or phone calls to patients. The other is a flat $50.52 fee for office visits. Total of both payments is adjusted up or down based on quality performance.

AAFP collaborated with CMS on the development of the model. “You basically eliminate coding and documentation to justify 99213, 99214, 99215 [lengthier visits for established patients]. You create a global payment for your attributed lives,” says Martin. “[CMS] kept the idea there should be a wraparound fee for service for in-office procedures—that’s driving that comprehensiveness.”

But while PCF represents a different way of funding primary care delivery, it doesn’t put any new money into it. So noted Laura Sessums, MD, Sanjay Basu, MD, and Landon in a commentary published July 17 on the New England Journal of Medicine website. What’s more, they argue, PCF may move some less technologically advanced primary care practices toward population-based payment before they are ready for it.

Whatever the strengths or weaknesses of the model, the authors were at least encouraged that primary care is at last getting attention from CMS. To Martin, it’s a welcome change.

“Look at any academic medical center in the country. Primary care is in the basement. I think it’s symbolic of our approach to health care. We just don’t prioritize health and maintenance and we way overprioritize these miraculous interventions downstream.”