Primary Care: Getting Systematic

Health policy can be an imprecise practice. Nowhere is that imprecision more prevalent than in the promiscuous use of the word “system” in health-policy-speak.

I teach a public health course optimistically called “The US Health Care System.” (A primary aspiration for me is that students leave the semester skeptical that there is anything systematic about US health care.) Once a hospital acquires a second piece of real estate, it boldly proclaims itself a “health system.” Likewise, clinical providers bemoan the tyranny and obtuseness of the “electronic health records systems” they now use.

Labeling an item does not make it so, but you would not know that from the way we throw “system” around in health care. This is likely because, as a policymaker or a patient, we want US health care to have at least some system attributes: relatively stable inputs and outputs, interrelated elements that interact regularly within real or figurative boundaries, common goals, and — perhaps most importantly — accountability and reliability. A “system” also implies that some sort of central control or governance function exists.

Only the rosiest of glasses could detect any those characteristics in US health care.

A notable exception is the Veteran’s Administration, which runs a true health care system. With a fixed budget and a defined patient population, it has clear inputs and outputs and common goals. Its employed clinicians and its delivery sites constitute interacting elements within it. There is oversight and accountability. As a result, its performance is relatively reliable and can be adjusted to meet those agreed-to goals.

On a smaller scale, community health centers in the country contain some aspects of a system as well. The Bureau of Primary Health Care, in exchange for federal grants and a statutorily mandated payment formula, has strict accountability requirements that help move the health centers toward specific access and clinical quality goals.

But most of health care in the US has no such system-ness, even within rapidly consolidating mega wannabe health systems. Autonomous health providers treat individual patients on a fee-for-service basis with no overall resource constraints or population-level goals. Patients are free (or doomed) to wander among these providers. Licensing, accreditation, credentialing, and liability suits attempt to weed out bad apples, but that is the extent of accountability and oversight.

That is why the continued progress of the Maryland Primary Care Program (MDPCP) is notable. A younger sibling of the state’s 36-year-old Medicare waiver program establishing all-payer payment for hospitals, MDPCP consists of four interventions for participating primary care practices: an aligned hybrid (per capita and fee-for-service) payment method, shared practice transformation and care coordination resources for the practices, aligned quality measures, and a functioning health information exchange (HIE).

The interactions of these system elements are coordinated, not merely allowed to happen. Program management at the Maryland Department of Health oversees the size and structure of the payments, the nature of the practice improvement activities, the quality measurement activities, and the reporting from the HIE.

The result for participating practices is something that starts to resemble a primary care system in the Old Line State: populations (or inputs) attributed to primary care teams, relative budget stability not driven by visit volume but per-person payments, better alignment of goals, and more consistent interactions among clinician teams and patients in pursuit of those goals.

Maryland has reaped the benefits of this system-thinking during the COVID 19 pandemic. MDPCP-enrolled practices serve about half the population. As MDPCP researchers outlined in a recent Milbank Memorial Fund issue brief, when the pandemic struck:

• The state department of health had an established communication channel for pandemic guidance, offering multiple webinars a week
• The HIE identified each practice’s high-risk patients and prioritized them for outreach, testing and vaccination
• COVID-related activities could be carried out by the care transformation organizations set up under MDPCP as shared resources for smaller practices
• MDPCP was able to quickly facilitate telehealth licenses for participating practices, allowing them to stay open and deliver care in the uncertain early months of the pandemic

This emerging system of primary care permits the actual integration of primary care and public health — a long-hoped-for concept expected to deliver preventive services and keep people healthy.

Has it worked? It seems so.

Analysis of Maryland data for Medicare beneficiaries shows that compared to similar patients in other primary care practices in the state, patients enrolled in MDPCP practices, at statistically significant levels, were less likely to:

• Get COVID
• Be admitted to the hospital with a COVID diagnosis
• Die from COVID

These are remarkable outcomes. It appears that lives are being saved. Could individual primary care practices have achieved these results? Probably – there will always be high performers. And if some of our private so-called health systems are truly systematic (bounded, stable, aligned in inputs, outputs, and goals) they might perform that way as well. But Maryland, by taking a universal approach, has developed a primary care system that can partner with public health systems in a comprehensive way that individual providers cannot.