Getting More Primary Care-Oriented: Measuring Primary Care Spending

We all have our obsessions. Many are best not shared, although sometimes airing them is helpful, especially when they are informed by research.

One of my persistent professional preoccupations is the US health care system’s stubborn aversion to orient itself to the fundamental importance of primary care. Fortunately, I am not alone in being somewhat haunted by this issue—and a new report by the Fund attempts to shed some light on this topic using the time-tested method of following the money.

Primary care is a great investment for a high-performing health care system. Research demonstrates that greater use of primary care is associated with lower costs, higher patient satisfaction, fewer hospitalizations and emergency department visits, and lower mortality. Within the United States, health care markets with a larger percentage of primary care physicians spend less and provide higher quality of care. Globally, almost all developed countries have a higher ratio of primary care to specialty care spending than the United States, and enjoy lower costs and higher life expectancy.

But the US health system does not reflect this evidence. Historically, researchers have looked at two measures to demonstrate this inconsistency—comparing average salaries of specialists to primary care physicians and comparing the total supply of specialists to primary care doctors. Both ratios are higher in the United States than elsewhere—indicating an underinvestment in primary care. Policymakers and payers alike, however, have been evidently unmoved by these measures.

A new measure of primary care orientation—a phrase first coined by Eric Schneider and Mark Friedberg in a 2010 Health Affairs article—has emerged in more regular use in the United States in the last few years. “Primary care spending,” or the ratio of spending for primary care services to spending for all medical services, has the potential both to sharpen the focus on the problem of underinvestment in primary care and motivate some policy solutions.

The primary care spending measure has several advantages over other measures of primary care orientation.  It is easily understood by policy wonks and lay people alike; when I ask them, non-health care types reliably and significantly overestimate the portion of the health care pie that goes to their primary care doctor. The measure encourages clear, financial accountability for whoever is spending the health care money—whether it’s insurers, integrated delivery systems, or public payers. Finally, it creates opportunities for learning about and possibly changing the political and economic forces that create low rates of spending on primary care in the United States.

While measures of primary care spending have been used internationally, their take-up in the United State’s multi-payer, financially fragmented environment has been slow. Not surprisingly then, it’s the states themselves that have been the policy innovators in developing and deploying this measure. Rhode Island, persuaded by the evidence and after five years of monitoring primary care spending levels, requires their commercial insurers to spend at least 10.5% of their dollars on primary care. Oregon has no requirement but has been measuring primary care spending for all of its commercial and Medicare insurers and Medicaid coordinated care organizations. Various state all-payer claims database (APCD) analysts have started to look at it as well.

Several issues have emerged with this state experimentation, however. Definitions of primary care that have been used are inconsistent and highly variable. This makes cross-state comparisons and benchmarking difficult. Administration of payments to primary care physicians—for insurance claims, alternative payments, or salaries—has been inconsistent and sometimes hard to capture. Finally, people interested in this question must have access either to the raw data or willing (or compelled) payers to produce the figures themselves in consistent ways.

Our new Fund report, Standardizing the Measurement of Commercial Health Plan Primary Care Spending, by Michael H. Bailit, Mark W. Friedberg, and Margaret L. Houy moves the research forward by systematically defining primary care with the help of expert advisors—in eight different ways (based on the specialty of the providers and the scope of services provided), developing full and publicly available data specifications, and then testing them on nine commercial insurers who voluntarily compiled and submitted data.

Among the authors’ findings:

• It is possible to measure primary care spending using insurers’ financial information and based on expert consensus definitions of primary care translated into data specifications.
• Primary care spending as a percentage of total spending averaged 7.7% for the broadest definition in preferred provider organizations (PPO) plans and varied greatly across high-performing health insurers. It is not clear if this variance is real or a result of the insurers’ reporting methods.
• Differences in spending between narrow and broad definitions of primary care providers were small. More providers could be included in a primary care provider (PCP) definition, and consensus on a definition could be reached more easily without a significant overall increase in costs.
• Primary care spending as a percentage of total medical spending is influenced by population characteristics, including age and chronic conditions.
• Voluntary reporting was challenging to obtain. Working with health plans one-on-one is resource intensive for the plans and the researchers. Alternatives may include using state or regional APCDs or state legal/regulatory requirements for plans to submit data in a standardized manner.
• New payment models and delivery system structures will create new measurement challenges. For example, shared savings models may involve payment to an accountable care organization which is then distributed among primary care providers.

How can we apply this research? With our new report, policymakers interested in following Rhode Island and Oregon’s lead now have tested definitions and specifications for the measure that will promote comparison and benchmarking. Evaluators of provider payment reforms have an easily implemented measure to see if value-based payments are achieving their stated goals by promoting what is known to be more valuable—primary care.  And advocates for primary care have a new, practical tool and credible initial results for organizing and focusing their colleagues.

Consistent with its mission of helping leaders and decision makers improve population health, the Fund will continue to explore ways to promote the refinement and adoption of the primary care spending measure to help strengthen the United State’s primary care infrastructure. After all, in the late 1920s, the Milbank Memorial Fund was investing in some of the early precursors to today’s community health centers in New York City and other locations. Some obsessions just don’t go away.