In two recent reports, the Health Care Cost Institute (HCCI) appears to use an oversimplified analytic approach and draws overly broad conclusions about price variation and price growth variation. Their analysis does not address many of the key factors that can contribute to price variation and growth, while also ignoring broader trends in the health care market.
To start, there may be a number of valid reasons why price variation occurs across metropolitan areas; among them is that the cost of providing a service can vary among providers. For example, labor costs are among the largest factors in hospital cost growth, and labor costs vary geographically. Another example is a provider’s unique service mix and whether the organization is balancing very costly services.
Some of the differences within a market could be attributed to the insurance plan type. HMOs, which typically use a more restrictive network to leverage their purchasing power for lower negotiated prices, may pay lower prices for the same service compared to other plan types that have broader networks and less patient volume. HCCI’s analysis lumps all plan types together, despite stark differences between them. Naturally, one would expect to see variation in the prices paid by different plan types. In addition to this, certain insurer practices, such as excessive denials, may play a role as well. For example, if a health plan abuses prior authorization and medical necessity reviews to frequently deny reimbursement for medically necessary care, a provider may need to account for that in its pricing in order to break even.
In addition, the geographic boundaries HCCI uses are not designed to reflect health care markets, and may not bear any resemblance to patient preferences or the patterns in which people actually seek care. HCCI uses core-based statistical areas, or CBSAs, as the foundation of their work. These are created by the Census to measure population density, and use geographic markers like county lines. Price variation within a CBSA could be accounted for by differences in labor markets, state regulatory environments, and a number of other factors, but HCCI doesn’t control for any of these possible factors.
Patient acuity — that is, how sick or well a patient is — can also result in price differences. For inpatient services, HCCI focuses on C-section delivery and vaginal deliveries. But even within these services, patients can vary in terms of age and health status. And the clinical care provided can vary as well — complications during delivery can require additional care or result in a longer length of stay. HCCI appears not to have adjusted to control for these factors, which could result in oversimplified and flawed findings.
Significantly, the report also conspicuously fails to examine the role that insurer concentration plays in driving up health care prices. We know that 75% of health insurance markets are highly concentrated, that in 48% of markets one insurer had a combined market share of 50% or greater, and that the share of markets that are highly concentrated increased from 71% to 75% between 2014 and 2018.
The reality is that hospital price growth is at historic lows. According to the Centers for Medicare & Medicaid Services, price growth for hospital care services was just 2.4% in 2018, and non-price factors such as intensity of services and inpatient bed days grew slowly as well. These factors combined for historic low growth in hospital spending. More recent data from the U.S. Bureau of Labor Statistics shows hospital prices have consistently grown less than 3% per year over the last decade and have frequently grown by less than the average rate of inflation. In fact, even when excluding the artificially low rates paid to hospitals by Medicare and Medicaid, annual price growth has still been below 3% in recent years.
Aaron Wesolowski is AHA’s vice president of policy research, analytics and strategy.
 National Health Expenditure Data, 2018. CMS, Office of the Actuary, National Health Statistics Group. https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/NationalHealthExpendData/index.html