Rand releases updated study of hospital payment variation in commercial market

• A new study published by the RAND Corporation finds that employers and private insurers across 49 states and District of Columbia pay, on average, 247% of what Medicare pays for hospital services. The report finds greater variance in the prices between hospitals within one health system than the prices between different health systems.
• The report focuses on services provided by hospitals in both the inpatient and outpatient setting. The updated study also includes physician payments for services provided in hospitals.
• The study evaluated commercial spending from 2016 to 2018 from 3,112 hospitals in 49 states and the District of Columbia. Maryland was excluded given its all-payer rate setting system.

On Sept. 18, Rand Corporation released the third iteration of its evaluation of healthcare prices paid by private health plans.

The report focuses on services provided by hospitals in both the inpatient and outpatient setting. In addition to expanding the number of states covered in the analysis, the updated study also includes physician payments for services provided in hospitals.

The study evaluated commercial spending from 2016 to 2018, accounting for $33.8 billion in spending from 3,112 hospitals in 49 states and the District of Columbia. Maryland was excluded given its all-payer rate setting system. While $33.8 billion over three years is a lot of healthcare spending, it’s helpful to put this in perspective. According to CMS’s Office of the Actuary’s most recent National Health Expenditure data, commercial insurers spent $1.366 trillion on hospital services from 2016 to 2018 alone. So even before you add in the physician data, the study looks at less than 2.5% of commercial spending nationally over the three year period.  

Similar to prior years, the report reprices commercial claims using Medicare payment systems in an attempt to study the variance in hospital prices nationally, across states and within healthcare systems. It finds the following:

1. Commercial plans pay more than Medicare: According to the report: “In 2018, relative prices (includes FFS payment adjustments for DSH and IME) for hospital inpatient services averaged 231 percent of Medicare and 267 percent of Medicare for hospital outpatient services.” The report finds that the difference between commercial payments and Medicare payments has increased from 2016 and 2018. There is significant variance across states, across systems and even within systems. The report finds that across systems, price variance from the 25^th to 75^th percentile is 20% while the variance between hospitals within a system over the same interquartile range is 32%. Keep that last nugget about intra-system variation in your back pocket … we’re going to revisit it in a bit.

Commentary: The report uses Medicare rates to compare prices paid to different hospitals. While in one section the explains that it is merely using Medicare rates as an attempt to create a basis for comparison. However, in another, it quotes a MedPAC report stating that hospital prices are, “set with the overarching goal of compensating providers fairly based on their costs of doing business and the services they provide.“ There are a lot of reasons why Medicare prices aren’t an appropriate benchmark to compare variance in pricing between health systems.

First, Medicare allowable cost does not include all of a hospital’s cost to provide services, so suggesting that Medicare rates are intended to cover a hospital’s actual costs to deliver care is not accurate.

Second, MedPAC’s most recent analysis of payments finds that even relatively efficient Prospective Payment System (PPS) hospitals’ margin on inpatient services is -2%. The average hospital loses 9.3% on inpatient services. So long term, paying Medicare rates for the commercially insured isn’t sustainable. Which brings us to the cost shift.

2. No evidence of a cost shift: The report analyzes case mix adjusted discharges for Medicare and Medicaid patients and compares them to commercial prices. It finds no evidence of a relationship between the two. The report further states that, “Even if cost-shifting were a reasonable response to shortcomings in government payments, many employers would not consider it to be their responsibility to make up the shortfall when shopping for health care or for other goods and services.”

Commentary: Many, if not all, businesses outside of acute care hospitals have the ability to exit customer segments that aren’t profitable. However, hospitals can’t do that for both mission and legal reasons (EMTALA). And when governmental purchasers pay less than the cost to provide care some level of cross-subsidization is occurring. So you can run all the regressions you want. If governmental purchasers are a large percentage of a hospital’s payer mix, some cost shifting is necessary. My evidence? I submit to you the 120 rural hospitals that have closed since 2010. And there are another 453 hospitals considered vulnerable. You might question whether or not these facilities had a sufficient population in the catchment area to support them, or posit that they’re mostly in states that have refused to expand Medicaid because that’s what’s lead to these closures. That may have been true. But we’re starting to see more urban safety net hospitals struggle and close, so the tread on that explanation is starting to wear thin.

A recent Kaiser Health News article highlights the challenges facing safety-net providers in densely populated urban areas. And I’ll note that all of the urban areas are in Medicaid expansion states. This problem is only going to grow as more traditionally hospital-based procedures (e.g., joint replacements) can be done in freestanding settings. If we as a society, which includes employers, don’t support safety-net hospitals, we will be standing by idly as disparities in access to basic healthcare services between the haves and have-nots increase.

3. Insinuating consolidation is a culprit: The report points to several studies that suggest that consolidation between providers is a primary driver of price variation. However, it does not appear from the report that the authors attempted to use the claims data they have at their fingertips to understand the role consolidation may have played in the price variation they identify.  

Commentary: A couple of random thoughts on this one. It would be interesting for Rand to use its claims data to construct provider a  Herfindahl–Hirschman Indexes (HHIs) for the plans/markets it has access to and see if it is consolidation that’s driving the pricing variation. However, they may not have enough data to do so.  It’s also interesting to note that price variation within systems is greater than across systems or states. Granted this aspect of the analysis only looks at 40 systems, so it’s a small pool. But if the researchers could divide the systems between those who are predominately in one state versus multi-state systems, it might be interesting. It’s hard to tell from the whisker plot in the chart. But if there’s as much variation between prices in single state systems as multi-state systems, it would weaken the assertion that it’s market power that drives differences in pricing.