Explaining regional variation in private health care spending

Health Status and Hospital Prices Key to Regional Variation in Private Spending

Research Brief by Chapin White
National Institute for Health Care Reform, February 15, 2012

Differences in health status explain much of the regional variation in spending for privately insured people, but differences in provider prices — especially for hospital care — also play a key role, according to a new study by the Center for Studying Health System Change (HSC) for the nonpartisan, nonprofit National Institute for Health Care Reform (NIHCR).

Based on claims data for 218,000 active and retired nonelderly unionized autoworkers and their dependents, the study found that health spending per enrollee in 2009 varied widely across 19 communities with large concentrations of autoworkers, from a low of $4,500 in Buffalo, N.Y., to a high of $9,000 in Lake County, Ill.  The autoworkers’ health benefits are essentially uniform nationally, so spending differences do not reflect benefit differences.

Differences in service quantities accounted for two-thirds of the overall spending variation, while differences in prices accounted for one-third, according to the study.

On the quantity side, differences in health status and other demographic factors explained most, but not all, of the variation in quantity. About 18 percent of the total variation in spending was a result of unexplained differences in service quantities. On the price side, the cost of doing business explained very little of the price differences, with almost all of the differences in prices unexplained, the study found.

From the Research Brief:

Sources of health care spending variation across autoworker communities, 2009

Differences in spending due to quantities
37% - Differences in health status
10% - Differences in age and sex
18% - Differences in quantities (excess)

Differences in spending due to prices
2% - Differences in providers' cost of doing business
33% - Differences in prices (excess)

Comparing Private Prices to Medicare

Physician office visits. The prices paid by the autoworker plan for physician office visits, on average, are only 3 percent higher than what Medicare would have paid for the same services.

Hospital inpatient care. The prices for inpatient hospital care paid by the autoworker plan are, on average, 55 percent higher than what Medicare would pay, and the price gap varies widely across communities.

Hospital emergency department care. The prices paid by the autoworker plan for hospital emergency department care are, on average, more than double the Medicare price, and the price gap varies even more widely across communities than for inpatient care.

Research Brief
PDF version


By Don McCanne, MD

Policy wonks likely will download this Research Brief since it will prove to be a landmark study explaining the wide variation in spending amongst privately insured patients - very useful information when designing policies to improve our health care financing.

We already have several landmark studies demonstrating the wide variations in quantity of services and in pricing of those services - both factors in determining the variations in health care spending.

What has not been so clear is whether the quantities vary based on the health status of the patients or based simply on variations in intensity amongst populations of comparable health (i.e., providing excess services to drive up revenues, though deficient services when system capacity is inadequate is an important but separate concern).

What also has not been so clear is whether price differences are primarily due to differences in costs of health system resources, or simply due to charging higher prices where private insurer market control is weaker.

This study provides very convincing data to explain the differences. Look at the percentage differences listed above.

Health status (37%) and age and sex (10%) explain almost half (47%) of the differences in spending. These are entirely appropriate differences and do not require policy intervention. The 18% that is not otherwise explained may well be excess quantity of services that perhaps could be moderated by judicious application of health policy.

But let's not kid ourselves. Those who suggest that, based on the Dartmouth studies, close to 100% of the difference in quantity of services is excessive tend to support policies that would come as close as possible to eliminating the differences. Since this study shows that the excess quantity is only about 18% of the difference in spending (only slightly over one-fourth of the increased quantity in higher spending areas), it would be very difficult to design policies that would tease out this modest amount while leaving in place the other services that are clearly beneficial. That doesn't mean that efforts shouldn't be made to do so, but it does mean that great care must be taken to be certain that people still receive the services that they should have - a much more important priority than budget trimming.

Excessive prices (33%) appear to account for most of the differences in prices for services since differences in the providers' costs of doing business account for a mere 2% of spending differences. Most of this excess pricing is found in hospitals and in their emergency departments. This finding screams out for a better method of controlling prices. All other wealthier nations use some form of government pricing to eliminate these excesses. We should too.

Uwe Reinhardt has provided convincing evidence that an all-payer system, such as the hospital price-setting system in Maryland, would capture much of this excess in pricing. That is true, though it would be more difficult for an all-payer system to address the problem of excess quantity.

Our objection to all-payer systems is that they address only this one issue and leave out the many other advantages of a single payer national health program. When we have a solution that would fix so many problems at once let's go with that.