Tuesday, March 19, 2013

Last week, Altarum released our Health Sector Economic Indicators (HSEI)SM for March 2013 the latest monthly information on health care spending, prices, utilization, and employment. We headlined that health care price inflation had dropped to 1.5 percent in January 2013, one of the smallest increases on record, and the second consecutive month under 2 percent.[i] We thought this would be an opportune time to elaborate on the data presented in our monthly price brief. We begin with some background on our health care price index (HCPI) and show how it differs from the medical care component of the consumer price index (CPI). Next, we look at health care price inflation from January 1990 through January 2013, compared to economy-wide inflation. Finally, we analyze the past five years and contrast price growth for hospital services, physician services, and prescription drugs.

The Health Care Price Index

We model the HCPI after the personal health care price index developed by the Centers for Medicare and Medicaid Services (CMS). The CMS index is a weighted average of Bureau of Labor Statistics (BLS) price indexes for various categories of health care goods (prescription drugs, durable medical equipment, etc.) and services (e.g., hospitals, physicians, nursing home, home health etc.), in which the weights are proportional to the total amounts spent. The HCPI is essentially a monthly version of the CMS index using our monthly total spending estimates as weights. It is important to distinguish the HCPI from the more familiar medical component of the CPI, which uses out-of-pocket spending to weight the component price indexes.[ii] This CPI weighting scheme is not representative of health spending as a whole (since so much health spending is not out-of-pocket) and, as a result, the HCPI is a better measure of overall health care prices.[iii]

Health Care Prices, Growth, and Economy-Wide Inflation: January 1990–January 2013

The chart below, from Exhibit 5 in our March price brief, plots health care price inflation versus economy-wide inflation, using the gross domestic product (GDP) deflator, from January 1990 through January 2013. Let’s begin by focusing on the gap between health care and economywide inflation. This gap is important because health care price growth in excess of economywide inflation is a key contributor to health spending growth in excess of GDP. It is a more important statistic than the level of health care price inflation.  

The general pattern is for a positive gap, i.e., excess health care price growth, except for three periods in which the gap oscillates around zero. The first such period is the middle of the managed care era (1996–1997) when the well-known temporary bend in the health care cost curve occurred. The second period was from 2005 to 2007, the three years leading up to the most recent recession, and a period we identified as representing a pre-recession bend in the cost curve. The final period, from 2011 to present, is characterized by very low growth in health spending.  

As would be expected, periods of rising and falling health care price inflation align with rising and falling economywide price inflation. It is no accident that the only periods in which health care price inflation dropped below 2 percent are periods in which economywide price inflation was also below 2 percent. In this light, the sub-2 percent growth in health care prices in recent months is at least partially explained by sub-2 percent economywide inflation. There are, however, notable breaks in the relationship between these two measures. Consider 2009 and 2010, during which health care and economywide prices diverged dramatically. The large gap clearly distinguishes the zero gap endpoints characterizing the pre- and post-recession periods. Why did health care price inflation peak at more than 3 percent while economywide inflation dropped to zero in late 2009? For a possible answer, we turn to the most influential components of the HCPI.

Hospital, Physician, and Prescription Drug Price Growth: 2008 to Present

The three sectors with the largest weights in the HCPI are: hospital services (31 percent), physician and clinical services (21 percent), and prescription drugs (10 percent). As shown in the Addendum charts in this blog, for the 2009-2010 period, hospital price growth most closely matches that of the HCPI, first rising and then falling. Note the downward cliff in October 2010, which is the month when Medicare and many private payers change their hospital payment rates.

BLS provides separate price indexes for hospitals by payer type (Medicare, Medicaid, Other). As shown in the chart below, the 2009-2010 patterns of price growth for Medicare and Other (largely private insurers) match that of the HCPI and tend to dominate the very different Medicaid pattern. We conjecture that Medicare hospital payment policies are a significant contributor to this pattern, with private payers generally following suit. This chart also shows that, for the past three years, Medicare hospital prices have grown more slowly than private insurer hospital prices (the “Other” category). This could be evidence of cost shifting but could also simply represent the government’s greater influence over prices.[iv]

The Addendum charts also show that physician prices, and especially prescription drug prices, contributed to the jump in the HCPI in 2009. Medicare payment policies added to the rise in physician price inflation but we don’t have a good explanation for the rise in prescription drug price inflation. It is worth noting that BLS confronts challenges in accurately measuring the growth in prescription drug prices, including the effects of rebates (which are generally not captured, since they don’t affect the transaction price), and the shift to generics, in which timing is an issue. Prescription drug prices, as measured, show the highest inflation rate, running at close to 4 percent for much of the past five years. Physician prices, on the other hand, show anemic growth—rarely exceeding 2.5 percent—often well below economywide inflation, and currently at a miniscule 0.6 percent. Hospital price growth has been more resilient, generally exceeding economywide inflation.

From January 2011 through November 2012, health care price inflation was steady at around 2 percent. In the past two months, it has averaged only 1.6 percent. This recent drop can be traced to lower price inflation for hospitals (a half percentage point drop) and prescription drugs (a full percentage point drop). Will this sub-2 percent HCPI growth continue? We think there is a good chance it will. Discussions with BLS suggest that the slower growth in prescription drug prices is tied to some popular brand name drugs going off patent, so this slower growth has the potential to persist. Medicare payments for hospitals will continue to be stingy under the provisions of the Affordable Care Act, and physician prices show no signs of increasing. Tune in next month for a new installment of our price brief and see whether February 2013 becomes the third consecutive month of sub-2 percent health care price inflation.

Addendum: Hospital, Physician, and Prescription Drug Price Growth


[1] Our inflation measure compares a given month with the same month for the previous year. Thus we show that health care prices in January 2013 are 1.5 percent higher than they were in January 2012.

 [2] “Out-of-pocket” includes spending financed by insurance premiums, including Medicare Part B.

 [4] A difference between Medicare and private payer price growth does not, by itself, denote cost shifting. Hospitals could be absorbing the slow growth in Medicare prices through cost cutting and/or lower margins. Frakt gives an in-depth discussion.

All postings to the Health Policy Forum (whether from employees or those outside the Institute) represent the views of the individual authors and/or organizations and do not necessarily represent the position, interests, strategy, or opinions of Altarum Institute. Altarum is a nonprofit, nonpartisan organization. No posting should be considered an endorsement by Altarum of individual candidates, political parties, opinions or policy positions.