Thursday, July 21, 2016

Latest CMS Forecast Shows Big Drug Spending Growth Through 2025

Last week, the Centers for Medicare & Medicaid Services (CMS) released its latest forecasts for national health expenditures. See the Health Affairs article: National Health Expenditure Projections, 2015–25: Economy, Prices, And Aging Expected To Shape Spending And Enrollment. (Available for free to subscribers and to others for purchase.)

The boffins at CMS have again raised their forecasts for spending on outpatient prescription drugs. Their new projections show that through 2025, drug spending will grow slightly faster than will total U.S. healthcare spending. By 2025, the U.S. is expected to spend about $615 billion annually on outpatient prescription drugs.

Despite this growth, outpatient prescription drugs will still account for only about one of every nine U.S. healthcare dollars. By comparison, spending in 2025 on hospitals and physicians is projected to be $3.2 trillion, or more than five times the spending on outpatient pharmaceuticals.

Below, I delve into the latest numbers, explore how the new forecasts compare with previous prognostications, and highlight the variables driving CMS’s model.


Every year, the Office of the Actuary at CMS publishes annual projections for U.S. National Health Expenditures (NHE). These projections include spending on prescription drugs sold through outpatient retail, mail, and specialty pharmacies.

U.S. drug spending in the NHE is roughly equivalent to total outpatient pharmacies’ prescription revenues minus manufacturer rebates to third-party payers. Thus, the CMS-reported figures differ from pharmacies’ prescription revenues, manufacturers’ revenues, and the “invoice price spending” data reported by IMS Health. Note that the patient’s cost share (copayment or coinsurance) is included in NHE, because patients are one of the payer types tracked by the CMS.

The NHE data are also not comparable to IMS Health’s “net price spending,” which attempts to measure net revenues received by pharmaceutical manufacturers. As I understand the data sets, the NHE figures should be roughly equivalent to:
  • IMS’s “invoice price spending” for outpatient dispensing channels
  • minus off-invoice discounts and rebates provided by wholesalers (but not manufacturers) to pharmacies
  • plus channel margins earned by pharmacies
  • minus the share of manufacturers' rebates that are passed through to third-party payers
Note that the NHE drug expenditure figures do not measure total U.S. spending on prescription drugs. Inpatient prescription drug spending within hospitals and nearly all provider-administered outpatient drugs are reported in other categories. This additional spending accounts for an estimated 3 percent to 4 percent of national health expenditures. (source)

Make sense? For more, see my comments on “Drug Disambiguation Details” in Key Insights on Drug Prices and Manufacturer Rebates from the New 2015 IMS Report.

FYI, you can go the full wonk with the complete historical and projected data files on the CMS National Health Expenditure page.


CMS estimates that net outpatient prescription drug spending increased to $321.9 billion in 2015, up 8.1% compared with the 2014 figure. Annual U.S. prescription drug expenditures are expected to grow by $292.6 billion (+91%), to $614.5 billion in 2025.

The chart below presents our analysis of the most recent forecasts for annual growth in total national healthcare expenditures and outpatient prescription drugs. From 2016 to 2025, total health spending is projected to grow at an average rate of 5.9%, while prescription drug spending is projected to grow at an average rate of 6.7%.

[Click to Enlarge]

In Health Affairs, CMS provided the following commentary on the factors behind its forecast:
“In 2015 there were forty-five new drug approvals in the United States, up from forty-one in 2014 and twenty-seven in 2013. Many of these drugs have small target patient populations. Thus, the impact of new drugs approved in 2015 is likely to be smaller than in the previous two years—when fewer new drugs were approved, but several of them were intended for wide use.

Over the projection period, the impact on spending growth from newly approved drugs each year is expected to be lower than that observed in 2014 and 2015. The number of new drugs approved annually is anticipated to decrease. Moreover, a few of these new drugs are expected to be biosimilars, which are typically priced lower than the originator drug.”
In a supplementary forecast summary, CMS notes that the drug spending growth rate is projected to peak in 2018. That’s when relatively fewer brand-name drugs are expected to lose patent protection.


Over the past few years, CMS has consistently raised its projections for drug spending. Here’s a look at the three most recent projections. (CMS hates when I do things like this.) As you can see, projected drug spending is higher in more recent forecast years.

[Click to Enlarge]

Here’s what happened.

CMS told me that its team did a comprehensive revision to all of the services and source of funds. Consequently, the total healthcare spending data were restated all the way back to 1960 and the drug data back to 1992.

This revision lowered historical prescription drug spending levels. Last year’s projections, for example, pegged drug spending in 2014 at $305.1 billion, while this year’s figure for 2014 is $7.4 billion less, at $297.7 billion. CMS also increased its projected growth rates, most notably for 2016 and 2018.

To understand how CMS’s prescription drug model works, see page 25 of this wonktastic methodology guide. Briefly, CMS projects Real aggregate drug spending per capita using the following independent variables:
  • Real disposable personal income (3-year moving average) (+)
  • Relative drug price * Share paid out-of-pocket (3-year moving average) (−)
  • New drug introductions (+)
  • Generic dispensing rate (−)
The signs in parentheses show the effect on the dependent variable. But as CMS admits:
“Patterns of growth over the most recent 10 to 15 years of data are difficult to explain, as the effects of several different factors must be disentangled.”
And in its review of forecast accuracy, CMS adds: “The projections of drug spending growth have, on average, overestimated historical spending…”

Yes, it’s true. If you live by the crystal ball, then you must be prepared to eat broken glass.

Hopefully, that lesson will make you hungry for my follow-up article in which I examine who will be paying for our prescription drugs over the next 10 years. Crunch!

No comments:

Post a Comment