Drug Forecasts: Oops!...They Missed It Again

The boffins at the Centers for Medicare & Medicaid Services (CMS) just released their annual computations of National Health Expenditures (NHE), including the latest figures for prescription drugs.

Two notable highlights that I haven't seen mentioned elsewhere:

• CMS’ projections for the growth of drug spending have been much, much higher than actual growth rates—even for forecasts made only two years ago. As someone who makes his living from the crystal ball, I understand that sometimes you need to eat broken glass. But the disparity is quite striking and so far unexplained by CMS.
  
  
• The data confirm last year’s projection that Federal and state governments are becoming the primary payers of outpatient prescriptions. A major downside to this government takeover is the risk that budget pressures could encourage the government to dictate or mandate lower payment levels for pharmacies. Be careful what you wish for, health reform fans!

Here are links the original data sources in case you want to check my math:

• “Health Spending Growth At A Historic Low In 2008” (Health Affairs summary article)
• CMS’ National Health Expenditure Data (raw data)

REALITY VERSUS PROJECTIONS

The table below compares the actual growth rates in the latest NHE data with annual projections made by the Office of the Actuary in CMS during its three most recent forecasting attempts. The “n.a.” indicates CMS had actual growth rates for those years instead of projections.

As you can see, actual growth in drug spending has been far, far below CMS projections. For example, CMS' 2007 forecast called for 7.9% growth in drug spending in 2008. Actual growth turned out to be less than half of that rate (3.2%).

Perhaps CMS will enlighten us on the differences when they release the next set of NHE projections in February. CMS may pin the blame on the recession, but I think the primary factor was their underestimation of generic dispensing rates and the corresponding deflation in drug trend.

FYI, the big 3 pharmacy benefit managers (PBMs) retrospectively computed the 2008 drug trend to be around 3%. See Specialty Spending Soars (for now).

WHO PAID FOR DRUGS IN 2008?

The chart below shows how the payment source for outpatient prescription drugs (share of dollars) has changed since 1968. Public funds—Medicare Part D, Medicaid and State Children’s Health Insurance Program (SCHIP)—are crowding out both public and out-of-pocket payments. Public funds paid for 37% of total retail drug spending in 2008. The private health-insurance share peaked in 2001 at 50% and has been declining since then.


A FEW DATA MUSINGS FOR DRUG WONKS

NHE does NOT measure total U.S. spending on prescription drugs. Sales through outpatient retail channels are only about 75% of total pharmaceutical manufacturer sales. Unfortunately, there is no way to figure out spending for inpatient drugs in other NHE categories because spending is bundled with inpatient procedure fees. For example, some portion of "Hospital care" (and other NHE categories) also include an unknown amount of drug spending. CMS does report Medicare Part B outpatient spending, but only retrospectively and with a few years lag.

CMS’ estimates for total outpatient drug spending are consistently lower than estimates from private sources, while CMS’ estimates of growth rates tend to be higher than the growth rates reported by private sources. For instance, CMS estimates that outpatient drug spending was $234.1 billion in 2008 versus an estimate of $253.6 billion from NACDS. I believe that CMS underestimates prescription sales in food and mass channels. Why? Technically speaking, the current NHE Prescription Drug series is benchmarked to product line sales data from the 2002 Economic Census of Retail Trade (NAICS 44-45), which appears to underestimate prescription sales in non-pharmacy-dependent retailers. The 2007 economic census data should be available soon and will probably be incorporated into NHE by 2011. I expect an upward revision to the historical NHE data when that occurs.

---

P.S. The classic (circa 2000) Britney Spears reference above should tip you off that I have a 13 year-old daughter in my house!