Friday, April 11, 2014

Weighing Express Scripts’ Drug Trend Forecast Errors, Sovaldi Pricing, and PBMs’ Pricing Control

Yesterday, I focused on Untangling Express Scripts’ 2013 Drug Trend Report. Today, I explore the report’s unexpectedly large forecast errors for traditional drug trend, and what they mean for PBMs’ formulary management.

As I document below, 2013 drug trend was positive and higher than Express Scripts’ forecast for three major traditional drug therapy classes, none of which had a major generic launch. Express Scripts blamed the higher-than-expected trend on manufacturers’ drug pricing.

To coincide with its new drug trend report, Express Scripts launched an attack on drug prices, focusing on Gilead’s hepatitis C pill Solvaldi (of course). Dr. Steven Miller, Express Scripts’ chief medical officer, even implied that Sovaldi could “lose all its market share” when competitors launch. (Full quote below.) No surprise if you had read Express Scripts Plans a Specialty Drug Price War last December.

Given Express Scripts’ difficulty in projecting traditional drug trend for therapy classes without new generic launches, the tough talk may be premature. Read on and see if you agree.


Forecasting is very hard, especially about the future. Those of us who live by the crystal ball often eat broken glass. Nonetheless, the gap between Express Scripts’ projections and this year’s results surprised me.

In October 2013, Express Scripts released an updated version of its 2012 Drug Trend report. This update projected that 2013’s drug trend would decline by -1.0%. In reality, Express Scripts reported that 2013’s traditional drug trend increased by +2.4%.

The table below compares the October 2013 forecast with actual 2013 results, by therapy class.

[Click to Enlarge]

Three classes—high blood cholesterol, asthma, and depression—had unexpectedly large unit cost declines, due to generic launches. Drug trend was therefore lower than projected.

For three other therapy classes, however, drug trend was positive and higher than the forecast. Express Scripts explains these discrepancies by pointing the finger at manufacturers:
  • Diabetes: “Medications used to treat diabetes were the most expensive for the third year in a row.”
  • Pain: “Although generic medications continue to dominate the class, PMPY spend has not declined in accordance because some manufacturers of branded, tamper-resistant formulations have been successful in blocking generics to older, regular-release formulations with claims of superior safety for newer versions.“
  • Infections: “The primary trend driver in this class was an increase in unit cost that was in part the result of drug shortages for commonly used therapies including doxycycline and tetracycline.”

My question: If the biggest PBM was unable to control unit costs without generic launches, how will it control the costs of new specialty drugs?

Consider the diabetes category, which accounted for 13.3% of traditional drug spending in 2013.

Express Scripts estimates that diabetes drug costs could be cut in half by eliminating “pharmacy-related waste.” It defines waste as “using higher-priced medications when more affordable, clinically equivalent alternatives were available” and “using the most cost-effective and clinically appropriate pharmacies, including narrower networks of retail pharmacies, home delivery and specialty pharmacies like Accredo.”

Yet these apparently large savings opportunities, in the biggest category of traditional drug spend, could not be contained in 2013. As a result, 2013 diabetes trend was 14.4%, compared with the 8.9% forecast.


Think about the diabetes example as you read Dr. Miller’s comments in Bloomberg’s Express Scripts Raises Pressure on Gilead for Drug Price:
“Gilead could have a great year this year and lose all its market share a year from now,” Miller said. “The FDA has fast-tracked several other medications and we believe early 2015 is when there will be competitors in the marketplace.”
“The companies that will be second and third to the market here will have to play catch up,” Miller said. “We could shift the market share as soon as a competitor comes out. We need to start a national debate on fairness in drug pricing.”
Wow, that’s pretty tough talk. Per Express Scripts and the Inevitability of Formulary Exclusion, Express Scripts has shown that it is willing to exclude products from its national formulary. Is Express Scripts really ready to lower the boom on Gilead’s highly effective new therapy? And if they do, will it work?


  1. Adam - great post. Do you think ESRX's exclusion of many of Novo's diabetes products in favor of Lilly going to be able to put a dent in their diabetes trend? It's interesting because share of medications like Victoza, while taking a step down early in the year, has steadily rebounded, and I don't know how that works out given they were just kicked out of one of the largest formularies in the country. thoughts?

  2. Adam, the point that you didn’t talk about is that while the PBM’s can talk tough and roll out closed formularies du jour, if Plan sponsors aren’t ready to take the hard steps that drive member complaints the PBM’s are a paper tiger. I haven’t looked in a while but the percent of clients that actually take these national formularies is really small and they are really more of a ‘stake in the ground’ rather than an iron hammer.

  3. I presume that's what they are hoping, although the forecasts don't show much decline.

    Here are Express Scripts' Diabetes drug trend forecasts (source):
    2014: +11.2%
    2015: +12.2%
    2016: +10.5%

    Unfortunately, the forecast doesn't separate utilization vs. unit costs. Stay tuned.

  4. I not sure even ESI van convince physicians to with hold therapy on a drug that has superior clinical outcomes. Where is patient benefit in this discussion and long term reduction of other medical costs? I assume it is excluded from ESI cost projections because they have no line of sight into medical expenditures.

  5. I wonder what drug Steve Miller would want, for himself or a family member, if they had a serious Hep C infection. It's great to rail against drug costs, when you have no stake in medical/hospital costs.

  6. Items 2 (high cholesterol) and 4(anti ulcer), raise an eye brow and (IMO) speak to either an addiction to rebates or a lack of straight talk by PBMs to their clients around the trade-off between choice and cost. With the multiple generic options available, I can't imagine why any employer (or health plan)--when presented with the facts--would not choose to PA a few select branded options in key indications, and NDC block the rest of the brands. As for #1 (diabetes--or to be more direct, we can probably point a finger directly at analog insulins--or maybe any insulin, here), I can't blame the model or forecasters, as the tools are dependent upon historic data to tease out trend and seasonality. When manufacturers decide to steepen the rate of price increases (relative to historic data), models break. I wonder if we can chalk up the tend in diabetes spend as an externality of the "donut hole deal" struck by manufacturers. Someone had to subsidize the value of the additional rebate. By the way, what ever happened to biosimilar/"generic" glargine insulin?

  7. +1... see my thoughts above! At the end of the day rebate/admin fee addiction, result in some/many/all (take your pick depending on how jaded you are or what kind of a day you are having) PBMs having a low level of 'motivation' to aggressively sell and educate clients on formularies that approach things from the perspective of lowest net cost. Exhibit 1: on ESIs national formulary where is Nexium placed? Why would that be given the existence of generic/OTC omeprazole and lansoprazole??

  8. As long as any PBM plays the spread game - it's hard to forecast against the spread versus forecasting against actual acquisition costs. If Dr. Miller wants to debate fairness in drug pricing let's throw in a debate on spread pricing from the PBMs.

  9. Solvadi or Sovaldi? Or did I miss the joke?

  10. Oops. Error corrected. Editor fired.

  11. Any indication what the price increases are net of additional rebates due to exclusions, price caps, etc...? In a class like diabetes, you'd HOPE that the unit price increases would be offset to some extent by additional rebates in the class.

  12. I suspect that Gilead's thinking is: we better have a great year this year "because" we are going to lose our market share next year."

  13. Adam,

    Why do you think there's such a huge difference between ESRX's forecasts (e.g. traditional drugs +2% v 2013) and their actual reported results?

    I just glanced thru ESRX's Q1 earnings, and they reported a utilization of -9% adjusted for UNH yr over yr. Last quarter was the same story, but not as bad.

    It doesn't even seem possible? (It can't be due to last minute client defections (right?) considering that the selling season was over months ago...)