Monday, October 27, 2008

Prescriptions and the Economy: A Contrarian View

Last week, the New York Times ran a story on the impact of the economy on prescriptions. (See In Sour Economy, Some Scale Back on Medications.) The article provided many compelling personal stories to illustrate the following macro data trend:

“Through August of this year, the number of all prescriptions dispensed in the United States was lower than in the first eight months of last year, according to a recent analysis of data from IMS Health, a research firm that tracks prescriptions.”

But what if the data are wrong?

I’m not suggesting that the prescription market is booming. However, I have some concerns about the ability of IMS Health to measure a changing retail marketplace with incomplete data, so perhaps the news is not quite as bad as reported.

Here's why. The top six dispensing pharmacies – CVS Caremark (CVS), Walgreens (WAG), Medco Health Solutions (MHS), Rite-Aid (RAD), Wal-Mart (WMT), and Express Scripts (ESRX) – now account for more than half of all retail prescriptions. Yet two of these six do not report sales data to IMS Health.

  • As far as I know, Wal-Mart does not sell its prescription records to any third-party data provider.
  • At least one of the major mail order pharmacies apparently stopped selling its data to IMS Health in January 2008.

These gaps would not bother me if market share was stable at the top six players because estimation would be a straightforward “fill in the missing number” exercise. But we know that’s not the case, especially when it comes to Wal-Mart.

Wal-Mart does not release any specific data on its pharmacy department, but anecdotal evidence suggests that the company has picked up share in selected regional markets. What do you think is going to happen to Walgreens market share in Peoria, IL, as the Wal-Mart/Caterpillar partnership gets going? Look at a map of Walgreens stores in the Peoria, IL area. Ouch.

The Times gave some lip-service to other explanations for the drop in scripts when briefly noting “…safety concerns over some previously popular drugs and the transition of some prescription medications to over-the-counter sales…” But the main point of the article was to link the economy to the slowdown in script growth.

I suppose it would be unfriendly to mention that most consumer-related metrics of the economy – retail sales, new unemployment claims, etc. – did not start to show signs of recession until this year. Look at the chart in the NYT article again, which shows the monthly year-over-year decline starting in early 2007. Well, this won't be the first time that I've highlighted partisan reporting about the industry by the New York Times. (See Sloppy reporting about Wal-Mart from way back in 2006.)

I’m sure the folks at IMS Health try hard to fill in the gaps, but there’s simply no way for them to do anything but guesstimate when they lack actual data for a significant and fast-changing part of the pharmacy market.


  1. Agreed that the NY Times chose to look at the prescription data through only one lense. The IMS data actually shows a slowing that began in 2005, took a big break in 2006 as Part D was introduced, and then started to decline again in 2H07. One cannot ignore the impact of poor outcomes research and the negative press that goes with it. Or the impact of higher out of pocket costs for many drugs.

    IMS may find it particularly difficult to keep track of scripts following the introduction of $4 generic programs at many retailers. We believe that some of these pharmacies are filling prescriptions as cash rather than adjudicating them through a patient's pharmacy plan. That makes it hard for IMS to compare data coming from the PBM's (for those that give IMS data) with that provided by retailers. A number of industry observers and participants have had concerns about the accuracy of the IMS data, yet the stock market and the press seem to treat it as gospel.

  2. Adam, I agree with your view and the first comment. If I go to Wal-Mart and pay cash for a $4 Rx, I don't want it submitted to my insurance. I think the number of scripts is likely under-counted.

  3. Great insight -- how come no one else is talking about this??

    As someone who has personally worked with IMS data, I am always surprised how people assume that the data are perfectly accurate. It's not like scanner data from the supermarket -- these are a combination of surveys and electronic data collection. Big gaps are "estimated" (as you point out), sometime imperfectly.

    Nature abhors a vacuum, so until the "perfect data" comes around (never), then IMS has got a good cash cow from pharma companies.

  4. Thanks! About time someone pointed out the flaws with IMS data. Filling in the blanks is guesswork at best. Clearly, something is missing and they refuse to own up. Good job, Adam!


  5. I am an IMS employee so I have some bias but the comment section has taken this argument over the edge. I agree that NYT has focused on linking the decline in Rx in an inappropriately direct way to the economy while ignoring drug safety issues. I do not agree that IMS data is total guesswork. We employee rigorous statistical procedures to estimate the market. I can confirm that Walmart does not report Rx but this does not mean that trends at a supplier (of data) as large as CVS do not indicate market trends that can be forecast with some degree of precision to the market. To say it is total guesswork and write off the insights that IMS data can provide is unfair and represents a an exaggerated skepticism.

  6. I use IMS and track our companys Rx data and it has always been within 1% of actual sales to wholesalers. The Rx data provided by vendors is very accurate and used by wallstreet analyst to track companys value