Tuesday, June 24, 2008

Accenture's Track-and-Trace Straw Man

A new report from Accenture attempts to assess the financial costs of implementing a serialized track-and-trace system for the U.S. pharmacy supply chain. I commend NACDS and NCPA for correctly highlighting an important barrier to authentication of serialized products at the point of dispensing.

Full report: Current Status of Safety of the U.S. Prescription Drug Distribution System

But even if we assume that Accenture did a credible and impartial job building their pharmacy cost models, the estimates in this report reflect an extreme situation that no one is seriously advocating. In other words, Accenture’s calculations may not be technically wrong (in the way that the recent AMP study was wrong), but its conclusions are highly misleading given actual proposals and practices regarding supply chain security. Thus, you should think of this report as the inflated, upper bound, "worst case" costs of a track-and-trace thought experiment. Caveat lector.

BAR CODES AND DOLLARS AND DATA! OH MY!

Accenture created detailed pharmacy cost models for a 100% compliant “complete track and trace system” that is federally mandated to be implemented at the unique unit-level for all products everywhere at the same time.

I can’t evaluate the assumptions hidden in the pharmacy costs models because Appendix F, a detailed Excel spreadsheet with all of the calculations, was not made publicly available. But my math (see footnote) shows that first-year implementation costs to be +/- 3% of revenues for the following four pharmacy archetypes defined by Accenture:

  • Large chain pharmacy (total annual retail sales = $18.8 billion): 2.6% of revenue
  • Medium chain pharmacy (total annual retail sales = $450 million): 3.4% of revenue
  • Small chain pharmacy (total annual retail sales = $60 million): 2.4% of revenue
  • Independent pharmacy (total annual retail sales = $6.5 million): 3.1% of revenue

Note that these figures are much higher than Accenture's previous cost estimates. At the 2007 NACDS/HDMA RFID/Track & Trace Health Care Industry Adoption Summit (yes, that’s really the title), Accenture estimated that one-time implementation costs for pharmacies would be only 0.5% to 1.5% of revenue. (See page 9 of Accenture’s State of the Industry presentation from November 13, 2007.)

I think that some of the newly-discovered costs come from additional pharmacy labor for scanning serialized products in the newer estimates as well as the assumption of complete year one implementation. But given the surprisingly large inflationary bump, I also surmise that NACDS and NCPA got their money's worth from Accenture.

SOMEWHERE OVER THE RAINBOW

Careful readers will note that Accenture’s “complete track and trace system” is much more comprehensive than anything being seriously proposed or considered right now. In my opinion, Accenture has defined a “track and trace” model in an unrealistic manner and then proceeded to explain that such a model is cost prohibitive.

In other words, the report is built upon the Straw Man logical fallacy, which goes as follows:

  1. Person A has position X.
  2. Person B presents position Y, which is a distorted version of X.
  3. Person B attacks position Y.
  4. Person B concludes that X is false/incorrect/flawed.

This fallacy plays out in the interplay between the actual study and the press release about the study:

  1. Legislators, regulators, and many industry participants want to implement serialized e-pedigree with national standards in the pharmaceutical industry.
  2. Accenture has modeled the pharmacy costs of implementing a 100% compliant “complete track and trace system” that is federally mandated to be implemented at the unique unit-level for all products everywhere at the same time, i.e., the Straw Man version of current proposals.
  3. Accenture's pharmacy cost models show that the Straw Man track-and-trace system would be ludicrously expensive to implement.
  4. Therefore, there is no need for serialized e-pedigree with national standards in the pharmaceutical industry, given the many other steps already taken to secure the pharmacy supply chain. Q.E.D.

I’ll mention just four important ways that the actual implementation of “track and trace” will likely differ from the models estimated in Accenture’s report:

To be fair, Accenture's language is fairly neutral regarding the implications of the study, whereas the NACDS/NCPA press release mistakenly links the "results" to a specific legislative proposal. Hence my view that this report represents unrealistic, upper bound costs of a track-and-trace thought experiment.

I'LL GET YOU, MY PRETTY...AND YOUR LITTLE BLOG, TOO!

I am very sympathetic to the motivation behind this study. A truly closed-loop, interoperable track-and-trace security solution based on serialization requires a massive infrastructure upgrade at the 150,000+ points of pharmacy dispensing in the U.S. I have persistently criticized the RFID hypesters who ignore this practical aspect of the pharmacy supply chain. I keep reminding technology firms, legislators, and regulators that pharmacies do not want to absorb the costs of reading pedigree or serialized data. (See Pharmacists Haggle over Pedigree Costs, among other posts.)

The report correctly highlights the fact that counterfeit drugs are still extremely rare in the United States and summarizes some of the business changes that have occurred to make the supply more secure. However, I do not believe that the cost estimates in this report make an accurate contribution to the debate over supply chain security.

On the other hand, the report should spark some good discussion at this November’s NACDS/HDMA RFID Track & Trace Healthcare Summit. Notice that the word “adoption” has now been discreetly dropped from the event title!

Read the report, make up your own mind, and let me know what you think. Maybe you'll get an honorary degree of Th.D. (Doctor of Thinkology**), too!

----

* Total Corporate Implementation Costs as % of Revenue = [(Cost Per Distribution Facility * No. of Distribution Facilities) + (Cost per Pharmacy Store * No. of Pharmacies) + Pharmacy Data Center Costs] / Annual Revenues

Example: Medium Pharmacy Chain = [($2,752,771 * 1) + ($103,939 * 100) + $2,288,265)] / $450,000 = 3.4%

** Yes, that's the honorary degree awarded by the Wizard of Oz to the scarecrow.

5 comments:

  1. AnonymousJune 24, 2008

    Adam-Another brilliant analysis! Funny, too. I'm impressed that you continue to publish so much material on a regular basis. A fan

    ReplyDelete
  2. AnonymousJune 24, 2008

    Oh the cost, the cost of it all. Yet, there was a note in the news today:

    CVS tests shrink-management system
    CVS is one of several retailers testing a new shrink-management program. Checkpoint Systems' Evolve system features RFID and software-defined radio. Drug Store News

    Seems they can invest in RFID if they want! Hmmm..Does this get me a Doctor of Thinkology?
    -A

    ReplyDelete
  3. AnonymousJune 25, 2008

    Oh Wizard, which path leads to OZ. Europe seeks point of dispense authentication. WHO has good distribution practices fostering supply chain partners to maintain lot control. PDMA pedigree still has not been instituted. ePrescriptions now take center stage. Bar codes used to prevent hopital medication errors. Can linear programmers plot time to implement, # of safety concerns addressed and total costs to find the magic portal.

    ReplyDelete
  4. AnonymousJune 25, 2008

    Great job in highlighting the change in Accenture's cost projections from November, 2007 to June 2008, and providing the link to back it up. How curious that implementation costs were projected to go up during a time when RFID tag and reader prices are coming down.

    The biggest flaw I found in the report is that it projects no benefits from using RFID and bar code to track or manage inventory. If bar codes are so cost prohibitive why do pharmacies scan them at the POS, rather than recording sales on a clipboard?

    ReplyDelete
  5. Nick Basta has an amusing way to describe Accenture's change of heart regarding track-and-trace: Project JumpStop. See Let’s call it Project JumpStop: NACDS/NCPA sponsors study slamming track-and-trace systems.

    I still can't understand why Accenture agreed to take on this assignment for the pharmacy associations.

    Adam

    ReplyDelete