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:

• A risk-based approach to serialization (RBATS) for high-risk products with the greatest likelihood of counterfeiting or diversion, which is included in H.R. 5839 Safeguarding America’s Pharmaceuticals Act of 2008.
• “Normal Channel of Distribution” rules that exempt certain transactions in many states. Even California may incorporate this exemption, judging by the Schwarzenegger’s Administration new legislative proposal for California pedigree regarding an “accredited distribution chain.”
• Central-fill operations and mail order pharmacies that aggregate prescription fulfillment would dramatically lower the per pharmacy costs. (Neither is contemplated in the study.)
• The willingness of wholesaler’s and other third-parties to provide data center services as a fee-based service.

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.