Brazil: The Anvisa 2019 Pilot Report

Earlier this month Anvisa published an initial report on their drug traceability pilot as required by law there (see “Brazil Gets Rational With Their New Pharma Traceability Law” and “Brazil Publishes RDC-157 To Regulate 2017 3-Lot Pilot”).  I finally had time to use Google Translate on that report.  Google keeps improving their translation software and it is now to the point where I think the translation of this document is pretty good.  I still don’t recommend you use it for compliance decisions, but it’s great for getting a preview of what the document—originally in Portuguese—says.

The National Drug Control System (SNCM) pilot ran from last August through April of this year, although the report makes it clear that little piloting was actually done until the last two weeks of April.  That appears to be because the pilot test environment set up by GAESI was not fully available until the beginning of April, even though there is also a sentence that says the test environment was available for eight months prior to that (translation problem?).  GAESI is a group that does R&D into automation and process management technologies at the University of Sao Paulo (USP).  Apparently they are providing the technical horsepower for the development of the government’s central reporting system.

You can download the RxTrace Google Translate English version of the report here.  Sorry, I didn’t have time to translate and update the text in the figures.

The pilot tested a number of supply chain scenarios, including “activation”, “expedition”, “receiving”, “finalization”, “replacement”, returns and “revocation”, but not others, like theft, forgery, or drug recall.  Situations like movement of overdue product, activation of medication by a company that does not have registration, movement by unauthorized company, among others, were not tested, nor did public sector entities or patients participate in the tests.  For the first pilot, none of those excluded tests are very surprising.  First pilots always test a tiny amount of product moving in mostly normal, expected ways in the supply chain.  The problem is, almost nobody follows up those first pilots with tests of those exception scenarios. 

On the other hand, the report indicates that some exception handling was tested.  These include:

  • “movement of product not activated;
  • “drug activated with batch with incorrect validity;
  • “absence of packaging with IUM reported in the message shipment identified during the reading on receipt;
  • “packaging with a different IUM than that reported in the shipment identified during the reading on receipt;
  • “packaging with illegible IUM identified during reading in the receipt;
  • failure completion;”

The Anvisa database reportedly received 43,645 IUM (Unique Medicine Identifier) “activations” during the short duration of the pilot (see “The ANVISA Unique Medicine Identifier (IUM) on Drug Packages”).  Fully 30,000 were activated by a single participant.  They quickly came across the problem where the expiration date reported in data did not match the data encoded in the barcode.  That problem has been experience in the EU, here in the US and most recently in the Russian Federation (see “Global Differences In Expiration Date Encoding”).

Apparently, Anvisa hopes to continue testing, but without an announced schedule.

“As agreed between ANVISA and GAESI / HCFM-USP, deadlines and environments will be kept available for new tests to be carried out, new designs to be adjusted and increments to be developed.”

Anvisa and the participants learned some valuable lessons, despite the short duration of the working pilot.  Here are my three favorite lessons listed in the document:

  • “periods of unavailability of the prototype in the Anvisa environment negatively affected the tests. This caused actual drug handling events to be transmitted later than the one defined in DRC 157/2017, ie 3 days for record holders, 5 days for dispensers and 7 days for dispensers, leading the prototype to generate unnecessary extemporaneous movement;
  • “difficulty in treating substitution events by the prototype;
  • “The delay in finalizing the specification document in relation to the testing period reduced the number of participating technology companies, which can introduce a bias to the tests.”

My hope is that Anvisa and other regulatory agencies around the world realize that the ability for well-meaning companies to comply with their traceability laws often depends on the quality of the government reporting systems offered by those agencies and its documentation.  If you expect companies to report some action within 3 days of a supply chain transaction or event, then your central reporting system had better be up and available to receive that data.  Replicating the operations of a very large, multi-tier supply chain in digital form is incredibly complex, as the Anvisa pilot demonstrates.  We can only hope these lessons will result in heightened awareness that the design of the government reporting system must be robust, and that will be very expensive.  If the final sentence in the report is any indicator, maybe they have:

“…the experience of installing and maintaining the prototype in the Anvisa environment has reinforced the prior assessment of its limitations on availability and stability, as well as the need for SNCM to be installed in an environment that can guarantee these requirements.”

Well said.

There will be no RxTrace essay next week due to the Memorial Day holiday and I am taking some days off. Watch for the next essay on Monday June 3rd.