Wholesalers Find Troubling Results In DSCSA Barcode Assessment

Last month GS1 US published the results of an assessment of the implementation progress by drug manufacturers of the Drug Supply Chain Security Act (DSCSA) serialization requirements.  It contains troubling indications of the next problem the industry will face in its quest for DSCSA compliance:  too many unusable barcodes.

Big deal, you say?  That kind of complacency could come back to haunt you later this year.  Now is the time to look at your own packages and address any deficiencies.  Here is a look at what GS1 US found in their assessment.

In fact, GS1 published the assessment (called “Assessing Current Implementation of DSCSA Serialization Requirements”), and they helped facilitate it’s development, but the actual data was compiled by two of the largest wholesale distributors in the US, AmerisourceBergen (ABC) and McKesson back in May of 2017.  These three organizations conducted a DSCSA barcode assessment of a large percentage of the products in the inventories of one of their respective warehouses.

McKesson checked 13,571 products, representing 66% of the prescription drug items in one of their larger pharmaceutical distribution warehouses, including all refrigerated, all narcotics, and all mid- and high-turnover items.  ABC checked 3,047 products, representing 100% of the items in one of their specialty pharmaceutical distribution warehouses.  This division of labor likely resulted in checking the widest range of products covered by the DSCSA on the market today.

They checked every available barcode on each product, including the legacy linear barcodes as well as the new DSCSA 2D barcodes.  They checked readability as well as data content.

What did they find?  McKesson found that only 6.5% of the products in their catalog had a readable 2D barcode containing all four of the DSCSA-required data elements known as the DSCSA Product Identifier (see “The DSCSA Product Identifier On Drug Packages”).  ABC found only 7.2% of the same.  Yes, the data for this assessment was collected in wholesale distributor inventories back in May, but that was at a time when the industry was working toward a serialization deadline of November of last year (see “FDA Delays Enforcement of DSCSA November Deadline: What It Means“).  I’m sure the number of products serialized would be significantly higher if the assessment were conducted today, but even from this small sample, a significant percentage of products that manufacturers think they are done with, still have readability problems.  That means those manufacturers are going to have to make adjustments to those barcodes again—some minor, some major.

Here are the issues they found:

  • Colors are a problem
    Some manufacturers are attempting to make the DSCSA Product Identifier barcode blend in with the marketing information on their packages by printing it in the same color.  A few of those colors might work, but many do not work at all.  The best readability comes from black on white.  Anything other than that—including white on black—is less than optimal.  White on black using laser ablation is a very common way of printing a 2D barcode, but I have personally seen several attempts that will not work because the dots produced were inconsistent.  See the report for more information about problematic colors.
  • Placement problems
    They found barcodes that were printed too close together, which makes it hard for workers to read whichever barcode they need to read.  They found smudging, which I have observed can result from the ablated waste material being redeposited onto the package in the laser ablation process.  They also found some packages where the prescribing information outsert attached to the package was blocking the DSCSA 2D barcode so it was unreadable or difficult to read.
  • Data content problems
    This is the most disturbing of the problems they found, because it is so pervasive.  McKesson found that an additional 8% of the products they analyzed did have an attempt at the DSCSA 2D barcode on them, but these barcodes either did not have all of the mandated data elements present, or the information was encoded incorrectly so it wasn’t usable.  Think of it.  Only 6.6% of the products McKesson scanned contained usable 2D barcodes, but 8% more contained unusable 2D barcodes.  That implies that the majority of attempts at compliant 2D barcodes are not actually usable for compliance.  If that percentage is maintained as the number of serialized products rises—and there is little reason to believe it won’t—this is an indication that we have a serious problem in the industry.

Perhaps the best news contained in the GS1 US report is that both McKesson and ABC intended to privately contact the manufacturers of the problematic products and share more details about what they are doing wrong.  Hopefully that will result in corrections.  I am also hopeful that all wholesale distributors do that on a regular basis as more products arrive at their doors with attempted DSCSA 2D barcodes on them.  Drug manufacturers must listen to these reports and act to correct any problems reported.

Virtual manufacturers must use these wholesaler reports as the basis of a conversation with their CMO/CPOs.  In these cases, the problems can only be solved at the CMO/CPO.  But it is the virtual manufacturer who will be held responsible for non-compliance by the FDA—and by the wholesale distributors next year when the DSCSA will not allow them to buy non-compliant products.

What can you do?  I recommend that drug manufacturers send samples of their artwork and sample barcodes to the larger wholesale distributors for preemptive analysis and comment.  The “Big-3” offer this as a service.  Take advantage of those offers to avoid compliance headaches down the road:

Other resources include GS1 US’s DSCSA guidance document (see “The New GS1 US DSCSA Implementation Guidance Suite”) and Healthcare Distribution Alliance’s (HDA) barcode guidance document (see “HDA Guidelines For Bar Coding In The Pharmaceutical Supply Chain”).  Peter Sturtevant of GS1 US, Joe Lipari and I are going to talk about this assessment and what it means to the industry in a webinar on February 22.  Register here.

Don’t be one of the companies who fail this important test.


4 thoughts on “Wholesalers Find Troubling Results In DSCSA Barcode Assessment”

  1. Dirk,

    There is a lot of work to do to get from 6% to 100%. However, the work is simple and straightforward. Those designing and printing the barcodes have to check them for compliance before sending them out the door. It is a quality conrol issue. It is no different than making sure the product itself is right.

    Some of the findings in this audit suggest that a large number of companies did not even reference the specifications when planning packaging changes, or at least not completely. It also indicates that they did not check them out for print quality and compliance before sending them to distributors.

    I applaud McKesson, ABC and Cardinal for trying to help fix the problem but disagree that this approach. It could cause a delay in implementations. Labelers must take ownership of this responsibility within their organizations. They must learn what has to be done and build competency to do it well. I hope McKesson, ABC and Cardinal give their suppliers that advice. As a start, they should tell suppliers to read the relevant (UDI and DSCSA) regulations and their identification standards specs (GS1 or HIBCC). That would have eliminated the color, size, position and data content problems noted in the audit. That would be a good start.

    Mike Nolan – AIS

    1. Mike,
      Thanks for your comment. One way to look at what the big wholesalers are doing is setting the stage for tougher action in the future if bad barcodes aren’t fixed. If you are eventually going to get tough, you need to have offered help before you get to that point. In my view, wholesale distributors will be forced to stop buying products that are not labeled correctly or are not machine-readable on November 27, 2019. That is, they will be forced to get tough at that time because the law gets tough on them. No one wants to see that happen. So I too applaud the wholesalers and GS1 US for taking this early step.


  2. Great article Dirk. We sell a lot of barcode quality verification equipment, both sampling as well as 100% in line inspection and you would be amazed at how little some manufacturers know and the great offense they take at times when you try to tell them what they are doing wrong. Also some are deliberately sacrificing quality to maintain the throughput rates that they want. With one of the top ten pharma manufacturers who insisted on using laser ablation rather than inkjet we had to program their lasers so that some cells were not ablated at all to increase the speed to match their desired throughput because they were willing to sacrifice some error correction for higher speeds. It was nuts in my opinion because they then did not have that much remaining error correction to throw away if they had a real problem that was unplanned using up their error correction. I have seen tons of other quality issues as well.

    We have created inkjet systems where I can print as many as 3400 serialized codes per minute at 360dpi that we intended to use to feed multiple packaging lines but I cannot get that concept of having a system feed multiple lines to take hold here in the USA. We have had some success with it in Asia however and will soon be doing some systems for South America. Regarding laser ablation, we have been able to achieve higher speeds than most by using higher powered lasers so we can slew quicker because we accomplish the ablation quicker with higher power up to 150 watts.

    1. Robert,
      Thanks for that anecdote. That strategy you have observed will work until next year when the wholesalers are going to have to get tough on readability. After November 27, 2019 drug manufacturers who cannot meet the minimum quality standards are going to get a lot of product back as unsalable. That will be an eye-opener.


Comments are closed.