820.120 UDI (Unique Device Identification) Labeling Verification Requirements

[DCONN1]

Hello all,

I am looking for guidance as to what exactly is required/mandated in terms of establishing barcode quality by the FDA UDI final rule. My knowledge on this topic is not comprehensive, and I certainly understand that using standards such as ISO/IEC 15415 and ISO/IEC 15416 could be of great use. However, I haven't been able to find any requirement mandating a specific barcode quality standard. The most relevant information I've encountered is:

§ 820.120 Device labeling

"(b)Labeling inspection.Labeling shall not be released for storage or use until a designated individual(s) has examined the labeling for accuracy[emphasis mine] including, where applicable, the correct unique device identifier (UDI) or universal product code (UPC), expiration date, control number, storage instructions, handling instructions, and any additional processing instructions. The release, including the date and signature of the individual(s) performing the examination, shall be documented in the DHR."

However, this section does not address quality; if you have a scanner which is able to read the barcode and show that the information is accurate, would this not satisfy this requirement?

Section (a) of § 820.120 remains unchanged by the UDI final rule and is relevant to the printed quality of the UDI:

"(a)Label integrity.Labels shall be printed and applied so as to remain legible and affixed during the customary conditions of processing, storage, handling, distribution, and where appropriate use."

"Remaining legible" in terms of the barcode (as well as human readable components) would certainly involve ensuring sustained readable barcode/human readable quality in some manner. However, since ISO/IEC 15415 and ISO/IEC 15416 are not referenced within the UDI final rule to my knowledge, is there a requirement to conform to these standards? Or is the method and scope of demonstrating compliance left up to the labeler?

My company uses GS1 as our accredited issuing agency. I am aware that GS1 has implementation guides for establishing verification programs for 1D and 2D barcodes which conform to /IEC 15415 and ISO/IEC 15416. However, is it necessary to follow these guidelines, or can a labeler establish a program as they see fit?

Again, I apologize for any ignorance on this topic and any help would be very much appreciated.

 

[Mark Meer]

Hi DCONN1,

I've also been wondering the same thing.

There is a thread started sometime ago discussing this (but no resolutions yet...).
See http://elsmar.com/Forums/showthread.php?t=63958

MM.

 

[ScottK]

I went with customer expectations on this issue.

Every one had an SOP that I had to review and they all required compliance to ANSI standards with an overall grade of C or better. As far as getting to that point - it was totally up to us.

While you don't need a bar code grader/verifier the expectation was that we would validate our printing system. And grader came in realty handy for that.

Most customers, in my experience, expect that you will own the grader/verifier to do regular checks.

 

[Gert Sorensen]

GS1.org has quite a bit of information about UDI on the website.

 

[Mark Meer]

GS1.org has quite a bit of information about UDI on the website.

They do, indeed.

...however, the question that I (and the OP, I presume) are more interested in is regarding the FDA's position and requirements/expectations. There is a lot of 3rd party guidance out there, but as far as anything official from the FDA...this is not clear.

...and, to be honest, it's not that clear in GS1 either. For example their technical overview of the Datamatrix states that print quality should be verified, and uses ISO/IEC 15415 as an example (but then go on to say it's "defacto" application standard...). They indicate throughout that verification needs to verify that (customer) requirements are met.

So back to the FDA, what does this mean? Is it acceptable to develop your own (simplified) tests for verifying print-quality with justifications such as risk, and redundant human-readable data?

...or, like many guidances seem to imply, is there an expectation that manufacturers need to invest in formal verification according to established standards (e.g. to ISO 15416, ANSI X3.182, ISO/IEC 15426-1, ISO/IEC TR 29158, ...etc.)?

 

[DCONN1]

Thanks for the replies all.

Yes, as Mark stated, I am most interested in determining the actual requirements of the UDI rule and what my company is mandated to do. I've continued to research this issue, but I've yet to come across anything resembling a clear mandate to use these standards. As such, if customers do not have relevant requirements pertaining to barcode quality testing, I would be hard pressed to justify the investment in expensive verification equipment, adding to production processes, and increasing training requirements.

21 CFR Part 830.20 states:

"A unique device identifier (UDI) must:
(a) Be issued under a system operated by FDA or an FDA-accredited issuing agency;
(b) Conform to each of the following international standards:
(1) ISO/IEC 15459-2, which is incorporated by reference at 830.10;
(2) ISO/IEC 15459-4, which is incorporated by reference at 830.10; and
(3) ISO/IEC 15459-6, which is incorporated by reference at 830.10.
(c) Use only characters and numbers from the invariant character set of ISO/IEC 646, which is incorporated by reference at 830.10."

While we will be sure to purchase and review these standards for compliance, does anyone know if ISO/IEC 15459-2, ISO/IEC 15459-4, ISO/IEC 15459-6 references any barcode quality standard such as ISO/IEC 15415 or ISO/IEC 15416?

 

[Mark Meer]

Hi DCONN1,

I've decided to adopt a risk-based approach, and just document our decisions.

As was described in a previous thread, it can be said that the printed AIDC codes are:

1. Indelible (we test this internally);
2. Properly coded (we validate the software that does the encoding);
3. Scannable (we test); and
4. Redundant (the human-readable text is always provided).

(3) may be the only item which proper verification may be challenged, however given the intended use and risk of the device, and the redundancy described in (4), I just don't see any reasonable argument for doing extensive print-quality verification as per ISO/IEC standards.

 

[DCONN1]

Just as an update on this issue, the FDA released a draft guidance on July 26, 2016 titled "Unique Device Identification System: Form and Content of the Unique Device Identifier".

The paragraph beginning on line 264 addresses print quality with respect to ISO/IEC standards:

If a labeler choses a bar code form of AIDC, the bar code form of the UDI should be tested for print quality. Please refer to the most recent version of the following standards for more information on how to determine the print quality: ISO/IEC 15416 Information technology -- Automatic identification and data capture techniques -- Bar code print quality test specification -- Linear symbols; ISO/IEC 15415 Information technology -- Automatic identification and data capture techniques -- Bar code symbol print quality test specification – Two-dimensional symbols; and ISO/IEC TR 29158 Information technology -- Automatic identification and data capture techniques – Direct Part Mark (DPM) Quality Guideline. For linear and 2-D bar codes, labelers should consult the most recent version of the standards listed above, and the guidelines of their FDA-accredited issuing agency, to determine the minimum overall symbol grade based upon ISO/IEC verification processes. For purposes of this draft guidance, we define "overall symbol grade" as the arithmetic mean of the grades of multiple scans of the symbol. The minimum acceptable grade should be satisfied under the expected handling and use life of the device. Labelers should discuss print quality requirements with their FDA-accredited issuing agency.

To my knowledge this is the first time UDI print quality has been directly addressed by the FDA. It now seems as if companies which choose not to verify their barcodes with appropriate equipment according to these standards would not be in compliance.

 

[Mark Meer]

Hmmm... this is always a frustrating element of regulatory creep: when "draft guidances" are treated (both by industry and the FDA) as de-facto mandatory.

Unfortunately, I suspect you are correct: that this draft guidance will become the standard expectation of compliance, and any deviation from it will raise issues.

(*Sigh*) I guess we'll all just have to budget for a few thousand dollars extra in additional equipment, training, and documentation maintenance just to comply (despite arguably little value to the company).

 

[Sam Lazzara]

As a basic quality control measure I believe it is important to at least inspect each run of printed labels to make sure the the barcode can be correctly read using a simple "scanner" (USD $50 - $400) if you do not have the budget to use a real "verifier" (USD $3000-5000). My recommendation would be to either scan or verify at least 1 label printed for each unique production batch/lot.

Use a scanner is infinitely better than doing nothing. Using a verifier instead of a scanner does buy you assurance, but nowhere near the gain achieved when you go from nothing to a scanner.

And using a scanner is still a way to "check" print quality, albeit without the intense rigor of a real verifier, and without being able to claim conformance with the ISO print quality standards.

By the way, after doing quite a bit of research I landed on LVS verifiers as a good solution, for those with the verifier-bucks.