Are Your Test Methods Scientifically Valid?

After a bit of a hiatus during the pandemic, FDA in-person inspections of dietary supplement manufacturers and distributors are once again increasing in frequency.  FDA recently released updated data for Fiscal Year 2021, which just closed on September 30th.

The first thing to notice is that Form 483 observations are once again rising after being in steady decline since 2018. The past couple of years has seen a subtle change in FDA Form 483s for dietary supplements, with the new appearance of 21 CFR 111.75(h)(2)1 in the top ten citations, which requires tests and examinations used for dietary supplements to be scientifically valid.  In case you’re not familiar, this section of the dietary supplement cGMP regulations reads as follows:

21 CFR Sect. 111.75(h)(1) You must ensure that the tests and examinations that you use to determine whether the specifications are met are appropriate, scientifically valid methods. (2) The tests and examinations that you use must include at least one of the following:
(i) Gross organoleptic analysis;
(ii) Macroscopic analysis;
(iii) Microscopic analysis;
(iv) Chemical analysis; or
(v) Other scientifically valid methods.

You could be forgiven for thinking that, provided you are using at least one of the above methods, you are complying with the regulation.  But lack of knowledge regarding the specific analytical methods used to test your products (and components) may turn out to be a big issue when you are inspected.

FDA’s focus during inspections is by now firmly fixed on examining the specifications you set for finished dietary supplements and their components, and ensuring that compliance with those specifications is tested using scientifically valid, fit for purpose methods.  What does that mean in practice? It means you must be prepared to do due diligence, whether you use an internal or external third party lab, to ensure that the analytical methods used to test your supplements are scientifically valid for the product under test.  In other words, you can’t just accept whatever method your lab prefers (or the cheapest method). You must understand the methods used and ensure the test will give accurate, reliable results in your particular situation.

Based on recent FDA Warning Letters, inspectors are finding several problems in this area. First, I think most firms have heard by now that “by input” is not acceptable as a “method.” A recent Warning Letter states “Relying solely on the amount input is not a valid scientific method of confirming that specifications for strength of the finished batch are met.”

The letter further states: “You may exempt one or more product specifications from verification requirements under 21 CFR 111.75(c)(1) if you determine and document that the specifications you select for determination of compliance with specifications are not able to verify that the production and process control system is producing a dietary supplement that meets the exempted product specification and there is no scientifically valid method for testing or examining such exempted product specification at the finished batch stageIn such a case, you must document why, for example, any component and in-process testing, examination, or monitoring, and any other information, will ensure that such exempted product specification is met without verification through periodic testing of the finished batch. However, you have not prepared such documentation to exempt any of these specifications from finished product verification. ” [my emphasis]

There may be circumstances in which no  scientifically valid method exists to determine the strength (or composition) of a finished dietary supplement.  For example, the product matrix (formulation) may contain multiple overlapping components, or substances that interfere with available analytical methods, and could result in inaccurate or unreliable data. In that case, though, you can’t just use “by input” — you must DOCUMENT the reasons why the analysis is not scientifically valid, and that no valid method exists. Then, that documentation must be signed and approved by QC personnel.  I suggest you don’t risk irritating your FDA inspector by claiming that you can’t do analytical testing on your final product without having the prerequisite documents in place.2

A similar regulation in 21 CFR 111.320 — not currently so commonly cited as 21 CFR 111.75(h) — mandates using methods that are appropriate for their intended use, i.e. “fit for purpose”. The subtle difference is that, for example, infrared spectroscopy may be a valid analytical method in some circumstances, but invalid for a particular product without the appropriate controls and reference materials.  A recent Warning Letter states:

“You failed to verify that the laboratory examination and testing methodologies are appropriate for their intended use, as required by 21 CFR 111.320(a). Specifically, your [REDACTED] dietary supplement contains [REDACTED] and [REDACTED]. You conduct infrared testing (IR) on the finished product, purportedly to confirm the identity and composition of the contents; however, the results of your IR tests do not identify the identity and composition of the contents encapsulated as the finished dietary supplement.”

An analytical method may also be generally scientifically valid to determine whether some specifications of a product are met, but not specific enough for the circumstances of use — in another Warning Letter, FDA states:

“For example, visual and organoleptic attributes are indicated to confirm the identity of your finished dietary supplement products, however those attributes are unacceptable to identify powdered products without a complementary identity test that is more specific to the analyte, such as HPLC with an indicated method or reference standard to establish identity. Other attributes indicated are limited to a calculated percentage of protein, fat and moisture content by NIR Spectroscopy, and pH by potentiometric measure. However, none of these attributes provide specific analytical standards to confirm the identity, purity, or strength of your finished products as these characteristics are shared by many of the finished products and would not be sufficient to distinguish them from each other.” [my emphasis]

Below are the 10 most frequent citations under 21 CFR Part 111 during FY 2021 (yes, there are 11 on this list!)3  Note that the relatively low number of citations reflects the limited inspections carried out during FY 2021, and we may see a change in the relative order as inspections ramp up again in FY 2022.  Stay tuned!

Specifications – identity, purity, strength, composition55 (citations)
Written procedures – quality control operations29
Specifications-component purity, strength, composition28
Specifications – component identity24
Written procedures – product complaint22
Batch record – complete21
Written procedures – holding21
Specifications met – verify; finished batch15
Component –  verify identity, dietary ingredient15
Master manufacturing record – each batch14
Tests, examinations – scientifically valid14

References:

FDA Warning Letters (Accessed 10/13/22)

https://www.fda.gov/inspections-compliance-enforcement-and-criminal-investigations/warning-letters/oregons-wild-harvest-inc-621842-08082022

https://www.fda.gov/inspections-compliance-enforcement-and-criminal-investigations/warning-letters/muscle-feast-llc-627030-07222022

https://www.fda.gov/inspections-compliance-enforcement-and-criminal-investigations/warning-letters/mcaco-ltd-624606-04212022

Notes:

  1. I don’t really understand why the 483s are citing 111.75(h)(2) instead of (h)(1) – perhaps someone at FDA can explain this? Actually, the 483 doesn’t give you the citation number, but the number is in the FDA records…you can use my lookup page to find the citation reference number if you get a 483, or ask your inspector at the closeout meeting.
  2. Contact me if you need help with documentation https://www.marianboardley.com/contact.php
  3. The figures above are taken from FDA’s downloadable spreadsheets. They may differ slightly from the data used to create my charts, as the charts use FDA’s data dashboard, which is updated more frequently. In both cases, data may be redacted due to pending FDA regulatory activity.