METADATA
last updated: 2026-03-19 RT
file_name: _context-commentary_regulatory-fl-fda-submissions.md
category: regulatory
subcategory: fl-fda-submissions
gfile_url: https://docs.google.com/document/d/1hDlXqrnIvU096_wlMpKcl9oFBAuSiIalUDnoIWi5FUk
words: 1944
tokens: 3338


CONTENT

## Context
This `regulatory/fl-fda-submissions` subcategory contains FloodLAMP's FDA Emergency Use Authorization (EUA) submissions and associated Instructions for Use (IFU) documents for its SARS-CoV-2 diagnostic tests. There are 15 files spanning from November 2020 to October 2021, representing the full arc of FloodLAMP's regulatory submission effort. For background on the EUA framework itself and how it differs from standard 510(k) IVD approvals, see the `regulatory/fda-policy` subcategory. For the correspondence with the FDA that accompanied these submissions, see `regulatory/fl-fda-correspondence`.

The subcategory documents a pre-EUA in 2020 and 2 rounds of multiple EUA submissions. The "QuickColor" test is the main FloodLAMP test, and one that was used for all of FloodLAMP's surveillance pilot programs. **When the "FloodLAMP test" is used throughout the files in this archive, the direct, extraction-free colorimetric LAMP visually-read "QuickColor" test is the one being referred to.**

### Pre-EUA (Nov 2020) - Purified "Glass Milk" LAMP Test
The earliest submission, a colorimetric RT-LAMP assay using silica ("glass milk") purification of nucleic acids from saliva and anterior nares swab specimens (from the Harvard Rabe-Cepko protocol). This was a pre-EUA — a preliminary package submitted to initiate FDA engagement and enter the review queue. It targeted asymptomatic screening with sample pooling at a baseline level of 10. The limit of detection (LoD) was 2 copies/uL. The outcome of this pre-EUA was an hour-long call with a lead reviewer who strongly recommended against submitting for asymptomatic and pooling, and instead and instead recommended the standard indication of individual person suspected of COVID. We followed that advice, perhaps mistakenly (see Commentary in `regulatory/fl-fda-correspondence/_context-commentary_regulatory-fl-fda-correspondence`).
- `regulatory/fl-fda-submissions/2020-11-06_Pre-EUA Sub - FloodLAMP Glass Milk LAMP Test.md`

### First Round EUA Submissions (March 2021) - Direct Extraction-free QuickColor LAMP and EasyPCR Tests
The first round of submissions included 2 extraction-free tests: "QuickColor" - colorimetric visually-read LAMP and "EasyPCR". Both tests use the same TCEP inactivation with the PCR test using the SalivaDirect primers and probes (a variation of the CDC test). An additional test, "QuickFluor" fluorimetric LAMP, was included in all of the wet-validation experiments as well as the Stanford clinical validation, but was not submitted due to both the performance and the lack of need for the test.

#### FloodLAMP QuickColor COVID-19 Test
(EUA Submission + IFU, v1.0 March 2021, v1.1 May 2021, v1.2 draft October 2021): An extraction-free, colorimetric RT-LAMP assay with a visual pink-to-yellow readout requiring no specialized instrumentation, based directly on the Rabe-Cepko assay (see `various/papers` and `various/lamp-tech`). It used 18 LAMP primers targeting three SARS-CoV-2 genes (ORF1ab, N, E) and the same TCEP-based inactivation as the EasyPCR test. The LoD was 12,500 copies/mL. Clinical evaluation at Stanford showed 90% positive agreement and 100% negative agreement on 80 specimens. The v1.2 IFU introduced triplicate repeat procedures for inconclusive results. 
- `regulatory/fl-fda-submissions/2021-03-22_EUA Submission - FloodLAMP QuickColor COVID-19 Test v1.0.md`
- `regulatory/fl-fda-submissions/2021-03-22_Instructions for Use - FloodLAMP QuickColor COVID-19 Test v1.0.md`

#### FloodLAMP EasyPCR COVID-19 Test
(EUA Submission + IFU, v1.0 March 2021, v1.1 May 2021): An extraction-free, duplex RT-qPCR assay using the CDC N1 and human RNaseP primer-probe sets, with TCEP-based chemical inactivation and heat treatment. It required standard RT-PCR instruments (QuantStudio, Bio-Rad CFX96) but no nucleic acid extraction equipment. The LoD was 3,100 copies/mL. Clinical evaluation at Stanford showed 97.5% positive agreement and 100% negative agreement against a high-sensitivity comparator on 80 specimens.
- `regulatory/fl-fda-submissions/2021-03-22_EUA Submission - FloodLAMP EasyPCR COVID-19 Test v1.0.md`
- `regulatory/fl-fda-submissions/2021-03-22_Instructions for Use - FloodLAMP EasyPCR COVID-19 Test v1.0.md`

#### FloodLAMP FLAMP (QuickFluor) COVID-19 Test (NOT SUBMITTED)
(EUA Submission + IFU draft, March 2021, NOT SUBMITTED): A fluorimetric RT-LAMP assay using real-time fluorescence readout on RT-PCR instruments. It used the same primers and inactivation as QuickColor but with fluorescent detection instead of colorimetric. The LoD was 50,000 copies/mL and clinical evaluation showed 80% positive agreement and 100% negative agreement. This test was prepared but never submitted to the FDA.
- `regulatory/fl-fda-submissions/2021-03-26_EUA Submission - FloodLAMP FLAMP COVID-19 Test NOT SUBMITTED.md`
- `regulatory/fl-fda-submissions/2021-03-26_Instructions for Use - FloodLAMP FLAMP COVID-19 Test NOT SUBMITTED.md`

### Second Round EUA Submissions (May 2021) - Re-sub QuickColor and EasyPCR plus Pre-EUA for Pooling and Asymptomatic
The second round consisted of revised May 2021 QuickColor and EasyPCR submission packages together with the separate pooling and asymptomatic screening pre-EUA materials.
- `regulatory/fl-fda-submissions/2021-05-18_EUA Submission - FloodLAMP QuickColor COVID-19 Test v1.1.md`
- `regulatory/fl-fda-submissions/2021-05-18_Instructions for Use - FloodLAMP QuickColor COVID-19 Test v1.1.md`
- `regulatory/fl-fda-submissions/2021-10-01_Instructions for Use - FloodLAMP QuickColor COVID-19 Test v1.2.md`
- `regulatory/fl-fda-submissions/2021-05-18_EUA Submission - FloodLAMP EasyPCR COVID-19 Test v1.1.md`
- `regulatory/fl-fda-submissions/2021-05-18_Instructions for Use - FloodLAMP EasyPCR COVID-19 Test v1.1.md`

#### Pooled Swab Collection and Screening Studies
(Pre-EUA submissions, May 2021): Three documents supporting FloodLAMP's pooling and asymptomatic screening strategy, including a pre-EUA for a direct-to-consumer pooled swab collection kit (allowing 1–4 self-collected anterior nasal swabs in a single tube), collection kit instructions, and a proposed validation study design for pooling and serial asymptomatic screening aligned to FDA guidance.
- `regulatory/fl-fda-submissions/2021-05-18_Pre-EUA Sub - FloodLAMP Pooled Swab Collection DTC.md`
- `regulatory/fl-fda-submissions/2021-05-18_Pre-EUA Sub - FloodLAMP Pooled Swab Collection Kit DTC.md`
- `regulatory/fl-fda-submissions/2021-05-18_Pre-EUA Sub - FloodLAMP Proposed Pooling and Asymptomatic Screening Study.md`

Each submission follows the FDA's EUA template structure: purpose, measurand, applicant information, regulatory status, proposed intended use, device description and test principle, controls, interpretation of results, manufacturing and component sourcing, and performance evaluation (LoD, inclusivity, cross-reactivity, interfering substances, clinical evaluation). The IFU documents mirror much of this content in an operational format for laboratory use.

All of FloodLAMP's tests shared a common TCEP-based chemical inactivation step and were designed as "open source protocol" tests — meaning all reagent components were fully disclosed (no "reaction mix 1", etc.) such that designated CLIA high-complexity laboratories could source all components directly from commercial vendors rather than purchasing proprietary kits, reducing cost and enabling supply chain redundancy. This open-source approach is central to FloodLAMP's strategy, as documented in the `regulatory/open-euas` subcategory. The EasyPCR and QuickColor tests were designed to work as an integrated pair: QuickColor for high-throughput colorimetric screening (45-minute turnaround, no instruments) and EasyPCR for rapid PCR-based confirmation (~1 hour 45 minutes). New England Biolabs supported both tests with their LAMP master mix and Luna RT-qPCR kit, and LGC Biosearch Technologies supplied production-scale LAMP primer sets.

The submissions evolved across versions. The v1.0 submissions (March 2021) proposed intended use for "individuals suspected of COVID-19 by their healthcare provider and from individuals without symptoms or other epidemiological reasons to suspect COVID-19 infection, when tested at a weekly interval." The v1.1 submissions (May 2021) reframed the intended use around "routine screening programs" in settings like schools and workplaces. None of these submissions received FDA authorization.

### Archive Files Not Converted to Markdown
`2020-11-04_Pre-EUA Sub Supporting Data - Glass Milk LAMP Test.xlsx`
`2021-03-22_EUA Sub Supporting Data - FloodLAMP EasyPCR COVID-19 Test.xlsx`
`2021-03-22_EUA Sub Supporting Data - FloodLAMP QuickColor COVID-19 Test.xlsx`
`2021-03-26_EUA Sub Supporting Data - FloodLAMP FLAMP COVID-19 Test NOT SUBMITTED.xlsx`
`2021-05-18_Pre-EUA Sub - FloodLAMP Pooled Swab Collection Kit DTC design file.sketch`


## Commentary
The primary commentary for submissions is combined with the `regulatory/fl-fda-correspondence` subcategory — see `regulatory/fl-fda-correspondence/_context-commentary_regulatory-fl-fda-correspondence.md` regarding FloodLAMP's FDA engagement. Here are comments on the more technical aspects of the EUA submission process.

The EUA submission process involves two main types of documents: the submission itself and the Instructions for Use (IFU). The submission is the regulatory document — it follows the FDA's structured template. The IFU is the operational document intended for the laboratories that will run the test. They overlap substantially in content, but serve different audiences and purposes. Both must be prepared and aligned, and both are substantial documents. For FloodLAMP, each test's submission ran 5,000–10,000 words, and each IFU ran 7,000–14,000 words.

A critical concept in understanding these documents is that the FDA authorizes test systems, not tests in isolation. The entire system — reagents from specific vendors, validated instruments, the exact workflow, controls, and interpretation criteria — is what receives authorization. Changing a single component can require revalidation. For open-source protocol tests like FloodLAMP's, this posed an inherent tension: the whole point was to enable broad deployment using commodity components from multiple suppliers, but the regulatory framework required specificity at every level. FloodLAMP addressed this by validating multiple instruments (QuantStudio 6 Flex, QuantStudio 7 Pro, Bio-Rad CFX96) and multiple primer and probe vendors (Eurofins Genomics, IDT, LGC Biosearch) within a single submission.

The bench science for EUA validation can actually move quite fast. In a conversation with another test developer, an academic who started a company during the pandemic and did obtain an EUA, they described doing the core validation work in a weekend. The bottleneck is not the assay work itself but the surrounding logistics and documentation. FloodLAMP also had a significant delay in getting registered with BEI Resources to obtain the cross-reactivity reagents. The FDA requires testing against a panel of related pathogens and respiratory flora, and BEI is the primary source for those reference organisms. The registration and shipping process takes weeks, and for a small organization running lean, that waiting period was a significant constraint.

As an aside regarding the fluorimetric version of the LAMP test, which we did all of the validation work for but did not submit, it showed a worse limit of detection (LoD) in our hands compared to colorimetric LAMP (50,000 vs. 12,500 copies/mL). However, we spent very little time optimizing the fluorimetric test, and our requirement was that all 3 tests (colorimetric LAMP, fluorimetric LAMP, and PCR) use the same inactivated sample so there was no flexibility there. Importantly, New England Biolabs (NEB) used a combination colorimetric plus fluorimetric LAMP test for its in-house workplace screening program (see `various/papers-lamp/Tanner (NEB and Mirimus) - Preprint - Extraction-Free Saliva SARS-CoV-2 RT-LAMPWorkflow for Workplace Surveillance (3-11-2022).pdf`). Their LoD was quite similar in magnitude (40,000 copies/mL), though the matrices are different (raw saliva vs. swab eluate in TCEP+saline), so direct numerical comparison has to be done carefully. Both are roughly aligned with what you'd expect from a sensitive RT-qPCR comparator — the Tanner paper reported a clinical sensitivity of 97% and 100% specificity, which is higher than what FloodLAMP saw from the Stanford Clinical Lab clinical validation run (80% positive agreement and 100% negative agreement). See the files and section of the context and commentary in `regulatory/fda-policy` related to the FDA Reference Panel for perspective on the problems related to test performance metrics.

Even with the FDA's templates available, the document preparation for COVID-19 EUAs was a heavy lift for anyone who has not done it before. FloodLAMP benefited from other groups sharing their actual EUA submission documents — these are not published, and they differ in important ways from the public-facing authorizations and IFUs that appear on the FDA website. Having real examples was helpful for understanding the level of detail and the conventions that the templates alone do not fully convey. If the FDA allowed or facilitated sharing of these submissions, for example, by simply letting developers choose to publish their submissions openly, some likely would, especially academics and public-interest/nonprofit developers. This would meaningfully lower the barrier for new entrants.

The new capabilities of AI could transform the document preparation burden of regulatory submissions and compliance in the disease testing field. EUA submissions and IFUs are highly structured, repetitive across test types, and draw heavily on standardized language. With progress on standardization and transparency from the FDA — such as machine-readable templates, published example submissions, and clearer guidance on acceptable variations — AI tools could handle much of the drafting, cross-referencing, and consistency-checking that currently consumes weeks of a developer's time. The combination of AI and regulatory modernization could significantly reduce the cost and time to bring validated, innovative tests to market. See also `regulatory/reg-articles-misc/_context-commentary_regulatory-reg-articles-misc` and `regulatory/reg-articles-misc/2021-01-18_Phillips and Dinakar - A Proposal for Increasing the Speed of Validating SARS-CoV-2 Diagnostic Tests`.