METADATA last updated: 2026-03-02 RT initial creation file_name: _AI_FloodLAMP FDA October 2021 Correspondence Analysis.md file_date: 2026-03-02 title: FloodLAMP FDA October 2021 Correspondence Analysis category: regulatory subcategory: fl-fda-correspondence tags: fda-correspondence, deprioritization, eua, quickcolor, october-2021, open-protocol source_file_type: md xfile_type: NA gfile_url: https://docs.google.com/document/d/1mT_Hg7gHwWScfWfmGuaJcWK9gVLXB-NV3j0UthTwuDw xfile_github_download_url: https://raw.githubusercontent.com/FocusOnFoundationsNonprofit/floodlamp-archive/main/regulatory/fl-fda-correspondence/_AI_FloodLAMP%20FDA%20October%202021%20Correspondence%20Analysis.md pdf_gdrive_url: NA pdf_github_url: NA conversion_input_file_type: NA conversion: NA license: CC BY 4.0 - https://creativecommons.org/licenses/by/4.0/ tokens: 9443 words: 6447 notes: Created by Claude Opus 4.6 Max and ChatGPT 5.2 Pro Extended during archive preparation. **NOT HUMAN VERIFIED - MAY CONTAIN ERRORS** Comprehensive analysis of FloodLAMP's October 2021 FDA correspondence sequence, synthesizing the full email thread, FDA deficiency letter, meeting notes, and final deprioritization decision, with background from FDA town hall transcripts, open EUA context, and the AI deprioritization research report. Includes factual timeline summary and five interpretive explanations of FDA's rationale. summary_short: Analysis of FloodLAMP's October 2021 FDA correspondence -- the concentrated sequence of emails, deficiency letter, Zoom meeting (with Anne Wyllie), FloodLAMP's response, and FDA's final deprioritization/closure of EUA210582 for the QuickColor COVID-19 Test. Presents a complete timeline from October 2020 through October 2021, identifies FDA's stated justifications (clinical data gaps, resource triage, deprioritization policy), and offers five interpretive explanations of the FDA's rationale ranging from legitimate resource constraints to structural bias against small/novel entrants, with attention to the Omicron-wave irony five weeks later. CONTENT ## Prompt (Verbatim) Review the following files and the thing I want you to hone in on is I'll drag it over as a separate file. Well, you have these files on there all the October 2021 ones there in the archive combined files. So there are really four of these from from 2021, 10/4, 10/13, well two from 10/13 and then one from 10/20. So all the rest of the files are provided as like background and context so you know read and review all of those and then what I want you to do is summarize the interactions in all of those emails and then particularly what justification the FDA gave for not continuing our review and closing our case and refusing our EUA submissions and then... and then try to read between the lines and and use what you understand about the FDA's process and what's happening during the pandemic etc. to try to explain why they would shut us down. I mean Anne Wiley joined the call that we had with them. They knew she was an advisor for us. She was very successful. We submitted multiple times. I believe I discussed that we had the surveillance programs going and the test was successful. I think we even offered to submit that data. So you know this was in October and you know I mean maybe if they knew the Omicron wave was coming they wouldn't have done this. You know who knows but I want to try to understand this better. So both present the facts and then also try to step into their shoes and explain their rationale and you know if if you think it's best create you know a few possibilities of what their rationale and or thinking may have been. **Files included in context window:** - _archive-combined-files_fl-fda-correspondence.md - _AI_FDA Deprioritization of COVID-19 Diagnostic EUAs.md - _context-commentary_regulatory-fda-townhalls.md - _context-commentary_regulatory-open-euas.md - project-description.md - _context-commentary_regulatory-fl-fda-correspondence.md ## Prompt (Cleaned) Review all provided files, with particular focus on the October 2021 correspondence in the archive combined files -- the four key communications from 10/4, 10/13 (two that day), and 10/20. The remaining files provide background and context. Summarize the interactions in all of the emails, and specifically identify what justification the FDA gave for not continuing the review, closing the case, and refusing the EUA submissions. Then try to read between the lines, using what is known about the FDA's process and the pandemic context, to explain why they would shut FloodLAMP down. Anne Wyllie joined the call with the FDA. They knew she was an advisor. FloodLAMP submitted multiple times, had surveillance programs running with successful results, and offered to submit that data. This was October 2021 -- perhaps if the FDA had known the Omicron wave was coming five weeks later, they would not have made this decision. Present the facts, then step into the FDA's shoes and explain their rationale. Create several possibilities for what their rationale and thinking may have been. **Files included in context window:** - _archive-combined-files_fl-fda-correspondence.md - _AI_FDA Deprioritization of COVID-19 Diagnostic EUAs.md - _context-commentary_regulatory-fda-townhalls.md - _context-commentary_regulatory-open-euas.md - project-description.md - _context-commentary_regulatory-fl-fda-correspondence.md ## RESPONSE - Claude Opus 4.6 Max ### Summary of FloodLAMP-FDA Interactions: The October 2021 Correspondence #### Complete Timeline **Phase 1: Early Engagement (Oct 2020 - Dec 2020)** - **Oct 22, 2020**: FloodLAMP sends initial Pre-EUA request describing pooled RT-LAMP screening program. - **Nov 10, 2020**: FloodLAMP emails Dr. Tim Stenzel proposing a meeting about expanding the open-source EUA approach beyond SalivaDirect, with Anne Wyllie's involvement. - **Dec 9, 2020**: FloodLAMP engages Stenzel directly at an FDA Town Hall. Stenzel is encouraging: *"I look forward to hearing more about your development"* and confirms the open-source protocol and Rights of Reference model has real potential, calling it *"a pathway that could be very successful."* **Phase 2: Submissions and First Deprioritization (Mar 2021 - Jun 2021)** - **Mar 22, 2021**: FloodLAMP submits two full EUAs -- QuickColor LAMP (EUA210252) and EasyPCR (EUA210254) -- as open-source protocol EUAs with clinical validation from Stanford. - **Apr 2, 2021**: FDA deprioritizes both, declining to review. Justification: *"authorization of the test would have relatively limited impact on testing accessibility or testing capacity."* Signed by Donald J. St. Pierre. - **May 18, 2021**: FloodLAMP submits Pre-EUA (PEUA210313) for pooled home-collection kit and mobile app, trying to qualify for prioritization. - **Jun 17, 2021**: FDA responds to Pre-EUA. Critically, states: *"we have not reviewed the validation studies nor the data provided in this pre-EUA."* Flags the QuickColor PPA of 90% as below the 95% threshold. Closes the Pre-EUA. **Phase 3: The October 2021 Push and Final Closure** This is the concentrated sequence of four key communications: **1. Oct 4, 2021 -- FloodLAMP letter to Dr. Tim Stenzel (4 pages)** Randy True makes a comprehensive case directly to the FDA Director of OIVD. Key points: - Introduces FloodLAMP and the two complementary tests (LAMP and PCR) - Describes the open-source protocol approach with Anne Wyllie as advisor - Reports resubmission of both EUAs (now EUA210581, EUA210582) for persons suspected of COVID and serial screening - Presents real-world surveillance data: 5 sites in 3 states, ~2,300 people screened, 800 pools, 3 unknown positives detected, no known false negatives - Highlights manufacturing readiness: 2M tests on hand, 3M more at LGC Biosearch, NEB master mix supply - Frames $1-2/test pricing as "transformative" - Describes active commercial sites with EMS and municipal fire departments - Makes the Test-to-Stay argument for schools - Requests prioritized review and assistance avoiding immediate denial **2. Oct 13, 2021 -- FDA sends two communications** *Morning (8:25 AM):* Lead reviewer Liang Li sends a detailed 6-deficiency letter on EUA210582: 1. **Clinical study design**: wrong specimen type (AN swabs instead of NP), no intended use population description, no low-viral-load samples 2. **PPA of 90%** -- below FDA's 95% threshold 3. **Outdated inclusivity/variant analysis** (Feb 2021 data, no longer representative) 4. **Incomplete cross-reactivity** -- organisms with >80% homology not wet-tested 5. **Insufficient endogenous interfering substances** testing 6. **External controls too concentrated** (8x LoD vs. recommended <5x LoD) 7. **RUO reagent risk** -- cannot recommend RUO-labeled components for authorized test **Deadline given: October 15, 2021 at noon** -- two business days. *Late morning (10:15 AM):* Liang Li follows up offering a 30-minute Zoom meeting the next day. **3. Oct 14, 2021 -- Zoom Meeting** FloodLAMP: Randy True (CEO), Anne Wyllie (advisor) FDA: Liang Li, Silke Schlottmann, Kris Roth Key exchange points: - Randy requests to **amend the intended use to serial screening only** (removing "suspected of COVID"), which has a lower PPA threshold (~80%) - Silke asks about designation and **says they would not want the IFU publicly available** -- a direct tension with the open protocol model - Kris mentions **6 criteria for serial screening** and that the risk-benefit differs for POC/at-home vs. lab-based tests - Anne Wyllie asks if they could validate with a blinded reference panel; Silke says **the reference panel is no longer available** - Agreed FloodLAMP would provide written response by Oct 20 **4. Oct 20, 2021 -- FloodLAMP's Written Response** Randy submits a point-by-point response: - Requests amendment to remove "persons suspected of COVID" -- serial screening only - Narrows specimen claim to AN swabs only - Argues PPA threshold for serial screening is 80%, and performance exceeds all authorized antigen tests - Presents real-world surveillance data as supporting evidence - Cites multiple independent validations of the same LAMP components - Commits to updating inclusivity analysis, addresses cross-reactivity concerns specific to LAMP vs. PCR - Offers to qualify RUO lots Crucially, Randy also asks: *"Is the risk-benefit analysis for a high complexity CLIA lab rather than POC setting able to be factored in?"* and notes no response was received from the Oct 4 letter to Tim Stenzel. **5. Oct 21, 2021 -- FDA Final Decision (one day after FloodLAMP's response)** Liang Li sends a deprioritization/closure letter signed by Donald J. St. Pierre: > *"Given the volume of EUA requests the Agency has received, FDA is prioritizing review of EUA requests for tests, taking into account a variety of factors such as the public health need for the product, availability of the product, extent to which the product would serve a significant unmet medical need, and availability and adequacy of the information concerning the likelihood that the product may be safe and effective in diagnosing the disease/condition. Based on these factors, our review of your submission thus far, and the anticipated resources needed to continue review of your EUA request, FDA has determined that further review of your EUA request is not a priority. FDA therefore declines to issue an EUA for the product at this time."* **6. Oct 28, 2021 -- FloodLAMP appeals** Randy asks whether the intended use amendment to serial screening was understood before the closure decision, and requests reconsideration. He proposes two paths: (a) if the current clinical data is acceptable for serial screening, FloodLAMP will quickly complete the remaining items; (b) if not, he requests feedback on a new clinical study design. **7. Oct 29, 2021 -- FDA confirms and closes the door** Liang Li responds: > *"We understood the intended use amendment when we sent the Deprioritization Letter. Your current clinical study is not acceptable to support testing a screening population with serial testing."* Directs FloodLAMP to the updated molecular EUA template or to submit a pre-EUA amendment for study design feedback. The exchange effectively ends here. --- ### FDA's Stated Justifications The FDA gave three layers of justification, used in combination: 1. **Deprioritization on public health impact grounds**: *"authorization of the test would have relatively limited impact on testing accessibility or testing capacity"* -- used in both the April 2021 and October 2021 closure letters. 2. **Inadequate clinical performance data**: PPA of 90% was below the 95% threshold for the "persons suspected of COVID" intended use, and the clinical study design had deficiencies (wrong specimen type, no population description, no low-viral-load challenge samples). 3. **Inadequate clinical study for serial screening**: Even after FloodLAMP amended to the serial screening intended use (with a lower ~80% PPA bar), FDA determined that the clinical study itself was not designed to support a screening population claim. The study used de-identified remnant clinical samples from Stanford, not a screening population. 4. **Resource calculus**: The October 21 letter adds a new phrase not in the April letter: *"the anticipated resources needed to continue review of your EUA request"* -- suggesting the FDA was also factoring in the additional work (new clinical study, updated analytics, etc.) that would be needed to get FloodLAMP over the finish line. --- ### Reading Between the Lines: Why the FDA Shut FloodLAMP Down Based on the full documentary record, the broader FDA deprioritization analysis, and the pandemic context, here are several possible explanations -- from most to least charitable to the FDA. #### Explanation 1: Legitimate Resource Triage Under Genuine Constraints **The FDA's perspective**: By October 2021, the agency had authorized over 400 COVID-19 tests. Vaccines were widely available. The Delta wave, while severe, appeared to be subsiding. FDA staff were being shifted back to non-COVID review work (GAO documents this starting around May 2021). The agency faced an unprecedented backlog across its entire regulatory portfolio. Every hour spent on a submission that wasn't going to clear quickly was an hour taken from something else. FloodLAMP's submission, however meritorious its mission, presented a significant review burden: the clinical study needed to be redone, the analytical package needed substantial updating, the RUO reagent question was unresolved, and the open protocol model raised novel designation and quality assurance questions. From a pure resource allocation standpoint, a submission requiring extensive additional work from a small company without an existing EUA track record ranked lower than submissions from entities with complete packages or near-complete packages. **What makes this plausible**: The GAO data shows 558 tests declined to review by September 30, 2021 -- FloodLAMP was one of hundreds. The FDA was applying this policy broadly, not singling FloodLAMP out. The two-day response window, while harsh, was apparently a standard mechanism during this period. **What complicates this**: The FDA's own prioritization criteria emphasized tests that "increase testing accessibility" and "expand testing capacity." FloodLAMP's offering -- instrument-free, $1-2/test, open protocol enabling lab proliferation, commodity supply chain -- arguably fit those criteria well. The FDA never substantively engaged with these claims. #### Explanation 2: The Clinical Data Gap Was Real and Dispositive **The FDA's perspective**: Regardless of mission, price point, or scalability claims, an IVD EUA requires adequate validation data. FloodLAMP's clinical study had genuine problems: - 90% PPA, not 95% - Anterior nasal swabs only, while claiming broad URT specimen types - No description of study population characteristics - No low-viral-load challenge samples (the hardest test of sensitivity) - De-identified remnant samples from Stanford -- not a prospective screening population study Even at the 80% PPA threshold for serial screening, the FDA's position was that the *study design* was wrong, not just the performance number. A screening population claim requires a study on a screening population, and remnant clinical samples from symptomatic patients tested at a hospital lab don't demonstrate that. **What makes this plausible**: This is technically sound. FDA templates were clear about these requirements. The June 2021 pre-EUA response explicitly flagged the PPA issue. FloodLAMP had months of advance notice. **What complicates this**: FloodLAMP asked for guidance on study design repeatedly -- in the May 2021 pre-EUA, in the October 5 interactive review request, and in the October 20 response. The FDA's response to the May pre-EUA was to not review the data at all and close the file. On October 14, when FloodLAMP asked about the clinical requirements, the reference panel that could have helped was "no longer available." And the final closure came one day after FloodLAMP's response -- suggesting the decision was already made before the response was evaluated. #### Explanation 3: Institutional Discomfort with the Open Protocol Model **The FDA's perspective**: The open-source protocol EUA was still a novel regulatory construct. SalivaDirect was the only example, and it was backed by Yale -- a major research institution with significant credibility and infrastructure. FloodLAMP was a tiny public benefit corporation asking FDA to authorize a test whose Instructions for Use would be publicly available, whose components were commodity RUO-labeled reagents, and whose deployment model involved designating potentially hundreds of labs. The October 14 meeting notes are revealing: Silke Schlottmann specifically said *"they would not want us to make IFU publicly available"* -- a direct challenge to the foundational concept of an open protocol EUA. If the IFU isn't public, the protocol isn't open. This suggests the FDA reviewers, or at least some of them, were uncomfortable with the regulatory implications of this model expanding beyond the SalivaDirect precedent. There were genuine regulatory questions: Who is responsible for quality when any lab can run the protocol? How do you ensure lot-to-lot consistency of RUO reagents? What happens when a designated lab deviates from the protocol? These are legitimate concerns, but they were never worked through -- the submission was closed before they could be. **What makes this plausible**: The regulatory system is not designed for open-source diagnostics. The standard model assumes a single manufacturer with a quality system controlling its product. The open protocol model disrupts that assumption, and FDA reviewers would reasonably be cautious. **What complicates this**: Tim Stenzel, the OIVD Director, had explicitly endorsed the open protocol approach on record in the December 2020 town hall. But Stenzel never responded to the October 4 letter, and the review team -- who made the actual decisions -- appeared to be less aligned with the concept. #### Explanation 4: The "Pandemic Is Ending" Assumption **The FDA's perspective**: In October 2021, the dominant institutional narrative was one of pandemic management trending toward resolution. Vaccines were widely deployed, boosters were being authorized, therapeutics were in the pipeline. The acute crisis of "we need every test we can get" had passed. The FDA's own town hall language reflected this: *"at this time the pandemic is very different"* (March 2021); tests were being deprioritized because the agency perceived less urgency. A small company's LAMP test -- which needed substantial additional validation work, wouldn't be POC or at-home, and was instrument-free but still high-complexity CLIA -- simply didn't register as urgent in a world where the pandemic was supposed to be winding down. **The devastating irony**: Omicron (B.1.1.529) was first reported to the WHO on November 24, 2021 -- approximately five weeks after the FDA closed FloodLAMP's submission. By January 2022, Americans were standing in lines that *"snake around entire city blocks"* trying to get tested. The U.S. experienced the worst testing shortage of the entire pandemic. The very thing FloodLAMP was offering -- massive-scale, low-cost molecular screening using commodity reagents and an open protocol that any CLIA lab could adopt -- was exactly what was needed. Would FloodLAMP's authorized test have prevented the Omicron testing shortage? Almost certainly not by itself. But the principle behind the decision -- "we don't need more tests right now" -- was catastrophically wrong, and it was applied systematically to FloodLAMP and hundreds of other developers. #### Explanation 5: Structural Bias Against Small, Mission-Driven Entrants **The FDA's perspective** (unstated): The FDA's EUA prioritization criteria implicitly favored established diagnostics companies. "High-throughput" meant companies with existing high-volume manufacturing. "Widely distributed" meant companies with existing distribution networks. "POC" and "at-home" meant companies with the capital to run the required usability studies and manufacture consumer devices. FloodLAMP's claims -- "$1-2/test," "millions of tests by end of year," "10K+/day without automation" -- were extraordinary, but they came from a company with no existing EUA, no manufacturing facility, and two active commercial surveillance sites serving EMS workers. From a risk-averse regulator's perspective, these were aspirational claims from an unproven entity, not demonstrated capacity. Anne Wyllie's presence on the October 14 call was meant to signal credibility -- she was the developer of SalivaDirect, the most successful open protocol EUA. But the review team had apparently already set their course. The letter to Stenzel went unanswered. The surveillance program data -- while genuinely impressive for a startup -- was a rounding error in national testing volumes. **What makes this plausible**: Regulators evaluate track record and demonstrated capacity, not potential. The FDA's job is not to pick winners or invest in promising startups; it's to determine whether a specific device meets specific criteria. **What complicates this**: This is precisely the critique that the GAO, legal scholars, and industry commentators made about the deprioritization regime. It created a circular barrier: you can't demonstrate capacity without authorization, but you can't get authorization without demonstrated capacity. For a small public benefit corporation operating on grants and mission-driven funding, this was a death spiral. --- ### Synthesis The most likely reality is a combination of all five factors, weighted roughly in this order: 1. **Resource triage was the dominant driver.** The FDA was overwhelmed and had a systematic policy of closing submissions that required extensive additional work. FloodLAMP's submission genuinely needed more work. 2. **The clinical data gap was real but should have been solvable.** FloodLAMP needed a prospective screening population study. The FDA never gave them meaningful feedback on study design despite repeated requests. The June 2021 pre-EUA response didn't review their data at all, and the October process gave them 48 hours to respond to a 6-deficiency letter before closing the case. 3. **The open protocol model created institutional friction** that the review team was not equipped or motivated to work through, particularly in a resource-constrained environment. 4. **The "winding down" assumption** about the pandemic made every submission from a small developer with an incomplete package look like a low-return investment of review resources -- an assumption that proved catastrophically wrong within weeks. 5. **Structural bias** against small/novel entrants was an ambient factor, not a deliberate choice, but it meant FloodLAMP's arguments about potential impact were likely discounted relative to what a large manufacturer with an existing supply chain would have been credited. The single most telling detail in the entire record may be this: the FDA sent a 6-deficiency letter on October 13, offered a 30-minute call on October 14, received a substantive written response on October 20, and issued a final closure one day later on October 21. The speed of that closure -- particularly the one-day turnaround on a response that proposed a narrowed intended use and addressed multiple deficiencies -- suggests the outcome was predetermined and the "interactive review" was procedural rather than genuinely deliberative. FloodLAMP was being processed through the deprioritization pipeline, not evaluated on the merits of its amended proposal. ## RESPONSE - ChatGPT 5.2 Pro Extended Below is a fact-first reconstruction of what happened in the October 2021 thread (the four “October files” you pointed to), plus the most important earlier context in the correspondence bundle. I then separate out what FDA **explicitly said** from what can be **inferred** (with clear labels), and I end with a set of plausible “FDA-in-their-shoes” rationales for why they shut this down. I’m basing this entirely on the archive files you provided; the October 10/14 “meeting notes” are *your* notes (not a verbatim FDA transcript), so I treat them as directional evidence, not quotes. --- ### What happened, chronologically #### The key pre-October setup (why Oct 2021 isn’t “out of the blue”) * **Mar 22, 2021:** FloodLAMP submitted two “open protocol” EUA requests (QuickColor LAMP and EasyPCR) and positioned them as low-cost, supply-chain-robust, scalable protocols modeled on SalivaDirect, including a planned blanket Right of Reference (ROR) strategy. * **Apr 2, 2021:** FDA issued a **final decision** on EUA210252 (and similarly on EUA210254) and **closed** the EUA(s), stating they were deprioritizing because FDA was in “a different phase of the pandemic…where many COVID-19 tests are now authorized,” and FDA was prioritizing tests that would **increase testing accessibility** (POC, home collection, at-home) or **significantly increase testing capacity** (high-throughput). FDA said FloodLAMP’s test would have “relatively limited impact” on accessibility/capacity. * **May–Jun 2021 (Pre-EUA PEUA210313):** FloodLAMP submitted a pre-EUA that bundled tests + pooled home collection + app. FDA said, due to volume/prioritization, they **did not review the validation studies/data**, but gave “general recommendations,” including (critically) stating FDA believes **~95% positive and negative agreement** is acceptable clinical performance and that QuickColor’s PPA was below that; FDA explicitly said they **did not recommend pursuing** an EUA for QuickColor unless sensitivity improved. They then **closed the pre-EUA** upon issuing the feedback. * **Oct 5, 2021:** FloodLAMP followed up asking for an **interactive review** and emphasizing real-world surveillance deployment metrics (sites, pools, people screened, positives detected/confirmed, etc.), while noting resubmitted EUAs (EUA210581 / EUA210582). That’s the backdrop: FDA had already signaled “not priority” in April, and in June had already flagged QuickColor sensitivity as below what they considered acceptable for the use as framed then. --- ### The October 2021 interactions (the four files you flagged) #### Oct 4, 2021 — FloodLAMP escalates to leadership (email routed to Tim Stenzel) FloodLAMP sent an email intended for Dr. Tim Stenzel via the FDA templates inbox, asking for a meeting and help avoiding another immediate deprioritization of the resubmitted EUAs. FloodLAMP’s main points in that email: * Positioned QuickColor (colorimetric LAMP) and EasyPCR (duplex RT-PCR) as **open protocol** tests using **TCEP + heat inactivation**, with claimed turnaround and throughput (e.g., “10K+/day without automation”), and stated LoDs (LAMP ~12,500 cp/mL; PCR ~3,000 cp/mL). * Highlighted that earlier EUAs were deprioritized and the May pre-EUA was closed with only brief comments; requested leadership assistance so the resubmissions wouldn’t be “immediate denial of review.” * Put major emphasis on **public health impact**: pooled home collection (family/exposure unit pooling), “test-to-stay,” low cost ($1–2/person), and ability to scale through manufacturing partners. * Cited **non-diagnostic surveillance deployments** (5 sites/3 states, ~800 pools, ~2,300 people screened, positives detected/confirmed, and “no known or suspected false negatives”). * Asked for FDA guidance on CLIA lab expansion / near-site testing and how surveillance programs might coexist with EUA-authorized clinical testing. #### Oct 13, 2021 — FDA requests additional information + offers a Zoom call There are two distinct FDA communications on Oct 13: 1. **Lead reviewer email attaching the “Additional Information” letter** FDA (lead reviewer Liang Li) sent FloodLAMP a formal letter requesting additional information for EUA210582. 2. **Follow-up email offering a 30-minute Zoom meeting the next day** The same reviewer offered a quick Zoom slot to discuss the submission. ##### What FDA’s Oct 13 “Additional Information” letter actually says This is the most detailed, “technical deficiencies” document in the October set. It asserts that FloodLAMP had not provided complete information adequate to validate the test and support the claimed performance characteristics, and it lays out deficiencies in six major areas (I’m summarizing, not reproducing the whole letter). **(1) Clinical evaluation problems** (FDA’s “big bucket”) * **Intended-use population not described**: FDA couldn’t tell whether the study samples came from the intended population (people “suspected” by their HCP). * **Specimen-type claims not supported**: FloodLAMP claimed multiple URT swab types (OP/NP/AN/mid-turbinate), but the study used AN swabs, and FDA stated NP is considered the most challenging matrix and expected for NP/general URT claims. * **Performance below what FDA considered acceptable**: FDA cited FloodLAMP’s PPA/NPA (90% / 100%) and reiterated that FDA believes **95% PPA and NPA** is acceptable clinical performance for the intended use “as you proposed,” and that 90% PPA was below that. * **No “low positive” / low viral load specimens** in the dataset as FDA defined them (Ct close to comparator LoD); FDA said without these they couldn’t assess performance appropriately. * FDA asked for root-cause analysis, possible design changes, and potentially new validation if the device/procedure changes. **(2) Inclusivity / variant analysis outdated** * FDA said the in silico inclusivity analysis (done Feb 27, 2021) was no longer representative of circulating variants and asked for an updated analysis + risk assessment. **(3) Cross-reactivity / microbial interference** * FDA noted specific organisms that had high homology in silico but weren’t covered by wet testing, and asked for tabular in silico data and additional interference testing where needed. **(4) Endogenous interfering substances** * FDA said the tested substances were insufficient for common URT cold remedies and listed additional recommended substances. **(5) External controls** * FDA recommended the positive control be “challenging” (low positive, <5x LoD) and said the described control (~8x LoD) wasn’t adequate; requested validation data. **(6) RUO reagents** * FDA said RUO components are not under 21 CFR Part 820 and therefore can vary; FDA said you may not recommend RUO-labeled materials as test components without additional mitigations, and they gave options (use IVD-labeled materials, qualify specific RUO lots and publish lot numbers, or bring materials under your own QMS by supplying them as kit components). Finally, FDA set a very short turnaround (by Oct 15 noon ET) and stated that if the response was late or inadequate, they could remove the test from the notification listing and expect suspension of distribution. (Your Oct 14 meeting notes indicate the written response deadline was effectively aligned to Oct 20 instead, which suggests there was at least some informal flexibility. ) #### Oct 14, 2021 — Zoom meeting (your notes) Per the meeting notes: * FDA raised concerns about the **open protocol / designation** concept; specifically, Silke asked about “designation” and “said they would not want us to make IFU publicly available.” * FloodLAMP described a plan: initially deploy like SalivaDirect, later produce a full reagent kit. * FloodLAMP communicated a request to **amend intended use** to serial screening only (removing “persons suspected of COVID”). * FDA asked about data supporting intended population; FloodLAMP said samples were remnant, de-identified, no clinical info. * FDA (Kris) referenced “6 criteria” for serial screening amendment and said risk-benefit differs for POC/at-home vs lab-based tests; also noted pipetting choices can affect whether something is “POC-like.” * Anne Wyllie asked about validating with a blinded reference panel; Silke said the reference panel was no longer available. #### Oct 20, 2021 — FloodLAMP responds (email + attached response letter) FloodLAMP’s Oct 20 email to FDA explicitly frames the central issue as **risk-benefit**, and that FDA’s thinking (as FloodLAMP understood it) might be: serial-screening “streamlined” pathways don’t apply because this is a **lab** submission rather than POC. FloodLAMP also asked for the “6 criteria” Kris referenced, and reiterated the “open protocol + scalable colorimetric LAMP” impact argument. FloodLAMP’s attached response letter (the separate Oct 20 response file) did several concrete things: * **Formally requested intended-use amendment**: remove “persons suspected of COVID,” keep serial screening, and asked for guidance on appropriate screening frequency given performance data. * **Narrowed specimen claim**: remove NP/OP; use nasal swabs only. * **Disputed the performance bar for serial screening**: FloodLAMP stated its understanding that the PPA threshold for screening with serial testing is 80% and argued their analytical + clinical results exceeded many authorized antigen tests. * Provided additional performance framing: noted only high-Ct samples failed in Stanford clinical evaluation; tied this to LoD; and again cited real-world pooled surveillance metrics (pools, people screened, positives found, “no known or suspected false negatives”). * Said updates on variants/inclusivity would be provided; pushed back on some cross-reactivity assumptions as “PCR-like” rather than LAMP-specific; asked for guidance on interference list given sample-type/intended-use changes. * On controls, argued the chosen positive-control concentration mirrored SalivaDirect and asked FDA to confirm suitability. * On RUO reagents, said FloodLAMP would qualify specific lots and post lot numbers. --- ### The actual “shutdown”: what FDA said when it closed the file Although you didn’t list Oct 21/Oct 29 as separate “October files,” they’re part of the same October correspondence record and they contain the **explicit closure rationale**. #### Oct 21, 2021 — FDA final decision: EUA210582 closed FDA told FloodLAMP the additional information “does not demonstrate” the device is adequately validated; the EUA was closed. In the attached letter, FDA combined two themes: 1. **Substantive deficiency framing** They said the submission did not adequately demonstrate the device is validated / safe and effective for intended use (implicitly referencing the prior deficiency list). 2. **Resource + prioritization framing (“why stop now”)** FDA said: given the volume of EUA requests, FDA is prioritizing based on factors including public health need, availability, unmet need, and the adequacy of the information about likely safety/effectiveness; based on those factors, their review so far, and “anticipated resources needed to continue review,” further review was “not a priority,” and FDA declined to issue an EUA “at this time.” So the closure is best read as: **(a) the file still had major gaps, and (b) closing it was also a triage decision about reviewer time.** #### Oct 28–29, 2021 — FloodLAMP challenges whether the intended-use amendment was recognized; FDA replies FloodLAMP pointed out that the closure letter still referenced the “suspected of COVID” intended use and asked FDA to confirm whether the amendment request was understood. FDA replied that they **did understand** the intended-use amendment when they sent the deprioritization letter, but that FloodLAMP’s “current clinical study is not acceptable to support testing a screening population with serial testing,” and directed FloodLAMP back to the new molecular template or a pre-EUA amendment/supplement for study design feedback. This matters because it suggests the amendment wasn’t the “unlock”; FDA’s decisive blocker (at least as stated in Oct 29) was **clinical-study acceptability for serial screening**, not merely the wording of intended use. --- ### FDA’s explicit justifications, boiled down If you strip out tone and just list the agency’s stated reasons across the thread (April → June → Oct), the “why we shut this down” story has three layers: #### 1) Deprioritization / triage due to pandemic phase + impact criteria * April 2021: FDA explicitly said the test would have “relatively limited impact” on U.S. testing accessibility/capacity, so review was “not a priority,” and they closed the EUA. * June 2021: FDA said they didn’t review the validation studies/data in the pre-EUA because of prioritization and closed the pre-EUA after providing limited guidance. * Oct 2021: FDA again used prioritization language, adding that the “anticipated resources needed to continue review” factored into the decision to stop. #### 2) “Not adequately validated” — with a very specific deficiency list The Oct 13 letter reads like FDA saying: **to authorize this, we’d need a lot more (and different) evidence**—especially around clinical performance, intended use population, low viral load positives, and updated variant inclusivity. #### 3) For serial screening specifically: “your clinical study isn’t acceptable” Even after FloodLAMP tried to narrow intended use to serial screening and AN swabs, FDA’s Oct 29 email says directly: the current clinical study is not acceptable to support screening with serial testing. --- ### Reading between the lines: plausible “FDA-in-their-shoes” rationales Everything below is explicitly **inference**—but I’m tying each hypothesis to something concrete in the record. #### Hypothesis A: FDA saw this as a “high-effort file” with multiple major gaps, so they cut it to protect scarce reviewer capacity Clues: * The Oct 13 letter is not “one missing table”; it’s a multi-domain deficiency set (clinical population/specimens/performance, variants, interference, controls, RUO supply chain/QMS). * On Oct 21, FDA explicitly referenced “anticipated resources needed to continue review” as part of why further review was not a priority. Interpretation: * Even if the test might ultimately be authorizable, FDA likely perceived that getting to “yes” would require **new clinical work + multiple analytical studies + QMS/controls/supplier mitigation**, i.e., a lot of back-and-forth. Under triage pressure, those files are the ones you close first. #### Hypothesis B: The “serial screening pathway” you tried to use may have been treated as mainly justified by *accessibility gains* (POC/at-home), and FDA didn’t see the same benefit for a lab-based test—so the risk-benefit balance didn’t “buy down” the evidence requirements Clues: * Your Oct 20 email indicates FDA’s risk-benefit framing on the call sounded like: lab-based reduces benefit relative to POC/at-home. * Meeting notes record Kris saying risk-benefit is different for POC/at-home vs lab-based tests, and referencing “6 criteria” for serial screening amendment. Interpretation: * FDA may have built the serial-screening indication to enable *more testing where testing otherwise wouldn’t happen* (especially at home/near patient), accepting different tradeoffs. If your test still requires a CLIA lab workflow (even if instrument-free), FDA may have categorized it as “not the kind of accessibility expansion that justifies a lower bar.” #### Hypothesis C: FDA likely discounted your “surveillance deployments” evidence because it wasn’t in the clinical validation form they rely on for EUA decisions (and because it was explicitly non-diagnostic) Clues: * You repeatedly offered real-world screening metrics (Oct 4 email and Oct 20 response). * FDA’s deficiency focus stayed anchored to classic EUA validation constructs: defined intended-use population, comparator-based clinical agreement, low viral load positives, etc. Interpretation: * Even genuinely strong field performance can be hard for FDA to credit if it’s not (a) prospectively designed, (b) comparator-grounded in the required way, and (c) clearly within the exact intended use. It may still be persuasive as “supporting evidence,” but not as a substitute for the missing validation pieces FDA flagged. #### Hypothesis D: QuickColor’s clinical sensitivity and “challenging positives” problem was the real immovable object Clues: * In June 2021 FDA already told you QuickColor’s PPA was below what FDA considered acceptable (95%) for the use as framed there, and they did not recommend pursuing unless sensitivity improved. * In Oct 2021 FDA again emphasized: PPA 90% vs the 95% they “believe is acceptable,” plus no low viral load specimens. * In Oct 29 FDA said your current clinical study is not acceptable for serial screening. Interpretation: * From FDA’s perspective, **false negatives in asymptomatic screening** can have outsized public-health consequences, and they may have judged that (i) 90% PPA plus (ii) a dataset without low viral load positives equals “we cannot responsibly authorize this” without substantial additional work. #### Hypothesis E: The “open protocol” posture created extra regulatory discomfort: uncontrolled replication + RUO quality + stewardship/designation issues Clues: * Your Oct 14 notes: FDA did not want IFU publicly available; asked about designation; this is directly about governance/control of an “open protocol” approach. * The Oct 13 letter is unusually explicit about RUO reagents and the quality-system implications, and requires a mitigation route. * The open-EUAs context file underscores that “open EUA” is not just publishing a protocol—it’s an authorization + stewardship pattern with lab designation burdens and supply chain validation complexities. Interpretation: * FDA may have been comfortable with SalivaDirect’s model (single steward, specific lab designation commitments, very clearly bounded method), but less comfortable with a commercial entity proposing “open protocol” + eventual kit commercialization + broad dissemination—especially if key inputs were RUO and if performance wasn’t already rock-solid. * In other words: open can be a public-good feature, but to a regulator it can look like a **control problem** unless governance and quality controls are extremely crisp. #### Hypothesis F: Timing/phase-of-pandemic thinking likely mattered—but not in the way people assume Clues: * FDA repeatedly referenced “different phase of the pandemic” and “many tests now authorized” when deprioritizing. * Your own internal commentary notes the experience felt like blanket deprioritization and triage driven by staffing/resource constraints. Interpretation: * It’s plausible FDA, in Oct 2021, believed the U.S. already had adequate diagnostic molecular capacity and was trying to concentrate on (a) at-home/POC accessibility and (b) high-volume industrialized products, not additional lab-developed-like workflows—even if your workflow was cheaper and operationally elegant. * This is consistent with your speculation about Omicron: in Oct 2021, they weren’t behaving like an agency preparing for a January 2022 demand shock; they were behaving like an agency trying to **wind down a crushing review backlog** under a “we have enough tests” narrative. --- ### A compact way to summarize “why they shut you down” If you want a one-paragraph explanation that is faithful to the record **and** captures the implied logic: FDA appears to have viewed FloodLAMP’s QuickColor EUA as (1) still **not adequately validated** for the claimed intended use—especially due to clinical study limitations (population not characterized, specimen claims not supported, PPA below FDA’s typical bar, and no low-viral-load positives) plus additional analytical/QMS issues (variants/inclusivity, interference, controls, RUO reagents)—and (2) **too resource-intensive** to continue reviewing in a late-2021 triage environment where FDA was prioritizing tests that most clearly expanded accessibility (POC/at-home) or capacity, so they closed the file as “not a priority” rather than continue iterative review. --- If you want, I can also do a second pass that’s more “forensic,” where I map **each** FDA deficiency (Oct 13 letter items 1–6) to (a) what you offered in the Oct 20 response, (b) what FDA implicitly accepted vs didn’t address, and (c) which items likely triggered “new clinical study required” vs “straightforward paperwork/testing.”