On 24 April 2026, in the early afternoon, Germany’s Paul-Ehrlich-Institut (PEI) — the federal agency for vaccines and biomedical medicines — posts from its official X account @PEI_Germany a card marking the end of European Immunisation Week. On the card: a cartoon family — mother, child, teenager, adult, elderly woman, male nurse in headscarf. Above it, in friendly green, with an exclamation mark:
“Now check your own vaccination status!”
The body text: “At the conclusion of the awareness week, the Paul-Ehrlich-Institut advises citizens to check their own vaccination status and, if applicable, catch up on missed vaccinations.”
Five hours later, engagement snapshot: 3,300 impressions, 18 likes, 13 retweets, 147 replies. For readers unfamiliar with social-media patterns: this is a ratio. A ratio occurs when a post draws substantially more disagreement than approval — here in the order of 8 to 1 replies to likes.
There is a reason. The reason is not that German citizens on average oppose vaccination. The reason is that the sentence “Now check your own vaccination status!” comes from an agency that has, during the decisive years, not documented its own core task — pharmacovigilance surveillance under § 77 of the German Medicines Act (AMG). This structural analysis sets the checking rhetoric against the checking gap that has now been confirmed by Freedom-of-Information records.
The Protocol Gap — Core Finding #
On 31 January 2026, the German investigative outlet Nachdenkseiten (author: Bastian Barucker) and Multipolar-Magazin published the result of a Freedom-of-Information request to the Paul-Ehrlich-Institut. The request: all meeting minutes of division SBD 1 (Pharmacovigilance) of the department for “Safety of Biomedical Medicines and Diagnostics” (SBD) since 2019.
The agency’s response:
- SBD 1 (Pharmacovigilance) — responsible for monitoring medicine safety: no written meeting minutes between 20 February 2020 and 19 June 2023.
- Entire SBD department (Safety of Biomedical Medicines): no written minutes between 2019 and 10 May 2023.
Those are exactly the years of the COVID-19 vaccination campaign. That is exactly the division responsible for detecting post-vaccination safety signals. That is exactly the division whose findings should have fed into the agency’s public communications, into authorisation extensions, into the booster recommendations, into parliamentary questions and into federal-government answers.
The PEI’s own justification, verbatim:
“Resources during the pandemic were prioritised on official duties, and the voluntary documentation of meetings was therefore deferred.”
Three words carry the work in that sentence: “voluntary documentation”. That is the PEI’s claim. The claim is legally contested. The Joint Rules of Procedure of the Federal Ministries (GGO) and the general record-keeping duty of the federal administration (§ 6 BArchG) suggest that meetings of a specialised division of a federal agency in a safety-critical field are indeed subject to documentation requirements. The Bundestag’s Covid Enquete Commission has formally requested the missing minutes — and is waiting for an answer.
The sharpest assessment is from Prof. Stefan Homburg (emeritus, public finance, Hannover; expert witness to the Covid Enquete Commission):
“It is simply out of the question that a senior federal agency would keep no minutes for three years.”
Homburg’s reading: either minutes exist and are being withheld — or they genuinely do not exist, which would constitute a breach of the division leadership’s official duties. Both are scandalous. Both warrant independent investigation.
Even former Health Minister Jens Spahn (CDU) — politically responsible for the 2021 vaccination campaign — has publicly declared himself “surprised” by the missing minutes. That is roughly the weakest possible show of solidarity with one’s own agency that an ex-minister can muster.
A secondary finding from the same FOI request worth noting: the PEI’s legal division (SBD 4) maintains 4.5 full-time lawyer positions whose workload, by their own statement, consists chiefly of handling FOI requests and lawsuits. An agency that deploys that much staff to resist transparency applications has a structural relationship to the public which the label “drug safety” alone no longer covers.
What the PEI Has Otherwise Not Delivered #
The protocol gap is not the only documented deficit. It is the hardest publicly documented one. Alongside it run further gaps, long raised in specialist and niche publications and systematically absent from leading-media reporting.
1. The Promised Safety Study on Statutory-Insurance Data #
At the start of the vaccination campaign the PEI publicly announced a risk-evaluation study on COVID-19 vaccines based on the routine data of the statutory health insurers. Multipolar documents in notice 0345 that five years later, no such study has been submitted. Germany has one of the densest statutory-health-insurance data bases in the world. The insurers have reports. Physician-visit rates before and after vaccination. Diagnosis codes. The data exists. The PEI announced the evaluation. The evaluation is not there.
2. SafeVac 2.0 — 700,000 Participants, Five Years of Silence #
The PEI’s own smartphone app SafeVac 2.0, launched in December 2020, was according to the PEI the largest vaccine safety study Germany has ever conducted. Over 700,000 people took part. Five years later the full evaluation has not been published. Dutch data analyst Wouter Aukema has reconstructed, from the EMA’s EudraVigilance database, that the SafeVac cases reported to Brussels deviate by about 42 per cent from the figures the PEI has cited in answers to parliamentary questions. On 7 April 2026 the federal government confirmed in BT-Drs. 21/5103 that the PEI’s figures on SafeVac 2.0 “in some cases did not coincide” — the analytical reconstruction of this scandal has been separately documented on this blog.
3. The Matthes–ImpfSurv Debate — Undercount as “Speculation” #
Prof. Harald Matthes, endowed professor at Berlin’s Charité and medical director of the Havelhöhe community hospital, found in his ImpfSurv study an adverse-event rate of approximately 0.8 per cent for serious reactions — against the roughly 0.02 per cent officially communicated by the PEI. The study has been methodologically criticised by Charité, Tagesspiegel and uebermedien.de (a different “severe” definition from EMA/FDA, open online survey). The critique is not baseless. But an analogous observation from the statutory-insurance sector comes from Andreas Schöfbeck (BKK ProVita), who derived a similar order of magnitude from his own insurers’ data — and lost his board position for saying so. Multipolar-Magazin documents that the PEI has classified the thesis of a systematic undercount as “speculation” — without presenting the in-house insurers’ study (see point 1) that could decide the question.
4. McKernan — DNA Contamination and the SV40 Sequence #
Kevin McKernan, former Human Genome Project team leader at MIT, together with other laboratories (including Phillip Buckhaults, University of South Carolina) has since 2023 detected residual plasmid DNA in mRNA-vaccine vials, including an SV40 promoter-enhancer sequence. The findings were peer-reviewed and published in September 2025 in Autoimmunity (Tandfonline). In its December 2023 statement, the PEI argued that “sufficient information” on the methodology was missing. An independent in-house re-measurement with publicly documented methodology has to this day not been presented by the PEI. This blog has covered the Bhakdi/Sterz line of the DNA debate separately.
5. Batch Heterogeneity (“Pfizer Roulette”) #
The Danish cohort study Schmeling et al. (BMJ 2023) and related work have shown that COVID-19 vaccine batches exhibit substantially heterogeneous serious-adverse-event reporting rates — inter-batch factors up to 1:30. The PEI has to date presented no public counter-analysis of the German batch landscape. This blog’s Pfizer-Roulette thread summarises the evidence.
6. ÄFI — The “Declaration of Bankruptcy” #
The German physicians’ association Ärztinnen und Ärzte für individuelle Impfentscheidung e.V. (ÄFI — Physicians for Individual Vaccination Decision) published in August 2025 at individuelle-impfentscheidung.de a piece under the title “The Complete Declaration of Bankruptcy of the Paul-Ehrlich-Institut”. The charges listed there: missing clinical efficacy evidence for STIKO-standard vaccinations, authorisation of the third mRNA vaccination for 12–17-year-olds without clinical trial data, handling of DNA contamination. ÄFI is a critical medical association, not the mainstream — but the documents it cites come from within the PEI.
The Division of Labour — and the Tweet in the Mirror #
Let us compile what is documented. The Paul-Ehrlich-Institut has:
- between 20.02.2020 and 19.06.2023 in the safety-critical pharmacovigilance division kept no written meeting minutes (FOI-confirmed);
- a five-year-old announced safety study on statutory-insurance data not delivered;
- SafeVac 2.0 with 700,000 participants not fully published five years after data collection began, and produced discrepancies of up to 42 % between its own parliamentary answers and EudraVigilance records;
- classified the thesis of systematic adverse-event undercount as “speculation” — without presenting the in-house study capable of refuting it;
- methodologically criticised an external peer-reviewed paper on DNA contamination in mRNA vaccines, without presenting its own controlled re-measurement;
- maintained 4.5 full-time lawyer positions in its legal division (SBD 4) predominantly occupied with the deflection of FOI requests.
This same institution sends into the world on 24 April 2026 a card with a cartoon family and the text: “Now check your own vaccination status!”
The rhetorical operation on display is the shift of the checking duty. The agency statutorily required by § 77 AMG to conduct drug- and vaccine-safety surveillance calls upon citizens to check for themselves. What they are to check is not the agency’s work. It is their own vaccination calendar. The direction of gaze is inverted. The checkers become the checked. The checked become the checkers. That is not only public-health communication. It is, in one word, institutional projection.
That this operation produced an 8:1 reply-to-like ratio on X within five hours is not the core finding — the core is the why. The X algorithm in such cases does not map an ideological camp; it maps a collective perception: the institution now demanding examination has not visibly performed the examination it is itself required to carry out. The replies, sampled, are not anti-vaccine pamphlets. They are questions about adverse-event reports, about batch testing, about the missing minutes. Precisely the questions to which the PEI has no published answers.
The Silence of the Leading Media #
The Nachdenkseiten scoop on the protocol gap dates from 31 January 2026. Multipolar confirmed it. The Covid Enquete Commission has formally requested the minutes. Spot-checks of tagesschau.de, spiegel.de, zeit.de, sueddeutsche.de, faz.net, welt.de, zdf.de for the combinations “Paul-Ehrlich-Institut minutes 2020 2023” and “PEI pharmacovigilance no minutes” yield zero relevant hits. The reporting runs via Multipolar, Nachdenkseiten, Tichys Einblick, Reitschuster, Alexander Wallasch, ÄFI, corona-protokolle.net — publications the mainstream frames as “contested” to “unserious”, which in this concrete case are the only ones publishing the primary-source FOI documentation.
This is the same pattern this blog has documented in other places this week. The BGH decision VII ZB 29/24 of 25 February 2026, which formally stripped Bavarian Broadcasting of its enforcement mechanism: 0 leading-media hits. The Hoppermann Enquete session of 23 April 2026, in which a former CDU judge delegitimised a witness through medicalisation of his childhood abuse: 0 leading-media hits (structural analysis here). The PEI protocol gap: 0 leading-media hits.
Three findings across three weeks, whose common denominator is not their content — but the fact that the leading media report on none of them. That is not conspiracy. That is, once again, omission. And omission is editorial work.
Who Wins, Who Loses #
Winners: The vaccination-campaign communications machinery, which in 2026 can still operate without structural counter-questioning from its own safety agency. The PEI’s legal division, whose 4.5 lawyer positions specialise not in answering the questions put to them but in litigating them. The leading-media desks that can present the PEI’s cartoon card to the public without the protocol-gap finding attached, because they are not sending it.
Losers: The injured, whose adverse-event reports land with a division that kept no written meeting minutes during the critical years. Researchers who would need the promised statutory-insurance study to locate the Matthes/Schöfbeck magnitudes on firm ground. The 700,000 SafeVac participants whose data has gone unevaluated in full for five years. Parliamentary oversight, which must react to figures differing by 42 per cent depending on who is asked. And — structurally — public trust in Germany’s institutional drug-safety architecture, which is not restored by a cartoon card when the checking duty of the checkers themselves has vanished from the written record.
We #
The card of 24 April 2026 is not the problem. The structure in which it appears is the problem. An agency statutorily required to examine, whose responsible division has filed no meeting minutes for three years, has displaced the checking duty onto the addressees of its own recommendations. That is not a public-health campaign. It is a role swap.
The PEI protocols of SBD 1 between 20 February 2020 and 19 June 2023 do not exist, on the agency’s own statement. The safety study on statutory-insurance data does not exist, five years after announcement. The SafeVac full analysis does not exist in full, five years after data collection began.
Whoever checks their own vaccination status now is not having a debate with the PEI. Whoever checks the PEI’s work is having one.
That is, on the last working day of European Immunisation Week 2026, this blog’s message to the Paul-Ehrlich-Institut. The duty to examine does not shift via a Twitter card. It stays where the legislature placed it. And when the replies run 8 to 1, that is not audience failure, it is feedback to the sender.
We know the sender. He has just not kept minutes for three years.
See also — from this blog #
- SafeVac 2.0: Five Years of Silence on Germany’s Largest Vaccine-Safety Study — the 42 % discrepancy, BT-Drs. 21/5103, EudraVigilance cross-check by Aukema.
- Pfizer Roulette: Batches, Placebo, BioNTech — Schmeling et al. (BMJ 2023), batch heterogeneity 1:30.
- Bhakdi/Sterz: DNA Contamination and Genetic Modification — the DNA-contamination thread.
- BGH VII ZB 29/24: The Director’s Name on the Line — parallel “leading-media silence” finding from the same day.
- Understand, Don’t Judge — The Müller Method — parliamentary delegitimisation of a Federal Interior Ministry witness.
Primary Sources #
- Paul-Ehrlich-Institut X account: @PEI_Germany — tweet of 24 April 2026 closing European Immunisation Week.
- Nachdenkseiten (Bastian Barucker, 31.01.2026): Ganze Abteilung des Paul-Ehrlich-Instituts legt während Corona-Impfkampagne kein Protokoll an.
- Multipolar-Magazin: Paul-Ehrlich-Institut: Keine Sitzungsprotokolle in Corona-Zeit angefertigt / Undercount as “speculation” / Missing statutory-insurance safety study.
- FragDenStaat: FOI request — pharmacovigilance protocols.
- Bundestag: Federal-government answer to AfD parliamentary question, BT-Drs. 21/5103, 07.04.2026.
- Peer-reviewed on DNA contamination: Speicher et al., Autoimmunity 2025.
- ÄFI — German physicians’ association for individual vaccination decisions: Declaration of bankruptcy of the PEI, 08/2025.
Note: This text is structural analysis on the basis of publicly documented primary sources (FOI answers from the PEI, Bundestag records, peer-reviewed publications). The quotations from Prof. Stefan Homburg and from the PEI’s justification are verified against the Multipolar notice 0367 and the Nachdenkseiten article. The assessment of leading-media silence rests on spot-check searches of tagesschau.de, spiegel.de, zeit.de, sueddeutsche.de, faz.net, welt.de, zdf.de. The political assessment is the author’s. Not medical or legal advice.
