How Are Vaccines Tested Before FDA Approval?

How are vaccines tested before FDA approval? A detailed look at the pre-licensure safety testing process — placebo controls, trial phases, follow-up periods, and what the evidence shows.

The clinical trial process conducted before a vaccine receives FDA licensure. Aaron Siri's work argues this process is structurally designed to avoid detecting adverse events, using inadequate safety monitoring windows, tiny trial populations, and non-inert comparators — making it impossible to detect chronic, autoimmune, or neurological harms.

Explanation

Before a vaccine is licensed, the pharma company seeking licensure funds and conducts clinical trials. It then submits trial data to the FDA. The FDA does not conduct trials itself.

A properly designed pre-licensure trial would:

• Compare a vaccine group to an inert placebo group
• Monitor safety for a sufficient duration to detect delayed adverse events (including neurological and autoimmune disorders, which may not manifest for months or years)
• Enroll a large enough population to detect rare events

Pre-licensure testing is particularly critical because, once a vaccine is licensed, conducting a placebo-controlled trial is considered unethical — meaning the licensure trial is the only opportunity to properly assess safety in a controlled manner.

Evidence and Examples

Recombivax HB (Merck Hepatitis B vaccine)

• Safety monitoring: 5 days after each injection
• Population: 147 infants and children
• Control group: none
• Stanley Plotkin was the principal investigator on the children's arm of this trial and did not recall these parameters during his deposition until shown the FDA package insert
• The FDA licensed a vaccine administered to newborns on day one of life based on a clinical trial that tracked safety for just five days — a fact confirmed by the FDA-approved package insert and ICAN FOIA records.
• After deposition, Plotkin was subpoenaed for proof of extended safety data; he moved to quash the subpoena and never provided data
• ICAN FOIA confirmed no additional data existed

Engerix-B (GSK Hepatitis B vaccine)

• Safety monitoring: 4 days after each injection
• Plotkin claimed in deposition that safety was surely collected beyond 4 days — then admitted: "Yes, that's speculation based on experience."
• ICAN sued FDA; after years of litigation, FDA never produced a single document showing safety review beyond a few days

HHS response

• HHS pointed ICAN to 7 studies on post-licensure Hep B safety in response to a formal safety data request
• 4 studies involved adults (not infants)
• 3 were short-window clinical trials — none reviewed safety beyond a few days

Key Deposition Exchange (Recombivax HB)

> Q: How long was the safety review period in the prelicensure clinical trial for this vaccine?

>

> A: [reads package insert] Five days.

>

> Q: Is five days long enough to detect adverse reactions that occur after five days?

>

> A: No.

>

> Q: Is five days long enough to detect an autoimmune issue that arises after five days?

>

> A: No.

>

> Q: Is five days long enough to detect any neurological disorder that arose from the vaccine after five days?

>

> A: No.

>

> Q: There's no control group, correct?

>

> A: It does not mention any control group, no.

> — Plotkin Deposition 2018, p. 19–20

Full Placebo Table and Pyramid Scheme

The Complete Placebo Table

No routine childhood injected vaccine on the CDC schedule was licensed based on a placebo-controlled trial. Every vaccine used another vaccine (often itself unlicensed) or an active substance as the "control." The following table summarizes each vaccine's comparator (citing FDA source documentation):

First 6 months of life (given 3 times):

Later childhood vaccines:

Siri challenged this publicly. Paul Offit initially claimed "all vaccines are tested in placebo-controlled trials before licensure" (June 22, 2023). After Siri posted proof, Offit quietly changed his claim to "most vaccines." Siri argues that is still categorically false. Offit also falsely claimed the Salk trial used "salt water" as a control — the actual official trial report describes injections of "199 solution" (synthetic tissue culture + ethanol), "phenol red," "antibiotics," and "formalin."

The Pyramid Scheme of Safety

This is a house of cards: each vaccine is licensed as "as safe as" the previous vaccine, none of which was ever validated against a placebo. The base of the house has no safety baseline.

Example — PCV pyramid:

• Prevnar 7 licensed as "as safe as" an unlicensed experimental vaccine
• Prevnar 13 licensed as "as safe as" Prevnar 7 → 8.2% SAE rate (vs. 7.2% for Prevnar 7) — "safe" because rates were similar
• Vaxneuvance licensed as "as safe as" Prevnar 13 → 9.6% SAE rate (vs. 8.9%)
• Prevnar 20 licensed as "as safe as" Prevnar 13

Example — DTaP pyramid:

• DTP (repeatedly shown to increase infant mortality in developing countries) was used as the baseline of safety for licensing DTaP (Infanrix)
• Daptacel (second DTaP): 3.9% SAE rate within 30 days — dismissed because no placebo control

Notable SAE rates from clinical trials (all dismissed because rates were similar to control, which was itself never placebo-controlled):

• Heplisav-B (Hep B, adults): 6.2% SAE (vs. 5.3% Engerix-B)
• Daptacel (DTaP, infants): 3.9% SAE within 30 days
• Prevnar 13: 8.2% SAE in 6-month review
• MMR-RIT (Priorix, new GSK MMR): 10.1% ER visits, 2.1% SAE, 3.4% new onset chronic diseases (autoimmune disorders, asthma, type I diabetes, vasculitis, celiac disease, thrombocytopenia, allergies) — buried in supplemental table

Gardasil: AAHS as Control

Gardasil used AAHS (Amorphous Aluminum Hydroxyphosphate Sulfate, a proprietary neurotoxic aluminum adjuvant known to induce autoimmunity in lab animals) as the control for 9,092 of 9,412 "control" subjects. Both the Gardasil group and the AAHS control group had 2.3% systemic autoimmune disorders in the 6-month trial period — deemed "safe" because rates were equal. Siri argues AAHS itself caused the autoimmune disorders in both groups.

Merck separately labeled 320 subjects' carrier solution injection as "Saline Placebo" in the package insert. The actual contents: L-histidine, polysorbate 80, sodium borate, and yeast protein — not a placebo.

For injection-site reactions (minor), Merck's table separated the three groups. For serious systemic autoimmune conditions, Merck combined the two control groups to hide the AAHS signal.

Safety Review Durations

Standard observation periods (solicited/unsolicited reactions):

Compare to Pfizer's top 4 drugs (all placebo-controlled): Eliquis 7.4 years, Enbrel 6.6 years, Lipitor 4.9 years, Lyrica 2 years.

The CDC itself acknowledged: "Because the childhood immunization schedule is essentially a long-term exposure … long-term adverse events may be more biologically plausible than short-term events." The IOM noted: there could be "a relationship between [known] short-term adverse events following vaccination and long-term health issues." Yet trials almost never extend beyond 6 months.

What Proper Trials Find (When They Exist)

The dengue vaccine (Dengvaxia, Sanofi) is the only routine vaccine licensed based on a placebo-controlled trial with 5 years of safety review. That trial found children under 6 and seronegative children had significantly increased risk of severe harm and death from the vaccine — harms that would never have been detected in a standard short-window trial. The vaccine is now restricted to older, seropositive children only.

Moderna's RSV infant vaccine: used a placebo control → found more harm in vaccinated group → Moderna abandoned the vaccine. Without a placebo, this vaccine would likely have been licensed and deemed "safe."

IPV (IPOL): Deaths Occurred, Cannot Be Assessed

IPOL (Sanofi, 1990) was licensed with no control group and only 3 days of safety monitoring. IPOL contains immortal (chromosomally modified, cancer-like) monkey kidney cells in every vial. Sanofi reported in the package insert: "Although no causal relationship has been established, deaths have occurred in temporal association after vaccination of infants." Siri argues no causal relationship could possibly be established because the trial was designed to be unable to detect it.

MMR-II: Human DNA from Aborted Fetal Cell Line

MMR-II (Merck, 1978): 834 children, no control, 42 days monitoring. Approximately 1/3 of participants developed GI issues, 1/3 developed respiratory issues — dismissed without a control. Every vial of MMR-II contains, according to available evidence, hundreds of billions of pieces of human DNA and cellular material from a cultured aborted fetal cell line. CDC's Vaccine Information Sheet for MMR-II acknowledged "brain damage" as a potential outcome; after Stanley Plotkin's deposition, he worked to have "brain damage" removed from the VIS.

Pharma Decides Its Own Safety Outcomes

When adverse events occur in clinical trials, the determination of whether they are "related to the vaccine" is made by the pharma company's own paid researchers. Examples:

• ActHIB (Hib vaccine, Sanofi): 3.4% SAE within 30 days — "None was assessed by the investigators [Sanofi] as related to the study of vaccines" — case closed
• Gardasil: Women reporting severe brain and immune dysfunction symptoms after vaccination were told "This is not the kind of side effects we see with this vaccine" — discovered by an 8-month Slate investigation
• Maddie de Garay (Maddie de Garay, 12-year-old, Pfizer Covid-19 trial): Within hours of second dose, ER visit, ended up in wheelchair with feeding tube. Pfizer reported her injury to the FDA as "functional abdominal pain." After ICAN's firm and Steve Kirsch forced FDA to ask Pfizer for more information, Pfizer disclosed more injuries but its paid principal investigator (Dr. Robert Frenck, Cincinnati Children's) concluded he did not "feel" the injuries were vaccine-related. The FDA accepted this. Siri's letters to the FDA were ignored; ICAN sued FDA to obtain internal communications showing how FDA handled Maddie's case.

Significance

Siri argues this is not accidental design but rather the result of pharma's incentive structure. Because pharma companies are not liable for vaccine injuries post-licensure (due to the 1986 Act (National Childhood Vaccine Injury Act)), they have financial incentive to conduct as little safety review as possible to avoid findings that could prevent licensure. Vaccinologists like Plotkin, whose careers were funded by pharma, adopted this approach.

The window problem compounds over time: many serious chronic conditions (autoimmune disorders, neurological conditions) are not diagnosed until months or years of age. A 4-or-5-day safety window cannot detect them.

See Also

Post-Licensure Safety Monitoring, Plotkin Deposition 2018, IOM Vaccine Safety Report, Stanley Plotkin, Paul Offit, FDA, ICAN, Regulatory Capture, Financial Immunity for Vaccine Makers, 1986 Act (National Childhood Vaccine Injury Act), Maddie de Garay

Frequently Asked Questions

This is the plain-HTML rendering of concepts/Pre-Licensure Safety Testing.md, served for search engines, AI crawlers, and accessibility tools. The interactive 3D version of this wiki lives at vaccine-safety.org.