Meta Description: A balanced look at one of the world’s largest COVID-19 vaccine safety studies, including rare safety signals, what the findings mean, and why long-term monitoring remains important.

Five years after the early stages of the COVID-19 pandemic, researchers have far more information about vaccine safety than they did when vaccination programs first began.

One study has received particular attention because of its remarkable scale. Researchers working with the Global Vaccine Data Network (GVDN) analyzed health data from approximately 99 million vaccinated people across eight countries.

Published in the scientific journal Vaccine in 2024, the study examined whether certain medical conditions occurred more often than expected during defined periods after COVID-19 vaccination. The researchers evaluated 13 predefined adverse events of special interest involving neurological, blood-clotting, and cardiovascular conditions.

The findings deserve careful discussion. They identified several previously recognized vaccine safety signals, while also illustrating an important principle of medical research:

A safety signal is something that deserves investigation. It does not automatically mean that every medical event occurring after vaccination was caused by the vaccine.

Here is a calmer, evidence-focused look at what researchers found and what the results mean today.

From a Global Emergency to Long-Term Safety Research

When SARS-CoV-2 began spreading internationally in early 2020, healthcare systems faced an extraordinary challenge.

COVID-19 vaccines were subsequently developed and introduced on an unprecedented global scale. Because clinical trials cannot realistically include tens of millions of people, extremely uncommon medical events may only become detectable after a vaccine is used widely.

This is one reason post-vaccination surveillance is so important.

The Global COVID Vaccine Safety project was created to coordinate large-scale vaccine safety research using healthcare databases from multiple countries. According to GVDN, very large populations are particularly useful for studying serious events that occur only rarely.

What Did the 99-Million-Person Study Examine?

The multinational study included 99,068,901 vaccinated individuals from 10 sites in eight countries.

Researchers primarily examined three vaccines:

• Pfizer-BioNTech (BNT162b2)
• Moderna (mRNA-1273)
• Oxford-AstraZeneca (ChAdOx1)

The researchers compared the number of selected medical events observed after vaccination with the number that would normally have been expected based on background rates.

The primary analysis considered events occurring within 42 days following vaccination.

This distinction matters.

Conditions such as myocarditis, blood clots, or neurological disorders already occur naturally in the population. Simply discovering that someone developed a condition after receiving a vaccine does not establish causation.

Researchers therefore compare observed versus expected rates to determine whether an event is occurring more frequently than its usual background level.

What Safety Signals Were Identified?

The study confirmed several safety signals that had already been recognized through previous vaccine surveillance.

1. Myocarditis and Pericarditis

One of the most closely monitored findings involves inflammation affecting the heart.

Myocarditis refers to inflammation of the heart muscle, while pericarditis affects the tissue surrounding the heart.

The multinational analysis found higher-than-expected rates of myocarditis and pericarditis during certain risk periods following the vaccines studied.

These conditions have therefore remained an important focus of vaccine safety research.

Symptoms that can occur with myocarditis or pericarditis include chest discomfort, shortness of breath, or an unusually fast or irregular heartbeat.

Such symptoms can have many possible causes. Anyone experiencing significant or persistent symptoms should seek medical assessment rather than trying to determine the cause independently.

2. Guillain-Barré Syndrome

Researchers also observed a safety signal for Guillain-Barré syndrome (GBS) following the first dose of the AstraZeneca/ChAdOx1 vaccine.

GBS is a rare neurological disorder in which the immune system affects peripheral nerves. Symptoms may include weakness, tingling, or difficulty with movement.

In the study, the observed-to-expected ratio following the first ChAdOx1 dose was approximately 2.49.

Importantly, a relative increase does not tell us how common an event is in absolute terms. A condition that is uncommon at baseline may remain uncommon even when its relative rate increases.

3. Cerebral Venous Sinus Thrombosis

Another previously recognized signal involved cerebral venous sinus thrombosis (CVST) after the first ChAdOx1 dose.

CVST is an uncommon type of blood clot involving veins responsible for draining blood from the brain.

The researchers reported an observed-to-expected ratio of approximately 3.23 following the first ChAdOx1 dose.

Again, this figure describes a relative comparison with expected background rates. It should not be interpreted as meaning that CVST was common among vaccinated people.

4. Other Signals Requiring Further Investigation

Researchers also identified additional potential signals.

For example, acute disseminated encephalomyelitis (ADEM), a rare inflammatory neurological condition, produced a statistical signal following the first Moderna dose in the analysis.

However, the authors specifically emphasized that some newly identified signals require additional investigation before firm conclusions can be reached.

That distinction is essential when discussing large safety studies.

Finding an unexpected statistical pattern is often the beginning of a scientific investigation rather than the final answer.

What Does “Observed More Often Than Expected” Actually Mean?

This is perhaps the most important part of understanding the research.

Imagine that a particular condition normally occurs occasionally among millions of people every year.

If millions of people are vaccinated, some individuals will naturally develop that condition shortly afterward simply because unrelated medical events continue to occur.

Researchers therefore need to ask:

Did the condition occur more frequently after vaccination than would normally be expected?

GVDN researchers established background rates using healthcare information collected before widespread COVID-19 vaccination and compared them with events observed after vaccination.

When an event appears substantially more frequently than expected, it may create a safety signal.

Researchers can then investigate that signal using other study designs, biological evidence, and independent datasets.

Does an Adverse Event After Vaccination Prove the Vaccine Caused It?

No.

This is an important distinction in vaccine safety discussions.

An adverse event following immunization describes a medical problem occurring after vaccination. The timing alone does not necessarily demonstrate that vaccination caused the condition.

An adverse event of special interest, meanwhile, is a medically significant condition selected for careful monitoring because there may be an established, suspected, or theoretical relationship that deserves investigation.

Some vaccine-event relationships eventually become supported by strong evidence, while other statistical signals may weaken or disappear when researchers conduct more detailed studies.

This is why ongoing pharmacovigilance remains valuable.

Why Study Nearly 100 Million People?

Rare medical events are difficult to study.

If an event occurs only a few times among hundreds of thousands—or millions—of people, an ordinary clinical trial may simply be too small to detect a meaningful pattern.

Large international databases can provide much greater statistical power.

The GVDN study included hundreds of millions of administered doses across the participating databases, giving researchers an unusually large dataset for detecting uncommon safety signals.

This type of research does not make scientific uncertainty disappear, but it can reveal patterns that smaller studies may miss.

What the Study Does NOT Show

The findings are sometimes presented online in ways that go beyond what the researchers actually demonstrated.

The study does not show that everyone who received a COVID-19 vaccine will develop long-term health problems.

It also does not establish that every illness diagnosed after vaccination was caused by vaccination.

Most importantly, the research focused on specific predefined medical outcomes during defined post-vaccination risk periods. It should not be interpreted as proof of an unlimited range of illnesses appearing years later.

The authors described their findings primarily as confirmation of established safety signals and identification of several potential signals requiring additional investigation.

Why Continued Vaccine Safety Monitoring Matters

Medicine rarely stops collecting safety information once a treatment or vaccine reaches the public.

Large-scale monitoring can help researchers answer increasingly specific questions, such as:

Who is most susceptible to a particular rare reaction?

Age, sex, genetics, previous medical conditions, vaccine type, dose number, and previous infections may all potentially influence risk.

GVDN has also conducted research examining possible genetic contributions to uncommon events such as myocarditis/pericarditis, Guillain-Barré syndrome, and thrombosis with thrombocytopenia syndrome.

Understanding these differences could eventually help healthcare systems make vaccination recommendations even more individualized.

What If You Have Persistent Symptoms?

Anyone experiencing persistent or unexplained symptoms following vaccination—or following COVID-19 infection—should discuss them with a qualified healthcare professional.

Keeping a simple medical record can be helpful. Consider noting:

• vaccination dates and vaccine type, if known;
• when symptoms began;
• whether symptoms have changed over time;
• medications currently being taken;
• previous infections or medical conditions;
• test results and specialist reports.

These details can help a healthcare professional evaluate possible explanations.

Urgent symptoms such as severe chest pain, significant difficulty breathing, fainting, sudden neurological changes, or rapidly worsening weakness require prompt medical attention regardless of whether they occur after vaccination, infection, medication, or for no obvious reason.

A More Balanced Way to Discuss Vaccine Safety

COVID-19 vaccination became an unusually polarized subject.

But acknowledging rare adverse reactions does not require concluding that vaccines are universally dangerous. Likewise, recognizing the public-health benefits vaccination provided during the pandemic does not require pretending that adverse reactions never occurred.

Both ideas can coexist.

Large safety studies are valuable precisely because medicine should continue looking for uncommon risks even after millions of doses have been administered.

Transparent communication about those risks can help people make healthcare decisions based on evidence rather than fear.

The Bigger Picture

The 99-million-person GVDN analysis represents an important milestone in post-pandemic vaccine safety research.

It confirmed previously established safety signals involving myocarditis, pericarditis, Guillain-Barré syndrome, and cerebral venous sinus thrombosis in particular vaccine and dose settings. Researchers also identified other signals that require additional study.

At the same time, the findings need to be interpreted in context.

Rare events remain rare, statistical associations do not automatically establish causation, and risks can differ substantially according to vaccine type, dose, age, sex, and other individual factors.

The most useful lesson from this research may therefore be broader than any single vaccine:

Medical safety is not a one-time conclusion. It is an ongoing process of collecting evidence, identifying signals, investigating them carefully, and updating recommendations when better information becomes available.

That approach allows concerns to be taken seriously without turning uncertainty into alarm.

This article is for general educational purposes and is not intended to diagnose a medical condition or replace individualized advice from a qualified healthcare professional.