Do Vaccines Cause Harmful Injuries to Children? The Evidence Shows Rare, Real Risks — Not Widespread Harm

The claim, as it circulates most forcefully, is that vaccines cause serious, widespread harm to children — most prominently autism, immune damage, and sudden death. The verdict is partially false. A small number of genuine adverse events are real, acknowledged, and actively monitored by regulators. The broader narrative of systemic, large-scale harm is not supported by the weight of scientific evidence across thousands of studies.

The clearest measure of confirmed vaccine injury comes from the National Vaccine Injury Compensation Program (VICP), tracked by HRSA through 2023. Since 1988, VICP has compensated approximately 10,000 petitioners across roughly 4.4 billion U.S. vaccine doses — a compensation rate of about 1 per 440,000 doses. These are cases where causation was accepted legally and medically. Separately, a 2014 systematic review by Maglione et al. in Pediatrics, covering 20,478 studies, confirmed specific serious harms: febrile seizures from MMR and MMRV vaccines, and anaphylaxis at approximately 1–2 per million doses. The CDC's Vaccine Safety Datalink, an active surveillance system covering over 12 million people, has also confirmed myocarditis following mRNA COVID vaccines in adolescent males at roughly 11–139 per million doses depending on age, sex, and dose number. These risks are not hidden — they are published, monitored, and factored into clinical guidance.

The steelman version of the claim points to VAERS, the CDC/FDA passive reporting system, which has logged approximately 1.6 million reports since 1990. If you treat every VAERS report as a confirmed injury, the numbers look alarming. But the CDC explicitly states that a VAERS report 'does not prove that the vaccine caused the adverse event.' VAERS is designed to detect signals worth investigating, not to establish causation. Anyone can file a report, and reports include events that happened after vaccination regardless of whether the vaccine was responsible. Using raw VAERS counts as proof of harm is the central analytical error driving the most extreme versions of this claim.

The most damaging specific allegation — that MMR vaccine causes autism — has been investigated exhaustively and rejected. Taylor et al., published in The Lancet in 1999, studied 498 children and found no association between MMR and autism onset. The 2011 Institute of Medicine report reviewed over 1,000 studies and found no causal link between vaccines and autism, SIDS, or immune dysfunction. The original 1998 Wakefield paper that launched this fear was retracted by The Lancet in 2010 after investigators found ethical violations and data manipulation. It is fair to concede that temporal coincidence is real: autism symptoms often become apparent around the same age children receive vaccines, which makes the association feel intuitive. That is correlation, not causation, and decades of research have confirmed the distinction.

The benefit side of the ledger is not abstract. The WHO estimates that vaccines prevent 3.5 to 5 million deaths per year globally. Against a confirmed serious adverse event rate of fewer than 1–2 per million doses for most vaccines, the benefit-risk calculation is not close. The manipulation pattern to watch for is this: take a passive reporting system not designed to prove causation, strip away the denominator of billions of doses, amplify with emotionally resonant individual stories, and present the result as evidence of a cover-up. When you see VAERS numbers cited without the caveat that they do not establish causation, or injury counts cited without the 4.4 billion dose denominator, you are watching that pattern in action.