How do VAERS reports get validated and what percentage have established causality with vaccines?
Executive summary
VAERS is a national, passive early‑warning system that accepts any report of an adverse event after vaccination but does not by itself establish cause and effect, and CDC/FDA follow-up and external studies are used to validate signals [1] [2] [3]. In published adjudications of sampled VAERS reports, expert reviewers were able to classify causality in most cases, and "probable" or "definite" causality was found in under one quarter of investigated reports, while most reports are judged unrelated or unlikely to be caused by vaccination [4] [5] [6].
1. How VAERS collects raw reports and why that matters
VAERS is intentionally inclusive: anyone — patients, parents, caregivers, or clinicians — may file a report and the system accepts events regardless of whether the reporter thinks the vaccine caused the problem, which means reports document temporal association not proof of causation [2] [1]. As a passive surveillance system it relies on voluntary submission and therefore is subject to both under‑reporting and reporting biases (more likely to capture severe events and events that attract public attention), so the raw counts are unsuitable for causal inference without further analysis [7] [8].
2. The first layer of validation: triage, follow‑up, and clinical review
CDC and FDA routinely triage VAERS reports and follow up on all serious events to obtain medical records, laboratory data, death certificates, and autopsy reports when available; these documents inform clinical review but do not automatically convert a report into a confirmed vaccine reaction in the public VAERS dataset [8] [9] [10]. Expert clinicians at CDC/FDA examine narratives and documentation; in some clear, biologically plausible cases (for example, expected local reactions) the evidence may be straightforward, but for most reports additional investigation is needed [11] [4].
3. Structured causality assessment and what studies show
Researchers have applied structured, WHO‑based causality frameworks and multi‑reviewer adjudication to VAERS reports: one research series found that causality could be classified in the majority (about 97%) of reviewed reports, but most of those were deemed unrelated or unlikely to be vaccine‑caused [4]. Another peer‑reviewed study using expert panels reported that "probable" or "definite" causality was present in fewer than one quarter of evaluated VAERS reports, and these probable cases tended to be mild, expected vaccine reactions or allergic responses [5] [6].
4. When VAERS signals trigger epidemiologic studies
VAERS is principally a signal detection tool: disproportionate reporting or clinical patterns can prompt formal epidemiologic investigations using independent data sources (claims databases, medical records, active surveillance) to test whether an observed association is causal and to estimate rates — a process CDC and FDA have followed historically for concerns such as febrile seizures after influenza vaccine [11] [3]. Only these analytic follow‑up studies, not the raw VAERS counts, can robustly establish population‑level risk and causality for specific events [3] [12].
5. Quantifying "what percentage" have established causality — limits and best evidence
There is no single, system‑wide percentage reported by VAERS that quantifies how many of all submitted reports are eventually confirmed as vaccine‑caused; however, sampled adjudication studies provide pragmatic estimates: in reviewed samples, definite/probable causality was under 25% and most classified reports were unrelated or unlikely to be causal [5] [4] [6]. It is important to stress that these figures derive from specific study samples and adjudication methods rather than a comprehensive accounting of all VAERS records, and VAERS itself cautions that its public data do not infer causality and change as follow‑up information is received [10] [12].
6. Alternative viewpoints, caveats and hidden agendas
Patient advocates and some clinicians argue that passive systems like VAERS undercount real vaccine‑induced harms and that stronger incentives or mandatory reporting would improve detection, while regulators emphasize the complementary network of active surveillance and epidemiologic studies to avoid false signals that can harm public confidence [7] [11]. Meanwhile, public misunderstanding arises when raw VAERS tallies are portrayed as proof of harm; VAERS administrators and scientific papers explicitly warn against that misuse, noting that the system’s strength is sensitivity for signal detection, not causal proof [1] [11].