What peer-reviewed studies track long-term safety of COVID-19 vaccines over 3–5+ years?

1. What counts as “long‑term” safety evidence — and why randomized 5‑year trials are scarce

Randomized pre‑licensure trials for COVID‑19 vaccines were large but compressed in time; regulators and academics have relied on post‑authorization safety studies and surveillance to capture later outcomes. Post‑marketing cohort studies, self‑controlled case series, and networked database projects are the standard method to detect rare or delayed adverse events over years — not new randomized trials running five years — and investigators have published multi‑year surveillance analyses such as the UK AZD1222 active‑surveillance study covering March 2021–April 2023 (published 2025) ^[1]. The literature and the public health community therefore treat ongoing surveillance systems as the primary source of long‑term safety data ^[3].

2. Representative peer‑reviewed, multi‑year surveillance studies and analyses

Examples in the recent record include the BMJ Open publication of a UK post‑authorization active‑surveillance analysis of AZD1222 spanning 2021–2023 (published online May 2025) that reported common adverse events like headache and fatigue and monitored events of special interest such as anosmia and ageusia in real‑world use ^[1]. Other peer‑reviewed population studies and journal articles across settings (e.g., cohort and self‑controlled case series work in Japan, Scotland studies on pediatric vaccination) have addressed safety and effectiveness across multi‑year timeframes or through successive seasons, although many results appear in surveillance reports or specialty journals rather than as single long randomized trials ^{[4] [5]}.

3. Large safety monitoring networks and administrative‑data studies doing the heavy lifting

Long‑term safety signals have largely come from continuous monitoring systems and large administrative cohorts. The Vaccine Safety Datalink (VSD) is an established collaborative system that conducts adverse‑event studies and has published incidence data, including Guillain‑Barré Syndrome analyses after COVID‑19 vaccination ^[3]. Analyses of Medicare beneficiaries (covering tens of millions of older adults) have been used to monitor 2023–2024 vaccine formulas and multiple health outcomes, reporting on millions of recipients in 2024 surveillance work ^[2]. These systems produce iterative, peer‑reviewed outputs rather than single 5‑year trial publications.

4. Projects aiming to be global and long‑term — some disrupted

Ambitious multi‑year international projects were created to pool billions of person‑years of data; at least one landmark global safety study faced funding cuts and was terminated before completion, undercutting a planned five‑year effort to evaluate COVID‑19 vaccine safety across huge populations ^[6]. That interruption helps explain gaps in having neat, peer‑reviewed 5‑year single‑study answers for every vaccine formulation.

5. What recent reviews and guidelines conclude about long‑term safety

Systematic reviews and guideline panels in 2025 assess the balance of benefit and risk using accumulated surveillance evidence and typically find that benefits outweigh risks for recommended groups; for example, IDSA guidance cited “little or no serious adverse effects” in supporting 2025–26 vaccine recommendations for immunocompromised patients ^[7]. CIDRAP’s synthesis and regulatory labeling changes (e.g., FDA adding Guillain‑Barré Syndrome warnings in early 2025) show the system updating recommendations as multi‑year evidence accumulates ^[8].

6. Disagreements, open questions, and areas needing more peer‑reviewed follow‑up

Controversies remain: small preliminary papers and preprints have suggested rare, lingering symptom clusters after vaccination and prompted calls for further research; these papers are contested and characterized as preliminary by reporting outlets ^[9]. Some population studies have flagged specific signals (e.g., pulmonary embolism in subgroup analyses) that warrant further scrutiny and confirmatory follow‑up in larger datasets ^[5]. Available sources do not mention a single, comprehensive set of peer‑reviewed randomized trials that follow vaccine recipients for a continuous 5+ years; instead, evidence accrues from multiple surveillance cohorts, VSD and Medicare analyses, and national studies ^{[3] [2] [1]}.

7. How to locate the best current peer‑reviewed long‑term evidence

To find peer‑reviewed multi‑year safety work: search for post‑authorization cohort studies and self‑controlled case series in major journals (e.g., BMJ Open, J Glob Health), monitor VSD and Medicare publications, and watch multi‑year network projects (some of which may be delayed or curtailed) — cite the AZD1222 UK active‑surveillance paper (BMJ Open, 2025) and recent Medicare surveillance preprint as entry points ^{[1] [2]}. For policy context and evolving safety labels, consult guideline syntheses such as CIDRAP and professional society guidance ^{[8] [7]}.

Limitations: Coverage is based on available reporting through 2025; many surveillance datasets produce ongoing analyses rather than single 5‑year papers, and major global projects have been disrupted, leaving gaps that future peer‑reviewed publications will need to fill ^{[6] [5]}.