What were the methods and sample size of the Cleveland Clinic flu vaccine study?
Executive summary
The Cleveland Clinic study was a prospective cohort of 53,402 Cleveland Clinic employees followed from Oct. 1, 2024, through late March 2025; 82.1% received a 2024–25 seasonal influenza vaccine during the study period and overall 2.02% tested positive for influenza, producing a reported vaccine effectiveness of −26.9% (i.e., a higher incidence in vaccinated staff) in the authors’ analysis [1]. The paper appeared as a medRxiv preprint and has prompted conflicting interpretations and critiques about methods, bias and causal claims [2] [3] [4] [5].
1. What the study actually did: prospective cohort of employees
The study enrolled Cleveland Clinic employees who were employed in Ohio on Oct. 1, 2024, and followed them through the respiratory viral season to compare cumulative incidence of laboratory-confirmed influenza in vaccinated versus unvaccinated states over the study period [2]. The public summaries and advocates note the cohort size as 53,402 employees and a study window ending in late March 2025 [1].
2. Sample size and vaccination uptake — big cohort, high vaccine coverage
The sample comprised 53,402 working-aged Cleveland Clinic employees; 82.1% received the seasonal influenza vaccine at some point during follow-up, leaving a minority unvaccinated for comparison [1]. That large N gives statistical power but also produces an imbalance between groups that matters for interpretation [1].
3. Outcomes measured and headline result
Investigators used laboratory-confirmed influenza as the outcome and reported that only 2.02% of participants were infected during the season; vaccinated staff had a higher cumulative incidence, yielding a calculated vaccine effectiveness of −26.9% in the authors’ analyses [1]. The authors’ formal conclusion framed this as an association: vaccination “was associated with a higher risk of influenza” during the 2024–25 season [2].
4. Why the finding has been controversial — preprint status and interpretation
The paper was posted as a medRxiv preprint and had not completed peer review when widely shared; communicators and fact-checkers warn that association is not causation and that preprints can change after review [2] [4]. Media and social posts amplified a causal-sounding headline that the shot “increases” flu risk; Lead Stories and the Cleveland Clinic spokesperson both stressed the study did not prove vaccination caused more infections [4].
5. Methodological critiques flagged by clinicians and journalists
Critics including MedPage Today’s commentator argued the paper has methodological flaws severe enough to undermine its conclusions, calling attention to potential biases in testing and analysis that the authors did not resolve in the preprint [5]. Public-health communicators similarly cautioned that testing behavior, exposure differences, and other confounders could explain an observed negative effectiveness without proving harm from vaccination [6].
6. Alternative explanations emphasized by experts
Independent experts and the Cleveland Clinic noted testing bias (vaccinated employees more likely to seek testing or be tested), differential exposure risk, and other unmeasured confounders as plausible reasons for the observed association; these interpretations aim to reconcile the data with longstanding evidence that flu vaccines reduce severe outcomes even when strain match is imperfect [3] [4] [6].
7. How to read the numbers — what they do and don’t show
The raw numbers (53,402 people; 82.1% vaccinated; 2.02% infected; VE −26.9%) are descriptive of this cohort and the season studied, but they do not on their own establish that the vaccine caused increased infections; the authors presented an observational association and media reports sometimes conflated that association with causation [1] [2] [4].
8. Practical context and public-health perspective
Public-health communicators and fact-checkers who reviewed the preprint urged caution: the finding alone is not grounds to change vaccination policy and should be weighed against broader evidence on vaccines’ role in reducing severe illness and death; they also flagged the risk of misinformation when preprints are amplified without methodological scrutiny [1] [6] [3].
Limitations and what’s not in current reporting
Available sources document methods, sample size, vaccination uptake and headline results but do not present the full, peer-reviewed methods appendix or any post–peer-review revisions; they do not provide detailed adjusted-model specifications or all sensitivity analyses from the final paper as published in a journal—those details are not found in current reporting [2] [1] [5].
Bottom line
The Cleveland Clinic preprint tracked 53,402 employees (82.1% vaccinated) and reported a small overall infection rate (2.02%) with a calculated VE of −26.9%; the result is an observational association that prompted scrutiny over testing bias and confounding and should not be read as definitive evidence that the flu shot increased infections without further peer-reviewed analysis [1] [2] [4] [5].