1. Introduction
The New England of Journal of Medicine Sounding Board article by Prasad and Makary, entitled “An Evidence-Based Approach to Covid-19 Vaccination” [1], outlines a significant shift in U.S. COVID-19 vaccine policy that continues to reverberate to today. The authors, serving as FDA Commissioner (Makary) and Director of the Center for Biologics Evaluation and Research (Prasad), advocate limiting COVID-19 vaccine access to individuals age 65 and older and those 18–64 years with high-risk conditions, citing as motivation for this policy declining public trust, low booster uptake, and the need for more robust data for broader approvals. While the idea of restoring trust and aligning policy with evidence is critical, the proposed framework raises concerns about the evidence base, consistency, lack of expert input and consensus, and implications for medical choice. This response critically evaluates their approach through four pillars: (1) the evidence base and risk-benefit considerations, (2) consistency of policy, (3) the value of content experts and consensus, and (4) preserving medical choice for those seeking immunization against COVID-19. We conclude with recommendations for a balanced path forward.
2. Pillar 1: evidence base and risk-benefit considerations
Prasad and Makary argue that the current “one-size-fits-all” approach to COVID-19 vaccination lacks evidence to justify use in lower-risk groups such as healthy children and younger adults. They propose a risk-based framework, prioritizing vaccine approvals for high-risk groups (e.g., those over 65 or younger persons with comorbidities) while demanding “gold-standard” randomized controlled trial (RCT) data for all others. Their justification for this is apparently grounded in their observation of declining booster uptake (e.g., only 22% of eligible Americans received the 2024–2025 booster) and waning public trust in vaccines rather than in concerns over actual data suggesting new benefits or risks. It also ignores two key observations: first, the vaccine has been used for multiple years with millions of doses given and hence robust safety data and real-world evidence are available, and second, the de facto realization that the balance of risks and benefits are such that the vaccine is acceptable in their proposed policy for all those age 65 and older and those age 18 and older with risk factors. This begs the question – what unique safety concerns are there in younger age groups (and pregnant women, and persons with Long COVID) that justify limiting vaccine access and choice?
While the call for data is laudable, the framework's evidence base is selective and overlooks several considerations. First, the authors cite low booster uptake as evidence of public skepticism but fail to address structural barriers, such as access disparities, vaccine fatigue [2], mis- and dis-information, conflicting policy recommendations and confusion, and politicization of policy which may contribute more significantly than distrust in vaccine performance per se. In addition, restricting access exacerbates inequities without addressing their root causes. Confusion now reigns for providers and patients – raising questions such as what about younger children who have never been immunized – perhaps in the face of a newer more transmissible or severe variant – will we start all over with demands for RCTs and develop new public policy? Why is “physical inactivity” an acceptable risk factor, but the presence of Long COVID isn't in their schema? These illustrate internal inconsistencies that would be avoided by having a broader group of content experts involved in shaping an evidence-based public policy.
Second, the risk-benefit analysis for moderate- and low-risk groups is not as clear-cut as the authors suggest. While severe outcomes (e.g., hospitalization, death) are certainly less common in healthy younger individuals; Long COVID, COVID-related complications, and milder infections still impose significant health, school, work, and economic burdens (much like influenza does). A 2024 meta-analysis estimated that 10–30% of non-hospitalized COVID-19 cases result in Long COVID symptoms, impacting quality of life and productivity [3]. By limiting access to COVID-19 vaccines, this policy risks undermining larger-scale immunity, particularly in settings like schools, where children can amplify community spread – particularly of newer highly transmissible variants – to the detriment of others at higher risk.
The demand for RCT data for variant-changed vaccines in low-risk groups is also impractical. RCTs are resource-intensive and simply are not feasible for rapidly evolving variants. By the time an RCT could be planned and conducted, the rapid evolution of new variant strains would make the data “obsolete” (as we have observed during the course of the pandemic). Real-world evidence, such as observational studies conducted in Israel and the UK, has demonstrated vaccine effectiveness in reducing severe outcomes across age groups, even for newer variants [4], [5], [6]. Dismissing such data in favor of requiring RCTs delays access to vaccines unnecessarily, especially when the safety profile of mRNA and protein subunit vaccines are well-established, with adverse events (e.g., myocarditis) occurring at rates along the lines of 28–38 per million due to mRNA vaccines, and myocarditis rates due to COVID-19 infection estimated at 7× higher that that related to vaccination [7]. There are also difficult ethical concerns that bear more detailed investigation. A more balanced approach would integrate real-world data with targeted RCTs to refine risk-benefit profiles without blanket restrictions.
