Michael A. Stoto

Evidence-based medicine (EBM) was developed to counter tradition, authority, and anecdote by grounding clinical decisions in the best available research. Yet in debates about COVID-19 and childhood vaccination, some of EBM's most visible champions have invoked its name to weaken, rather than strengthen, public health. This article examines how a narrow, randomized-trial-only interpretation of EBM has been used to cast doubt on vaccines, to delegitimize long-standing regulatory and advisory processes, and to justify sweeping policy changes that lack evidentiary support. I argue that these moves betray EBM's foundational principles, align with Health and Human Services Secretary Robert F. Kennedy Jr.'s broader effort to undermine vaccines, and risk eroding public trust in both science and public health. I then outline what a genuinely evidence-informed approach to vaccine policy should look like and why preserving the existing U.S. vaccine regulatory and advisory system is essential. Citing standards of evidence-based medicine (EBM), Vinay Prasad and Marty Makary announced early in their tenure at the Food and Drug Administration (FDA) that randomized clinical trials (RCTs) would be required for new COVID-19 vaccines. Prasad subsequently called for stricter regulation of all vaccines based on case reports (not RCTs) of ten, or maybe even zero, deaths in children who received a COVID-19 vaccine. Then, in January 2026 the Centers for Disease Control and Prevention (CDC) issued sweeping changes in its childhood vaccine recommendations, based on no evidence at all. Together, these actions reveal a striking inversion: the rhetoric of evidence-based medicine is being used selectively to question vaccine benefits while accepting weak or nonexistent evidence for vaccine harms. Beyond the dubious rationale and vast consequences of these changes for vaccine access and child health, some EBM advocates have become part of a project that casts doubts on the benefits and exaggerates the harms of vaccines and other public health interventions. Beyond the consequences for U.S. vaccine policy these actions undermine trust in public health in general. Understanding how EBM leaders have come to support Health and Human Services Secretary Robert F. Kennedy, Jr.'s actions that undermine public health is the first step towards true evidence-informed policy. Ironically, restoring confidence will require preserving the well-developed vaccine regulatory and advisory system Kennedy seems determined to dismantle, which I will show is actually a crown jewel of evidence-informed public health.

Ioannidis and colleagues' estimate that COVID-19 vaccines averted 2.5 million deaths globally through October 2024 is substantially lower than other estimates. Our analysis suggests that the authors' estimates of vaccine impacts are lower than others because of dubious model assumptions. These choices include model parameters, particularly for vaccine effectiveness; the infection fatality rate, including how it varies among countries and over time; and the use of reported COVID-19 deaths rather than excess mortality estimates. The lower estimates are also due to simplifications in model structure such as ignoring the impact of vaccines on transmission and of heterogeneity within and among countries. Other benefits, including reduced hospitalizations and non-health benefits to vaccinated individuals and the community, are harder to quantify but clearly substantial. Thus, the estimate of 2.5 million deaths averted markedly underestimates the benefit of the global COVID-19 vaccination program. In addition, the authors' results are presented in a way that deemphasizes vaccine benefits and overemphasizes harms, perhaps by a factor of 10. Historical analyses can be enlightening, but future decisions about vaccine recommendations or mandates will require a more comprehensive comparison of a vaccine's likely health and other benefits to the potential costs and harms, to the community as well as individuals, taking into account the current epidemiological situation.

There is no comprehensive synthesis of supportive care interventions for patients with hematologic malignancies receiving transplantation and cellular therapies (TCTs). This systematic review and meta-analysis aimed to characterize randomized controlled trials (RCTs) of TCT supportive care interventions and examine effects on patient-reported outcomes (PROs). We systematically searched MEDLINE, Embase, CINAHL, PsycInfo, Web of Science, and Cochrane Central Register of Controlled Trials. Risk of bias was assessed using Cochrane Risk of Bias 2.0. Findings were grouped by intervention and PRO type and synthesized narratively, with exploratory random-effects meta-analyses estimating pooled intervention effects on anxiety, depression, and quality of life (QOL). Fifty-three RCTs (1992-2025) met inclusion criteria. The search yielded only RCTs of hematopoietic stem cell transplantation (HSCT) recipients, although we included all transplantation and cellular therapies. Studies varied in sample size (11-711), participant demographics (mean age, 33-63 years; 22%-64% women), and PROs. There were 9 intervention categories: cognitive behavioral (n = 5), exercise only (n = 17), expressive helping (n = 2), mind-body/stress management (n = 7), multimodal (exercise plus another modality; n = 6), music/art (n = 12), palliative care (n = 2), positive psychology (n = 1), and self-management/self-efficacy (n = 5). Interventions were mostly feasible and acceptable, with promising improvements in psychological distress, QOL, and physical outcomes. Meta-analysis showed nonsignificant trends toward exercise interventions improving QOL (standardized mean differences [SMD], 0.09) and music/art interventions reducing acute depressive symptoms (SMD, −0.26). Among patients undergoing HSCT for hematologic malignancies, supportive care interventions show feasibility, acceptability, and promise for improving PROs. However, larger, rigorously designed RCTs using standardized PRO measures are needed to establish efficacy.

Seroprevalence surveys reveal the extent of humoral immunity against pathogens such as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and under some circumstances represent cumulative incidence of prior infection. However, antibody waning - or seroreversion - biases these estimates by reducing assay sensitivity in a time-varying manner. Because assay sensitivity decays over time, naively using serosurveys can substantially bias estimates of SARS-CoV-2 cumulative incidence and fatality rates. The Bayesian assay-specific, time-varying sensitivity adjustment developed in this paper can reliably correct for this bias and account for the delay between infection and serosurvey. In seroprevalence studies conducted in the United States in 2020, adjusting for time-varying sensitivity increased cumulative incidence by up to 1.4-fold, with an adjustment of 1.08 for a national study. Our estimates contrast with a previously published 2-fold adjustment that did not account for assay design. This suggests that previous analyses overestimated cumulative incidence by applying seroreversion corrections that did not account for assay-specific effects, or underestimated cumulative incidence by not applying seroreversion corrections. These biases imply fatality rate underestimation and overestimation, respectively. Our model provides a framework for design-specific time-varying sensitivity corrections in seroprevalence surveys for other pathogens.

Italy implemented various measures, including lockdowns and a mass vaccination campaign, to address the COVID-19 pandemic. This study aims to describe the temporal and regional differences in mortality trends between March 2020 and December 2021, along with associated socioeconomic, policy, and behavioral factors. We used National Ministry of Health data on COVID-19 mortality, excess mortality (EM), and vaccine uptake, along with data from the Italian arm of a European survey of preventive behaviors and attitudes, such as trust in institutions. The analysis was conducted across four macro regions and five study periods. Avertable mortality was calculated as observed EM minus the lowest EM at the macro-regional level for each study period. In 2020–21, the estimated total EM was 180,169 deaths, with 76.4% officially attributed to COVID-19. This proportion ranged from 13.5% in the South and Islands (June–October 2020) to 140.0% in the Northeast (March–July 2021). Excess and avertable mortality peaked in the North during the first two periods (March 2020 – February 2021) and in the South and Islands thereafter (March–December 2021). Survey data revealed reduced adherence to stay-at-home orders in the North and lower trust in hospitals and reduced vaccine uptake, especially among the elderly, in the South and Islands. After the initial period, 33,587 deaths (18.6%) could have been averted if each macro-region had matched the lowest rates observed in that period. An estimated 40.7% of avertable deaths occurred in Southern and Insular Italy, which constitutes 33.7% of the national population. Due to differential misreporting, EM estimates offer a more accurate view of regional and temporal patterns in COVID-19 mortality than official rates. The higher EM in northern Italy during the first year of the pandemic might be linked to lesser adherence to control policies, possibly associated with higher private-sector employment. The higher EM in the South and Islands post-March 2021, contributing to 40.7% of avertable EM, could be partly explained by the lower vaccination rates in the population aged 80 and older, who experienced the highest age-specific mortality rates and lower trust in the healthcare system in this macro-region.

Background Italy implemented various measures, including lockdowns and a mass vaccination campaign, to address the COVID-19 pandemic. This study aims to describe the temporal and regional impact of these measures on mortality between March 2020 and December 2021, partitioning the time series into five periods. Methods We used National Ministry of Health data on COVID-19 mortality, excess mortality (EM), and vaccine uptake, along with survey data. Avertable mortality was calculated as observed EM minus the lowest EM at the macro-regional level for each study period. Results In 2020–21, the estimated total EM was 180,169 deaths, with 76.4% officially attributed to COVID-19. This ranged from 13.5% in the South and Islands (June–October 2020) to 140.0% in the Northeast (March–July 2021). Excess and avertable mortality peaked in the North during the first two periods (March 2020 – February 2021) and in the South and Islands thereafter (March–December 2021). Survey data revealed reduced adherence to stay-at-home orders in the North and lower trust in hospitals and reduced vaccine uptake, especially among the elderly, in the South and Islands. After the initial period, 33,587 excess deaths (18.6%) could have been averted if each macro-region had matched the lowest rates observed post-June 2020. An estimated 40.7% of avertable deaths occurred in Southern and Insular Italy, which constitutes 33.7% of the national population. Conclusions Due to differential under-reporting, EM estimates offer a comprehensive view of regional and temporal patterns in COVID-19 mortality. The higher EM in northern Italy during the first year of the pandemic might be linked to lesser adherence to control policies, possibly associated with higher private sector employment. The higher EM in the South and Islands post-March 2021, contributing to 40.7% of avertable EM, could be partly explained by the lower vaccination rates in the population age 80 and older, who experienced the highest age-specific mortality rates and lower trust in the healthcare system in this macro-region.

In 2021 the “Red Covid,” narrative emerged: states and counties that are predominantly Republican and/or supported Trump experienced widespread vaccine hesitancy and eventually opposed mask and vaccine mandates. Our analysis of excess mortality quantifies this effect. After the first wave, death rates in the South were more than double those in the Northeast; 45% of deaths were in the South, with 38% of the population. Although there have been many correlational studies of this phenomenon, our analysis uses excess mortality estimates, aggregated by Census region and six pandemic phases, to establish the magnitude of this effect. We show that, if every region had the same mortality rate as the lowest regional rate in each period, more than 418,763 COVID-19 deaths were in this sense "avoidable." The analysis also shows that the U.S. has been a divided nation at almost since the start of the pandemic. We estimate that 198,642 deaths could have been averted during the Summer of 2020 and the Alpha wave, before vaccines were widely available. This strongly suggests that implementation of and adherence to NPIs made a difference. Similarly, while vaccination is a personal choice, as many Republicans argue, our estimates of 221,360 deaths averted after vaccines became available demonstrate the population-level consequences of that choice. Beyond the substantive results, this analysis demonstrates the importance of excess mortality estimates as part of the public health surveillance toolkit. Official mortality counts rely on complete recording of COVID-19 as a cause of death, COVID-19 deaths are under reported for many reasons. Between Jan. 3, 2020 and April 30, 2022, public health officials reported 978,567 COVID-19 deaths in the U.S. According to our calculations, however, the excess mortality associated with COVID-19 totaled 1,335,292, during that period, 37% more COVID-19 deaths than reported. Moreover, the proportion of COVID-19 deaths reported as such varied markedly over time and region, and reporting was generally less complete in the regions and periods with the highest mortality rates. Analyses of recorded COVID-19 deaths, just as of cases, therefore, would attenuate regional differences.

Background: Italy responded to COVID-19 with stringent lockdowns and a mass vaccination campaign. We describe the temporal and regional impact of these measures on mortality, along with social-behavioral correlates. Methods: The analysis uses National Ministry of Health data on COVID-19 mortality (CM), excess mortality (EM), and vaccine uptake, along with survey data. We calculate avoidable mortality (AM) as observed mortality minus the lowest regional EM rate in each period. Results: In 2020-21 EM was 180,169, of which 76% were counted as COVID-19. This fraction varied from 13% in the South in Jun-Oct 2020 to 140% in the Northeast in Mar-Jul 2021. Impact varied across Italy, with EM and AM highest in the North during the first and second waves (Feb-Jun 2020 & Nov 2020-Feb 2021) and in the South afterwards (Mar-Dec 2021). After the first wave, there were 33,587 avoidable deaths, 41% of which were in the South. Higher rates in the North in Nov 2020-Feb 2021 were possibly due to less social distancing (e.g. avoiding social events and staying home from work/school), perhaps because residents of the North are more likely to be employed in private industry, which were more likely than public agencies to remain open. Higher rates in the South after Mar 2021 were possibly due to lower trust in hospitals and reduced vaccine uptake, particularly among those aged 80+. Conclusions: Official CM was differentially under-reported, so EM estimates provide a more robust picture of regional and temporal patterns of COVID-19 deaths. Mortality was high in the North in the second as well as first wave. It was high in the South afterwards (Mar-Dec 2021), accounting for 41% of avoidable mortality. Key messages: • Excess mortality estimates provide a more robust understanding of regional and temporal patterns of COVID-19 deaths than reported COVID-19 deaths. • Heterogeneous impact of COVID-19 was associated with differences in preventive habits, trust levels, and vaccine uptake.

A number of scientific publications and commentaries have suggested that standard preparedness indices such as the Global Health Security Index (GHSI) and Joint External Evaluation (JEE) scores did not predict COVID-19 outcomes. To some, the failure of these metrics to be predictive demonstrates the need for a fundamental reassessment which better aligns preparedness measurement with operational capacities in real-world stress situations, including the points at which coordination structures and decision-making may fail. There are, however, several reasons why these instruments should not be so easily rejected as preparedness measures.From a methodological point of view, these studies use relatively simple outcome measures, mostly based on cumulative numbers of cases and deaths at a fixed point of time. A country’s “success” in dealing with the pandemic is highly multidimensional – both in the health outcomes and type and timing of interventions and policies – is too complex to represent with a single number. In addition, the comparability of mortality data over time and among jurisdictions is questionable due to highly variable completeness and representativeness. Furthermore, the analyses use a cross-sectional design, which is poorly suited for evaluating the impact of interventions, especially for COVID-19.Conceptually, a major reason that current preparedness measures fail to predict pandemic outcomes is that they do not adequately capture variations in the presence of effective political leadership needed to activate and implement existing system, instill confidence in the government’s response; or background levels of interpersonal trust and trust in government institutions and country ability needed to mount fast and adaptable responses. These factors are crucial; capacity alone is insufficient if that capacity is not effectively leveraged. However, preparedness metrics are intended to identify gaps that countries must fill. As important as effective political leadership and trust in institutions, countries cannot be held accountable to one another for having good political leadership or trust in institutions. Therefore, JEE scores, the GHSI, and similar metrics can be useful tools for identifying critical gaps in capacities and capabilities that are necessary but not sufficient for an effective pandemic response.

Background: COVID-19 pandemic provides a unique opportunity to learn the challenges encountered by public health emergency preparedness systems, both in terms of problems encountered and adaptations during and after the first wave, as well as successful responses to them. Results: This work draws on published literature, interviews with countries and institutional documents as part of a European Centre for Disease Prevention and Control project that aims to identify the implications for preparedness measurement derived from COVID-19 pandemic experience in order to advance future preparedness efforts in European Union member states. The analysis focused on testing and surveillance themes and five countries were considered, namely Italy, Germany, Finland, Spain and Croatia. Our analysis shown that a country’s ability to conduct testing at scale was critical, especially early in the pandemic, and the inability to scale up testing operations created critical issues for public health operations such as contact tracing. Countries were required to develop new strategies, approaches, and policies under pressure and to review and revise them as the pandemic evolved, also considering that public health systems operate at the national, regional, and local level with respect to testing, contact tracing, and surveillance, and involve both government agencies as well as private organizations. Therefore, communication among multiple public and private entities at all levels and coordination of the testing and surveillance activities was critical. Conclusion: With regard to testing and surveillance, three capabilities that were essential to the COVID-19 response in the first phase, and presumably in other public health emergencies: the ability to scale-up testing, contact tracing, surveillance efforts; flexibility to develop new strategies, approaches, and policies under pressure and to review and revise them as the pandemic evolved; and the ability to coordinate and communicate in complex public health systems that operate at the national, regional, and local level with respect and involve multiple government agencies as well as private organizations.