On June 9, about 300 employees of the National Institutes of Health, representing all 27 NIH institutes, issued a letter they called the “Bethesda Declaration,” calling on NIH director Jay Bhattacharya, MD PhD, to reverse the Trump administration’s course on multiple initiatives.
The declaration was explicitly inspired by the October 2020 “Great Barrington Declaration,” co-authored by Bhattacharya, an epidemiologist who was then a professor of health policy at Stanford, and signed at the American Institute for Economic Research’s Great Barrington campus. That declaration asked public health authorities to redirect anti-COVID efforts away from mandatory school and workplace closures, since children and healthy adults were at low risk of death, and focus efforts on protecting the elderly and chronically ill. For voicing this perspective, its authors (also including Dr. Martin Kulldorff, biostatistician and professor of medicine at Harvard, and Dr. Sunetra Gupta, epidemiologist and professor at Oxford) were blacklisted on Twitter (now X) and personally targeted for attack by federal health officials.
The authors of the Bethesda Declaration call on Bhattacharya to pay more than lip service to “establishing a culture of respect for free speech and dissent,” affirm that “dissent is the very essence of science,” and ensure that dissenting voices are “heard and allowed.” They accuse the Trump administration of politicizing research; cutting funding for research on health disparities, COVID immunization, health impacts of climate change, and the health needs of diverse populations; undermining global collaboration, peer review, and funding of indirect research costs; and firing essential NIH staff.
Both parties to the dispute can agree on one thing: the NIH’s annual budget of about $50 billion makes it the world’s largest and most influential organization supporting biomedical research. To be sure, pharmaceutical firms and biomedical device manufacturers invest several times this amount in research each year, but their programs are more directly tied to future profit, while NIH can support investigations with no immediate prospects of generating revenue. Many such projects, including Watson and Crick’s discovery of the structure of DNA, have subsequently turned out to produce immense economic returns.
Yet Bhattacharya’s mission, as he conceives of it, extends far beyond cost reduction to the very ethos and culture of biomedical research. For one thing, he is focused on the contemporary “replicability crisis.” A 2016 paper published in Nature showed that more than 70 percent of over 1,500 researchers polled have been unable to reproduce the published results of another researcher. One well-known 2021 review of cancer biology research showed that among studies that had been repeated, the effect sizes were on average 85 percent smaller than what had been reported.
Bhattacharya means to combat this reproducibility crisis by reversing the current heavy bias against publishing replication studies. As things stand, biomedical journals are almost 10 times as likely to publish positive as negative results, meaning that any medical researcher who wants to get a doctorate, get published, and get funded needs to show that the therapy under investigation is effective. The culture of science must shift to recognize that studies that fail to replicate published results are as important a part of good scientific practice as positive ones.
Bhattacharya himself incurred the wrath of leaders at NIH such as director Francis Collins, not only by co-authoring the Great Barrington Declaration but also by publishing results during the first months of the pandemic that showed that the prevalence of COVID infection was far higher than officials supposed (the Santa Clara study). The significance of this finding was this: calculations of mortality rates depend not just on numbers of deaths (the numerator) but also on numbers of cases of infections in the population (the denominator). Bhattacharya found estimates of the latter were systemically too low, which dramatically and incorrectly inflated COVID’s perceived lethality.
For many, it is difficult to believe that published scientific results could be wrong. But as Bhattacharya’s colleague John Ioannidis has long argued, most findings are almost certainly wrong, due to biases built into research methodology that Bhattacharya aims to correct. This is not mainly fraud but unrecognized bias. For example, most studies have small sample sizes and low statistical power, data sets are routinely not shared, the culture of scientific publication and citation favors positive results, study results are not subjected to replication, and researchers are under tremendous pressure to “succeed.”
The NIH has been perfectly designed to produce the results it is currently getting. What is it really good at? Keeping established researchers funded and ensuring that current research paradigms are carried forward into the future. To sit on a committee that evaluates research proposals, for example, scientists must be funded, which tends to reinforce current paradigms. Moreover, researchers must present preliminary data to show that their proposals are likely to produce positive results. Not surprisingly, the NIH culture makes it difficult for new researchers to secure funding to investigate new ideas. Likewise, it’s rare to receive funding for research that challenges or seeks to overturn NIH’s previously published findings.
Bhattacharya also criticizes NIH’s current pattern of funding for indirect costs. While there are nearly 6,000 US colleges and universities, only a relatively small number – dozens – receive substantial amounts of NIH funding, often charging indirect costs rates of over 50 percent. Every $1 million of direct research funding needs to be accompanied by more than $500,000 of additional funding to cover costs such as facilities, maintenance, utilities, and support personnel. This concentrates research in relatively few centers, further reducing diversity among researchers, their investigations, and scientific ideas.
The core question is this: what does the NIH, and more broadly speaking, the scientific enterprise, exist to do? Is its principal purpose to sustain funded researchers and their research programs? Or is it to promote discoveries that improve human health? If the latter is paramount, then the culture of the NIH must shift to put truth above success. What matters most is not how many grant dollars scientists garner, how many papers they publish, or how many awards they receive, but the degree to which they accurately illuminate what we really need to know.
Much hangs on Bhattacharya’s response to the Bethesda Declaration. Will he engage in character assassination and retribution, seeking to silence or terminate those criticizing the administration’s policies? Or will he speak out and act on behalf of good science, a word with origins in the Latin root meaning “to know.”
Science needs open and self-critical discussion that welcomes differing points of view, insists on the highest standards of evidence, shuns biases, and prizes the truth above all else. The NIH’s pursuit of knowledge has been ailing, and restoring it to good health will require strong medicine.