The lab-leak hypothesis for Covid-19’s origin, once a forbidden topic in the press and on social media, is now the subject of fierce debate among scientists and journalists. But the very possibility of a lab leak should be alarming, and not just retrospectively. A recent report based on documents from the National Institutes of Health reveals wide variation in how institutions respond to lab accidents involving dangerous pathogens. Regardless of whether the U.S.-funded Wuhan Institute of Virology released Covid-19, policymakers should change their attitudes toward this kind of research in the future.
What form should this rethink take? Some have called for a crackdown on “gain-of-function” research, in which an existing organism is modified to enhance a given feature. Others have advocated stricter safety standards for research funding or better training in labs that handle pathogens. Still others want independent oversight of especially high-risk research. All these proposals are worth serious consideration—but an under-discussed possibility would be to acquire better, earlier data on lab leaks. The current approach to detecting lab leaks is insufficiently proactive: if an incident happens, the world will know only if scientists and lab techs self-report it or if enough people develop symptoms. An early-warning system that uses large-scale genetic sequencing would strengthen security against lab leaks and other pathogenic threats.
In a way, the world was lucky that Covid-19 wasn’t worse. Though the disease killed perhaps 1 million Americans (and many more globally), its infection fatality ratio was quite low. Smallpox, on the other hand, killed around 10 percent of its victims, while untreated bubonic plague and Ebola kill around 30 percent. And novel pathogens with no known treatment have a non-zero chance of emerging every year. Industrial-scale animal farming, wet markets, and human encroachment onto animal habitats may increase the risk of new pathogens, which rapid intercontinental travel can spread around the world.
Of course, natural threats don’t tell the whole story. Pathogens engineered to be more dangerous—whether for benign or nefarious purposes—present another threat. In the 1980s, using biotechnology that would today be considered primitive, a Soviet bioweapons program developed antibiotic-resistant strains of anthrax and other diseases, estimating death counts in the hundreds of thousands with merely one successful city attack. In the future, technological development will make creating lethal diseases even easier: gene editing through CRISPR and its successors will improve; more predictive computer simulations will make lab work more efficient; and the proliferation of DNA-synthesis companies will reduce barriers to entry.
Policymakers have some available tools to blunt the impact of engineered or natural pandemics: stockpiling existing vaccines, funding therapeutics for entire families of viruses, and encouraging rapid vaccine development. But these are imperfect. Novel pathogens without an available therapeutic or vaccine make stockpiling less useful. Pathogens can be designed to evade treatment or vaccination, as researchers demonstrated with mousepox. Malign actors can use DNA synthesis to manufacture new diseases, as scientists showed by synthesizing horsepox. And even under ideal circumstances, making, testing, and distributing new vaccines and therapeutics takes time.
A different approach would plug leaks as they happen. Consider one proposal, the Nucleic Acid Observatory (NAO), from scientists at the Massachusetts Institute of Technology. Instead of waiting for clusters of disease cases to trigger government investigation and for disease-specific testing capacity to scale up—a major problem in the U.S.’s early Covid response—the NAO represents an always-on capability that detects diseases before epidemiologists notice them. Under the NAO, scientists would collect water samples from specific locations, extract genetic material from the samples, and amplify that material (or look for a given pathogen). Notably, the system allows for the detection of novel pathogens—even if we don’t know what we’re looking for, we might be able to find it. Once the genetic material is analyzed, computational methods will show whether specific sequences are becoming more prevalent over time.
This system wouldn’t be perfect. Sensitive monitoring systems for pandemics will inevitably yield false positives, risking complacency among the public and costly overreactions among policymakers. Balancing timely warnings with hypersensitivity will be difficult. But just as hurricane tracking systems have made us much safer (and saved money) despite the occasional false alarm, so might the NAO.
Detecting future epidemics faster is a national imperative. It means not only more time for drug development but also raises the odds of successful containment. Instead of the U.S.’s flawed Covid response or China’s draconian “Zero Covid” policy, which seeks to avoid an already-endemic disease via multiyear lockdowns and mass surveillance, policymakers should think about all the epidemics that didn’t happen.
An early-monitoring system for pandemics would also have useful secondary effects. The data it obtained could improve the detection of invasive species, for instance, or illuminate severe viral syndromes that often go undiagnosed. (A recent study linked severe pediatric liver failure to a common virus.) It could facilitate the discovery of novel microbes, a historically useful source of novel chemical compounds. And it could advance the strategically vital American biotech industry, vis a vis China’s own interest in it.
Private philanthropies are funding crucial research into validating the design of the NAO, but a pilot system on a national scale would be too costly even for the most ambitious philanthropist to fund indefinitely. After the national nightmare of Covid, pandemic preparedness seems like a worthy goal for government funding. More academic research is needed, and other solutions may ultimately prove more promising. But Congress should keep tabs on proposals for pandemic early-warning systems and support the most promising entrants.
Long ago, the telegraph made a national weather-reporting service possible. The barometer, which measures atmospheric pressure, was invented much earlier, but only the telegraph enabled far-flung stations to report back to a central service. DNA sequencing, meantime, was first performed in the 1970s, but only enormous cost reductions in the last 20 years, along with improvements in computation and data storage, made the NAO feasible. Just as technological progress allowed humans, to some extent, to predict the weather, it may eventually allow us to predict pandemics.