IN THE SPRING of 2004 Antonina Presnyakova was working at the Vektor Research Institute of Molecular Biology near Novosibirsk, a remote Russian city. The scientist accidentally pricked her hand with a needle while drawing blood from an infected guinea pig and was rushed to hospital. There was little to be done. Two weeks later she died of Ebola.

The Economist Today

Hand-picked stories, in your inbox

A daily email with the best of our journalism

Much has been made of the “lab-leak hypothesis”—the theory that covid-19 escaped from a research facility in Wuhan, China. In May, President Joe Biden ordered the American intelligence services to produce a report assessing its credibility, to be delivered to him on August 24th. According to the American Biological Safety Association (ABSA), which maintains a database of such incidents, laboratory-acquired infections (LAIs) are disturbingly common. The majority are relatively mild: the most common type of pathogen involved in LAIs is Brucella, a type of bacteria that can cause flu-like symptoms and is easily treated.

But some cases, such as Presnyakova’s, are severe. Since 1970 there have been eight other incidents involving Ebola, and five involving SARS. Laboratory workers have been infected with Dengue fever, HIV and Zika virus. In 2009 a microbiology professor at the University of Chicago died after contracting the plague.

It is not just those who work in laboratories who are at risk. Deadly pathogens escape beyond laboratory walls with disturbing regularity. In 1979 at least 68 people died when anthrax spores leaked from a Soviet military facility and drifted downwind. In 2007 an outbreak of foot-and-mouth disease in Britain was caused by a leaky pipe at the Pirbright Institute, a high-security lab in Surrey. Between 1989 and 2002 there were, on average, 13 “exposure events”, or incidents involving exposures to “potential agents of bioterrorism”, each year at the American army’s biological-warfare research centre in Fort Detrick, Maryland, although only five of those events resulted in infections. More than 200 people were infected with Brucella in 2019 as a result of an error during vaccine production in Gansu, China.

It takes only one severe incident to create a disaster. The chances of such a catastrophe happening are increasing. In the past decade, more than 20 new “maximum-biosecurity” (known as biosafety level 4, or BSL-4) labs have opened. One of the most recent, certified in 2017, is at the Wuhan Institute of Virology—the focal point of the covid-19 “lab-leak hypothesis”. Both the Vektor Research Institute and the Pirbright Institute have BSL-4 labs. High-risk work is also sometimes done in lower-security facilities.

What can be done to prevent lethal pathogens from escaping into the world? Between 2014 and 2017, gain-of-function research—in which pathogens are manipulated to make them deadlier, more infectious, or both—was banned in America. Some researchers reckon that this ban should be reinstated, not least because many countries take their biosafety cues from America. A ban might even prevent a pathogen escaping from an American lab. But to better understand deadly viruses and to develop treatments for them, it is necessary to study them. At present, there is scandalously little oversight. International standards ought to be agreed on, and a watchdog independent of those who fund research set up to enforce them. That would, at the very least, provide some inoculation against disaster.