As somewhat unclearly related in the Daily Mail:

Dr Keusch, Professor of Medicine and International Health at Boston University's Schools of Medicine and Public Health, stressed that no release of viruses from a high-level lab, such as the one in Wuhan, 'has ever happened'.

I think he's referring to the BSL-4 labs. As a quick breakdown/check on lab numbers by safety level; just in the the UK in 2007 there were 600 BLS-3 labs, but only ten BSL-4. Not all of them were operating and there's a safety sub-level for those that can operate at ACDP4--meaning BSL-4 and using pathogens that can infect humans (only 5 of those in the UK then, four of which were in operation); the other were SAPO4 (for animal pathogens). The Wuhan lab Keusch mentions next is BSL-4. So I'm thinking the claim refers only to BSL-4.

Examples of BSL-3 breaches include FMD at PIADC--DavePhd's answer-- or SARS from a lab in Beijing, 2004, included in LangLangC's answer. In fact that accident was one of the reasons that China built its first BSL-4, the one in Wuhan, operational at this level since 2018.

So is the claim (still) true for BSL-4 labs (worldwide)? (There were approximately 60 BSL-4 labs in the world, as of 2018.)


3 Answers 3


Q Has a virus ever escaped a high-level virus lab?

Yes, depending on definitions.

How deadly pathogens have escaped the lab — over and over again

Most appropriate candidates for an inappropriately cherry picking and restrictive reading of the claim in question:

  • the 1977 "red flu" occurring from China or Russia is thought o be most likely a failed experiment in such a lab. Local authorities deny it.

  • the 2007 Foot-and-Mouth disease outbreak in the UK was traced to have come from a designated BSL 4 lab researching the vaccine on a strain from 1967. Authorities tried to deny this. (But came around eventually: "Virus escape from Pirbright site")

What is an "escape"?

If it is a "lab acquired infection" (LAI), then we need to look at:

Countless examples exist, but Mr Keusch restricts his definition artificially. He seems to not count any cases of for example labs just putting deadly strains into the post and shipping them ouit of their supposedly secure facilities via mail.

In March researchers at the U.S. Department of Agriculture received a shipment of what they thought was a relatively harmless strain of animal flu from the U.S. Centers for Disease Control and Prevention. It was, in fact, highly pathogenic H5N1 flu virus. Months later the researchers only began to suspect something was amiss when their experiments yielded results that they didn’t expect and couldn’t explain. The CDC, it turns out, had sent the deadly samples inadvertently.
"CDC Botched Handling of Deadly Flu Virus The third recent mistake in handling of pathogens is a “wake-up call,” says Centers for Disease Control head", Scientific American, 2014.

And why not just post some anthrx, dengue, or plague with whatever packaging or label you want, or whether the receiver wants them or knoiws what to expect?

In 2015, the US Army revealed that its chemical and biological defence facility at Dugway Proving Ground in Utah unwittingly shipped live anthrax to seven countries after failing to kill the spores properly. Responding to Freedom of Information requests from the Guardian, the US army said it shipped one batch to a UK lab in 2007, and 13 more to three UK labs in 2009. One of the 2009 batches contained a vial of yet another unsolicited pathogen: live but weakened Yersinia pestis, the bacteria that cause the plague. – Guardian/Blunders

A 2013 Centers for Disease Control report is a significant source of recent data on LAIs (11). The report documents four undetected or unreported LAIs in registered US Select Agent, high-containment BSL-3 labs between 2004 and 2010. An undetected or unreported LAI implies an escape when the infected person leaves the lab. The report identifies an average of 292 registered Select Agent BSL-2, BSL-3, and BSL-4 labs operating over those 7 years, for a total of 292 × 7 = 2,044 lab years. Unfortunately, the study does not break down numbers into BSL-2, BSL-3, and BSL-4 labs or lab years.

Thus, the probability of escape for a single year, p1, can only be calculated as 4 LAIs/2,044 lab years = 0.002 or 0.2% per lab per year. This is clearly an underestimate since BSL-2 and BSL-4 labs contribute to the denominator. (The denominator used here, 2,004, equals the number of BSL-2 plus number of BSL-3 plus number of BSL-4 labs. But the denominator in our calculation should be just the number of BSL-3 labs, so the denominator is overestimated and the percent escape is then underestimated. Although requested, the CDC has not supplied us with the number of BSL-3 labs for us to do the exact calculation.) This basic probability is consistent with that for SARS escapes in Asia through LAIs (12) and with all known escapes from BSL-4 labs in the Soviet Union from LAIs and Great Britain from a mechanical failure (13).

The risk of a man-made pandemic from a lab escape is not hypothetical. Lab escapes of high-consequence pathogens resulting in transmission beyond lab personnel have occurred (20, 21). The historical record reveals lab-originated outbreaks and deaths due to the causative agents of the 1977 pandemic flu, smallpox escapes in Great Britain, Venezuelan equine encephalitis in 1995, SARS outbreaks after the SARS epidemic, and foot and mouth disease in the UK in 2007. Ironically, these labs were working with pathogens to prevent the very outbreaks that they ultimately caused.

– Lynn C. Klotz & Edward J. Sylvester: "The Consequences of a Lab Escape of a Potential Pandemic Pathogen", Front Public Health. 2014; 2: 116. Published online 2014 Aug 11. doi: 10.3389/fpubh.2014.00116, PMCID: PMC4128296, PMID: 25157347

Mr Keusch is quite wrong on this. Or depending on true scotsman fallacy: 'if something escaped, then the lab wasn't secure'. Any focus on viruses alone seems also quite overspecific. Biological Safety Labs should contain all germs — after all, most won't care that much whether they die of a virus or a bacterium. And whether or not a biohazard escapes from a BSL '4' or '3' lab is also quite foggy. It seems like a straw man when it is meant to refer only to 'direct escape from a BSL4 building leading to an epidemic'. If a pathogen is designated to be only handled in BSL4, but then transferred to another lab with even lower security, is that an 'escape' or 'merely serious breach of protocol?' Even if then from that lower security lab we do find environmental transmission? A PR person or lawyer worth his money will spin this as 'no escape from BSL4'.

– Bio-Unsafety Level 3: Could the Next Lab Accident Result in a Pandemic? So-called gain-of-function pathogen research will likely receive closer scrutiny after three U.S. biolab incidents, Scientific American, 2014

And such risks are easily reached:

Although the estimates of the probability of accidental escape are relatively low (0.3% risk of release per lab per year 15), the increased number of laboratories working on BSL 3 and 4 agents gives rise to estimates projecting an appreciable combined escape risk of potential pandemic pathogens (PPP) in a 10-year window 15. In addition, for PPP, the relatively small risk of release has to be weighted against the size of the population that could be affected by such an event, the risk of severe or fatal cases and the likelihood of containment before the event could escalate to global proportions. Furthermore, the quantitative analysis of the post-release scenario is complicated by the different social and environmental settings that apply to the more than 1,500 BSL 3 and 4 laboratories around the world 16.
– Stefano Merler, Marco Ajelli, Laura Fumanelli, & Alessandro Vespignani: "Containing the accidental laboratory escape of potential pandemic influenza viruses", BMC Med. 2013; 11: 252. Published online 2013 Nov 28. doi: 10.1186/1741-7015-11-252 PMCID: PMC4220800, PMID: 24283203

In 2004 the highest level lab in Russia (Vector) where the last remaining smallpox viruses are kept, had a lab worker contract ebola.

A Russian scientist working on an Ebola vaccine died last week following a lab accident. On 5 May, Antonina Presnyakova, 46, pricked her hand with a syringe after drawing blood from infected guinea pigs in an ultrasecure biosafety level 4 (BSL-4) facility at the Vektor Research Institute of Molecular Biology, a former bioweapons lab near Novosibirsk, Russia. She was hospitalized immediately, says a lab official, developed symptoms 1 week later, and died on 19 May. (science)

Such an escape is also not unimaginable, even for smallpox, when things just start exploding there, like in 2019.

A nice overview is here: – Martin Furmanski: "Laboratory Escapes and “Self-fulfilling prophecy” Epidemics", 2014 (PDF), short version online, detailing the SARS outbreaks/escapes: Threatened pandemics and laboratory escapes: Self-fulfilling prophecies By Martin Furmanski, March 31, 2014, complemented with 10 incidents discovered at the nation's biolabs.

“Overall, the incident shows that failures — even cascading, compounding, catastrophic failures of BSL-4 biocontainment labs occur,”

The Bernhard-Nocht-Institut recorded this:

A virologist working in the BSL-4 laboratory pricked herself in the finger during a mouse experiment on 12 March 2009. The syringe contained ZEBOV [Zaire ebolavirus] from culture supernatant that had been concentrated by ultracentrifugation and mixed 1:1 with incomplete Freund’s adjuvant for immunization of mice. The material was injected into the animal before the accident happened. When the laboratory worker tried to recap the needle, it penetrated the cap laterally and subsequently all 3 gloves. The puncture site on the skin was visible, but it did not bleed. The wound was disinfected after leaving the laboratory. … The patient voluntarily agreed on being hospitalized on 13 March.
– Stephan Günther et al.: "Management of Accidental Exposure to Ebola Virus in the Biosafety Level 4 Laboratory, Hamburg, Germany", The Journal of Infectious Diseases, Volume 204, Issue suppl_3, November 2011, Pages S785–S790. doi

And while USAMRIID only has had Ameri-Anthrax spread over the facility in 2001, the CDC reports that Ebola is such a lazy virus, it really took humans to just carry it to the hunt out of a BSL4:

In December, CDC reported that a small amount of material from an experiment that was part of an Ebola virus study was securely transported from a select‐agent‐approved BSL‐4 lab to a select‐agent‐approved BSL‐2 lab and may have contained live virus. The material mistakenly transferred during procedures for this study was on a sealed plate but should not have been moved from the BSL-4 laboratory into the BSL‐2 laboratory.
CDC Releases Report on Recent Lab Incident, 2015

Note the date, as this was after

In 2014, the U.S. Centers for Disease Control and Prevention (CDC) had a string of incidents that marked one of the worst occurrence numbers on record for the organization as three biosafety incidents were reported in a span of just two months. […] A moratorium on the transfer of infectious agents from CDC BSL-3 or BSL-4 laboratories to any other facility was instated pending further investigation of this incident and laboratory transfer practices.
Transfers Leading to Incidents,

Lab workers wandering around security zones without protective gear, worker shandling toxins and agents even without being properly trained, disposing biohazard waste without gloves. The USAMRIID people

systematically failed to ensure implementation of biosafety and containment procedures commensurate with the risks associated with working with select agents and toxins.
The violation specifically observed involved “entity personnel [...] propping open” a door while removing “large amounts of biohazardous waste” from an adjacent room, “[increasing] the risk of contaminated air from [the room] escaping and being drawn into the [redacted]” where the people working “typically do not wear respiratory protection.”

storing biohazard waste with unauthorised access enabled, stuff just going "missing"

The CDC reports that someone at the lab did not maintain an accurate or current inventory for a toxin.

And eventually a 'low risk level'!


Severity level: Low

The CDC reports that a building at the Fort Detrick laboratory didn’t have a “sealed surface to facilitate cleaning and decontamination.” This included cracks around a conduit box, cracks in the ceiling, and a crack in the seam above a biological safety cabinet.

– Quotes from Diana DiGangi: "Army germ lab shut down by CDC in 2019 had several 'serious' protocol violations that year", January 22nd 2020

The FMD outbreak in the UK 2007

Failures in bio-containment and bio-security received wide publicity in 2007, including at facilities where awareness and compliance with procedures were thought to be high. One such breach occurred on 3 August at a farm near Pirbright in Surrey, UK, where an outbreak of foot-and-mouth disease (FMD) was discovered. Pirbright houses the Institute of Animal Health (IAH), which uses small quantities of live FMD virus for experimentation, as well as two private biotechnology companies: Merial Animal Health Ltd and Stabilitech Ltd. The Merial facility was producing large quantities of FMD vaccine, while Stabilitech used only small quantities of live FMD virus—comparable to those used by the IAH. […]

On 7 August a second farm was infected with FMD. The strain of FMD was identified as an FMD reference strain that had been obtained from the 1967 FMD epidemic in the UK and is commonly used at reference laboratories and in the production of pharmaceuticals, as at the Pirbright site. The strain does not occur naturally.[…]

It stated that breaches in bio-security procedures, recent high precipitation in the area and lack of maintenance (i.e. cracked wastewater piping due to tree root ingress) contributed to the release of live FMD virus, which subsequently led to the infection of animals at nearby farms.

The bio-security breaches included: the incomplete inactivation of live virus through insufficient chemical effluent filtering; allowing some live virus to reach the public sewer system, and eventually the surrounding soil, through unsealed manholes; the lack of standard operating procedures for handling blockages in effluent drains; and the lack of control over human and vehicle movement in and around the facility premises because construction work was being performed at the time of the outbreak. The dissemination of the FMD virus to neighbouring farms occurred because the soil around the facility was contaminated by the overflowing sewer system and spread by vehicles entering and exiting the area.

In 2007 several potential exposures to Coxiella burnetii (the causative agent of Q fever) also occurred at Texas A&M University but were not reported to the CDC, as required by law.

– Stockholm International Peace Research Institute: Yearbook 2008, 9. Reducing security threats from chemical and biological materials // the Texas A&M exposure covered at – Jocelyn Kaiser: "Pathogen Work at Texas A&M Suspended", Jul. 2, 2007; – Lisa Schnirring: "CDC suspends work at Texas A&M biodefense lab", Filed Under: Avian Influenza (Bird Flu); Bioterrorism; Foot-and-Mouth Disease, 03, 2007.

Foot-and-Mouth disease is insofar very interesting, as in the second half of the 20th century 'the labs' were the prime originators of disease outbreaks in Europe. In Germany the procedure making an FMD vaccine up until the 60s was so flawed that infected large animals were slaughtered for getting the 'agent' and then sold! While human consumption meat was 'properly treated to inactivate the virus, animal fodder meat was just not treated, spreading the disease far and wide.

That changed only in 1966 as people learned a bit. But from 1970 to 1992 there were further 30 primary infections leading to epidemics. Of these 30 at least 22 and a probable 3 more were caused by the vaccination programme itself. The last of these outbreaks near Hannover in 1988 clearly linked to a vaccination plant. The vaccination practice ended only under economic pressure from the US market, leading to a prohibition on FMD vaccination.
(– Karl Strohmaier [former head of Federal Research Institute of Animal Virus Diseases, Tübingen]: "Infektiöse Impfungen. Ein traurig-teures Lehrstück aus der Veterinärmedizin und -verwaltung", Zeit, 17. Januar 1992. More scientific, but paywalled: Tierärztliche Rundschau "Die Maul- und Klauenseuche - Was ist nach Einstellung der Flächenimpfung zu erwarten?" 1995; 4 parts.)

That is an endemic problem of these sites. As the subheading in the following piece is called "Patterns of Problems":

The safety breach at a government lab that may have exposed 84 workers to live anthrax centered on a pivotal lapse in procedure: researchers working with the bacteria waited 24 hours to be sure they had killed the pathogens, half the time required by a new scientific protocol.

“At that point, they assumed the material was safe.

“We know that was a mistake,” said agency spokesman Tom Skinner. “What else, I don’t know.”

A week later, scientists in the same lab that prepared the original samples discovered live anthrax growing in lab dishes left in an incubator, kicking off a scramble to contain the exposure.
– Sharon Begley, Julie Steenhuysen: "Multiple protocol breaches behind anthrax exposure at U.S. federal labs", Reuters Health News June 21, 2014.

This pattern of systematic neglect leads to astonishing numbers:

Safety violations, occupational exposures
The report documents 12 losses of select agents, 9 of which were the result of record-keeping errors and 3 that involved mistakenly destroyed materials. No evidence of theft or criminal intent was found, the report said, adding that three anonymous safety and security tips received during the year were all found to be invalid.

Of greater concern were 199 instances of potential occupational exposure to pathogens, which resulted in the monitoring of 908 lab workers. Many possible exposures (671) occurred in people working with diagnostic specimens outside of registered labs prior to a sample's identification as a select agent, the FSAP said, adding that no illnesses, deaths, or environmental transmission occurred as a result of the breaches.

In the course of routine screening, two federal lab employees were found to have seroconverted to Coxiella burnetii, and one academic worker had evidence of Brucella seroconversion. All three people remain asymptomatic, the report said.

Last year, six registered labs were assigned a corrective action plan to amend unsafe work practices and poor recordkeeping. Three labs were placed under suspension because of evidence of poor containment and failing to inactivate biological agents, the FSAP said. Sixteen labs were referred to FBI investigators, who found no evidence of criminal activity. Four labs were referred to federal investigators; two involved repeated select-agent work outside of the registered lab.
– Natalie Vestin: "Federal report discloses incidents in high-containment labs", CIDRAP, Jul 01, 2016.

Leading to the conclusion:

The director of the CDC, Dr. Thomas Frieden, recently revealed that, in June 2014, CDC laboratory workers improperly handled and stored what turned out to be live anthrax and that earlier in the year, the deadly H5N1 (“bird flu”) influenza virus was inadvertently mixed with a far more benign influenza strain in a CDC laboratory and then shipped to an outside laboratory (3). Senior CDC officials were notified of that event only after the revelation of the anthrax accident. We have now learned that vials of smallpox virus labeled “1954” have been sitting for decades in a U.S. Food and Drug Administration laboratory in a red brick building at the heart of the NIH campus (4)— truly the stuff of science fiction movies.

The risks in BSL-4 laboratories today are exponentially greater than those in my father’s laboratory long ago. They are immediate and extend not just to the laboratory workers and their families but also to surrounding communities—and, in the case of smallpox and genetically altered H5N1 influenza virus, the entire globe. It is time to move past the fear created by the events of September 2001. We should suspend all research at BSL-4 laboratories until a thorough review of the CDC accidents can be completed. For the future, greatly limiting the number of BSL-3 and BSL-4 laboratories would probably better enable us to ensure their safety. We must contain the terror within.
– Deborah Cotton: "Biocontainment Laboratories: Addressing the Terror Within", 21 October 2014, Annals of Internal Medicine Volume 161 • Number 8.

  • 1
    And that's assuming that the safety levels are actually consistent across labs. Commented Apr 17, 2020 at 22:55
  • 2
    The 1977 flu is not certain to have escape from a lab; alternatives are improperly attentuated LAIV (live challenges in vaccine tests): medicalsciences.stackexchange.com/questions/22936/… Commented Apr 17, 2020 at 22:55
  • 2
    Also your answer is unfortunately not trying to sort to what extent is the claim relevant/true by lab safety level. Nor is it trying to distinguish between viruses and bacteria. It's basically "lab accidents happened", which is somewhat obvious to be true. Commented Apr 17, 2020 at 23:02
  • 7
    @Fizz 77red-flu is most likely an escape, and a virus. Where do you see the bacs? The bolded part is pretty focused. With Bacs or vax the list would of course be much longer, 77 SU Anthrax, 9/11-anthrax… Things like Cutter incident, Marburg 67 are also candidates, as is SV40-Polio etc. For a direct A on the Q, even the SARS escapes should suffice? Your claimant says: no BSL4 lab accident ever happened, until it did? Or is this true scotsman: 'if sth escaped, lab wasn't secure?' Commented Apr 17, 2020 at 23:08
  • 7
    @KlausN I read through the entire answer and find it adequate and also a little hard to summarize (if you want proper sources). The initial Q on "BSL-4" lab is unnecessarily restrictive. If a BSL-4 lab accidentally shipped out some viruses, is that an escape? Or does some worker have to take a bunch of vials, and toss it in the air on the roof for this to be an escape? BSL-4 labs HAVE shipped out live viruses to non-BSL-4 destinations, but those "didn't count" for some reason.
    – Nelson
    Commented Apr 19, 2020 at 5:43

According to Agro-terrorism :

"Accidental releases of FMD virus from research laboratories have occurred in the United Kingdom [reference 38]. Historically, FMD virus was accidentally released from PIADC [Plum Island Animal Disease Center] in 1978...

FMD virus is on the List of biosafety level 4 organisms

  • And did the BSL-4 standard exist in 1978 and was it used at PIADC? Commented Apr 17, 2020 at 15:18
  • 2
    @Fizz "Since 1954, the DHS S&T Office of National Laboratories (ONL) Plum Island Animal Disease Center (PIADC) has served as the nation’s premier defense against accidental or intentional introduction of transboundary animal diseases (a.k.a. foreign animal diseases) including foot-and-mouth disease (FMD). PIADC is the only laboratory in the nation that can work on live FMD virus (FMDV)."
    – DavePhD
    Commented Apr 17, 2020 at 15:20
  • It would improve the answer if you added that in the answer itself. Commented Apr 17, 2020 at 15:20
  • 2
    @Fizz CDC says the first BSL-4 lab was in 1967 cdc.gov/museum/app/ebola/ebola.html
    – DavePhD
    Commented Apr 17, 2020 at 15:34
  • 2
    The New Yorker says "In New York, Senator Hillary Clinton and Representative Timothy Bishop, whose district includes eastern Long Island, were opposing plans to upgrade the current Plum Island lab from BSL-3 to BSL-4. As a result, the Department of Homeland Security (D.H.S.) was searching for other sites." That suggests PIADC was only BSL-3 at the time of the outbreak. Commented Apr 17, 2020 at 23:18

According to a Slate article:

In 2007, FMD appeared again in Britain, 4 kilometers from a biosafety level 4 laboratory—a designation indicating the highest level of lab security—located at Pirbright. The strain had caused a 1967 outbreak in the United Kingdom but was not then circulating in animals anywhere. It was, however, used in vaccine manufacture at the Pirbright facility. Investigations concluded that construction vehicles had carried mud contaminated with FMD from a defective wastewater line at Pirbright to the first farm. That outbreak identified 278 infected animals and required 1,578 animals to be culled. It disrupted U.K. agricultural production and exports and cost an estimated 200 million pounds.

According to Wikipedia's detailed article on this outbreak,

An independent investigation carried out by Professor Brian Spratt found that due to the recognition that infected material could survive the initial citric acid disinfection stage within the Merial plant, the effluent system up to the final caustic soda treatment plant was considered by Defra inspectors to be within the scope of Category 4 containment, yet it appeared not to have been subject to regular inspection and there was evidence of leakage both from broken pipework and via unsealed, overflowing manholes.

So at least one breach of BSL-4 safety for viruses dangerous to animals seems to have happened. On the other hand, the Government seems to have disagreed with that conclusion

The county council's external legal advice is that a prosecution against either of the two laboratories at the centre of the outbreak is not possible. This was because:

Three Government-commissioned reports were unable to pinpoint the exact source of the outbreak

According to the same Slate article, some of the SARS lab breaches (BSL-3) were due to passing improperly inactivated samples to other labs of lower safety level:

SARS has not re-emerged naturally, but there have been six escapes from virology labs: one each in Singapore and Taiwan, and four separate escapes at the same laboratory in Beijing [NIV]. [...]

Many shortcomings in biosecurity were found at the NIV, and the specific cause of the outbreak was traced to an inadequately inactivated preparation of SARS virus that was used in general (that is, not biosecure) laboratory areas, including one where the primary cases worked. It had not been tested to confirm its safety after inactivation, as it should have been.

There's a brief Science piece on the Chinese investigation report basically confirming that.

The outbreak earlier this year, which sickened eight people in Beijing and Anhui Province and killed one (ScienceNOW, 27 April), started when two workers at the CDC lab, independently from each other, developed SARS. The most likely source of their infection, the report concludes, is a batch of supposedly inactivated SARS virus that was brought from a high-containment facility into a low-safety diarrhea research lab where the two [workers infected] were working. Apparently, the inactivation process--adding a mix of detergents to the virus--did not work properly, according to the study, of which only a five-paragraph summary has been released. In a breach of standard safety procedures, the researcher who carried out the inactivation [...] had not tested whether the virus was truly inactive, according to the panel.

("CDC" refers' to China's CDC in that quote.) According to the same source, the head of China CDC resigned as the expert-panel report was published.

Regarding the quotes on the US CDC BSL-4 from the LangLangC's answer, the US CDC denied that they were a threat to the public; the "overall" quote LangLangC used is from a critic, not from the CDC (or some official investigation). From the same source:

“Overall, the incident shows that failures — even cascading, compounding, catastrophic failures of BSL-4 biocontainment labs occur,” said Ebright, who has testified before Congress about CDC safety issues. “And the attempted cover-up within the CDC makes it clear that the CDC cannot be relied upon to police its own, much less other institutions.”

CDC officials say there was never any risk posed by the lab’s equipment failures. Although the scientific team just finished conducting an inventory of a freezer in the lab that contained numerous frozen virus specimens — likely including Ebola and smallpox viruses, no work with live agents had yet been done in the lab, which was still new at the time.

“Yes, there was some malfunction, but there was a clearly established protocol for how to deal with the malfunction and that was quickly and rapidly executed," said Steve Monroe, who last fall was permanently appointed to head a new CDC lab safety office charged with improving the safety culture at the Atlanta-based agency. Even if work with live pathogens had been underway inside the lab, he said, “the potential risk would have been exceedingly small.” [...]

Notably perhaps, Ebright recently said that Covid-19 probably escaped from a Chinese lab. So he seems to see [potentially] catastrophic lab failures in a lot of place.

Also worth pointing out (in re worker contamination examples in BSL-4) that the safety model for BSL-4 errors concerns the outside world. The examples of the workers getting contaminated in a BSL-4 lab is of course a [personal] work-safety problem, but not a containment-breach problem if the infected workers reported the event, e.g. in the US:

the uSSRA [unique site-specific risk assessment] treats all mechanical errors and human errors as equivalent by using a single numeric value for each in all the risk calculations in the BSL-4 section of the uSSRA.

Basically the workers are considered "part of the machinery" as long as they are inside. I.e. a known contaminated worker is considered equivalent to a contaminated piece of machinery from the outside-safety perspective.

Note that in the case of a US virologist exposed to a mouse version of Ebola at USAMRIID (BSL-4 lab), the person was not even taken to a hospital because it was considered an unsuitable facility. Instead the virologist was quarantined in "the Slammer" (MCS) at USAMRIID.

The 1,288-square-foot MCS includes 2 patient rooms and a treatment room and is equipped for intensive care monitoring, ventilator use, and teleradiology capability. It has an independent ventilation system and a chemical shower for decontaminating caregivers’ encapsulating suits (identical to those worn in the BSL-4 laboratories), and it is isolated from adjacent areas by doors fitted with airtight gaskets. This latter feature earned the facility the moniker “the slammer,” which was popularized in the book The Hot Zone.

The MCS is staffed by USAMRIID personnel, with augmentation and specialized care provided by nearby medical center staff, who serve on a medical augmentation team. Passage of consumables and supplies in and out of the suite occurs through 3 conduits: a double-door autoclave, an ultraviolet light passbox, and a disinfectant dunk tank, which enables decontamination and transport of specimens to external areas within the facility for laboratory analysis. The MCS is maintained under negative pressure. Air undergoes HEPA filtration upon entry to and exit from the facility. The septic system links into USAMRIID’s laboratory sewer system, which undergoes steam sterilization.

The details on the Russian case (human Ebola, leading to researcher's death) are bit more vague. I've only found a fairly vague mention of a "special hospital":

The Itar-Tass report said she was isolated in a special hospital and treated in consultation with Health Ministry specialists and a physician who has treated Ebola patients in Africa. [...]

The Itar-Tass report said those who were involved in treating the scientist and investigating the case would be kept under observation for 21 days, with daily medical exams and twice-daily temperature monitoring.

The same Russian lab suffered an explosion in 2019, but apparently no pathogens were released.

"Insider threats" i.e. a worker deliberately infecting themselves and/or not reporting an accident that is a potential infection (they were aware of) is considered a potential problem, but there don't seem to be known examples of it having happened in BSL-4 labs, meaning the worker not reporting the potential exposure (that later led to an outbreak outside the lab).

A 2004 US gov't publication said:

With regard to other BSL-4 (formerly designated P4) laboratories worldwide, the safety record is remarkable. In a 10-year period from 1959-1969, only one laboratory-acquired infection occurred in a worker in each of the two existing P4 facilities at Ft. Detrick, Maryland (Wedum 1996). Both infections were cutaneous in nature, did not require hospitalization, and posed no risk to the community. NIAID has performed a survey of BSL-4 laboratories worldwide with over 20 years of operating history to determine the number and severity of laboratory-acquired infections occurring within these facilities (Johnson 2003, see Appendix D). In the past 31.5 years (approximately 344,000 man-hours of work), in newer BSL-4 suit facilities at the U.S. Army Research Institute for Infectious Diseases (USAMRIID) at Ft. Detrick, Maryland, there have been no clinical or sub-clinical infections from any BSL-4 agent. There have been no environmental releases of infectious agents from these laboratories. The Centers for Disease Control and Prevention has operated P4/BSL-4 facilities for over 30 years (120,560 man-hours of work in BSL-4 laboratories). There have been no clinical or subclinical infections and no releases of infectious agents to the environment. The National Institute for Communicable Diseases in Johannesburg, South Africa, has operated BSL-4 laboratories for over 22 years (approximately 40,000 man-hours), where much of the work was devoted to searching for wild reservoirs of Marburg and Ebola viruses. No infections or environmental releases of infectious agents have been recorded. In summary, over 604,000 man-hours of work with exotic agents in BSL-4 laboratories have taken place without any evidence of laboratory-acquired infection or environmental release.

Based on the NIAID safety record over the past two decades; the safety record in general of P4/BSL-4 laboratories; the lack of occupationally acquired infections in employees working in these facilities during the past 30 years; and the fact that there have been no environmental releases of infectious agents from these facilities, the conclusion can be made that the risk to communities surrounding BSL-4 laboratories is negligible.

So yeah, the governmental and critics' assessments seem to diverge a bit. This latter report only focuses on a few BSL-4 labs so it might be "cherry picking". The focus on the two long-running US BSL-4 labs is perhaps understandable; there were only two before 1990 in the US and only five before 2000. A substantial building/expansion program was started thereafter with 15 such facilities in the US by 2007. (The number of BSL-3 labs in the US is much higher, at least 1350 by 2008.)

A 2009 US CDC paper says much of the same:

these [BSL-4] facilities have had a distinguished safety record for >30 years; no clinical infections, no major incidents in operation of the physical facilities, and no escape of any agent with clinical consequences for neighboring communities have occurred.

The same paper rejected the idea of a peer overwatch ("1 to work and 1 to observe") during work in BSL-4, saying that video monitoring of the workers is enough to ensure that they report any accident.

  • Note that FMD2007 and the SARS breaches are in my A as well. And on that the governments always disagree. In China nothing ever happened… (Nor the US But for FMD it is reported that from 1970 until official vaccination prohibition 22 of 30 outbreaks in Germany were coming from labs. And look how official institutions now argue/explain for why there are no vaccination campaigns for FMD… Commented Apr 18, 2020 at 11:25
  • @LаngLаngС: China did not contradict that the Beijing outbreak came from a (China CDC) lab. See added quote. Commented Apr 18, 2020 at 11:27
  • @LаngLаngС: I see you've expanded your answer significantly since I last read it. But you (still) didn't mention in it that the 2007 FMD was supposedly from a BSL-4. Commented Apr 18, 2020 at 11:37

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .