Are BSL3 and BSL4 labs safe?  This is what we found.  (If you have other examples to share, please let us know!)

The National Institutes of Health (NIH) and Boson University (BU) claim that BSL4 laboratories have a long history of safety. They base their claim on an alleged “excellent” safety record of three BSL4s in the United States as compiled by Karl Johnson, MD.  Dr. Johnson’s research, however, is anecdotal, not fact based.  He relies only on interviews with staff at the facilities. It is not a detailed review of all laboratory exposure events at the three facilities, and should not be relied upon to make claims about the safety of BSL4 laboratories. In fact, the record of reported accidents and infections in high containment biological laboratories is troubling.

An example of a false safety claim is found in the Supplemental Draft Environmental Impact Statement (SDEIS) for the BU BSL4 lab facility where the NIH and BU wrote that, “With the longest running experience with a BSL-4 (33 years) Ft Detrick Maryland has an out­standing safety record…. Previous documented exposures at Fort Detrick in their original lab facilities mention one laboratory-acquired infection between 1959-1969 and no clinical or other infections in the more recently constructed USAMRIID facility.”

That statement by NIH and BU is wrong. USAMRIID (Ft. Detrick) has an extensive history of both exposures and laboratory-acquired infections. According to a study by USAMRIID researchers, published in the Journal of Occupational and Environmental Medicine in August 2004, 234 employees at USAMRIID were evaluated for exposure to 289 biological agents classified as “bioterrorist agents,” resulting in 5 confirmed clinical infections between 1989-2002. The recorded infections were from exposures to glanders, Q fever, vaccinia, chikun­gunya, and Venezuelan equine encephalitis. There were also numerous exposures to anthrax, plague, Western and Eastern equine encephalitis, ortho­poxviruses, yellow fever virus, and Rift Valley fever virus which did not lead to infections, but for which postex­posure antibiotic prophylaxis was administered (when available). For some of these diseases there is no treatment.

The report, (Rusnak, et al. 2002) thoroughly reviewed all exposure records, and paints a significantly different picture of the safety record at USAMRIID than Dr. Johnson’s report, which implies that accidents are extraordinarily rare. In contrast data shows that there were an average of 16.7 persons evaluated per year for accidental exposures to bioterrorist agents. The authors of the study conclude:

In summary, we reviewed available medical and safety records at USAMRIID from 1989 to 2002 and reported on 234 evaluations of potential exposures and illnesses to bacterial, rickettsial, and viral disease agents. During this period, there were five confirmed infections. The large number of exposure incidents reported in this time period serves as a reminder that work in a laboratory of this type is inherently hazardous.

Therefore, it is imperative for laboratories that elect to work with highly hazardous agents to be fully cognizant of the risk of occupationally acquired illnesses and institute policies and proactive employee health procedures to evaluate potential exposures.

A recent study of anthrax releases at Fort Detrick supports the conclusion that high containment biological laboratories present real risks to the community. An October 14, 2004, USA Today article reported on the U.S. Army report on the anthrax releases from the Fort Detrick BSL3/4 laboratory. Three strains of anthrax escaped the supposedly secure BSL3 laboratory, which is designed to enable scientists to safely work with deadly microbes. Two of the strains were used in biodefense work.  The report and statements of experts in the article serve to show that NIH and BU are incorrect to assert that there would be no human health or environmental damage from an anthrax release from a high containment laboratory.  Highlights of the article include:

Researchers expressed relief that no one was hurt or killed in the episode, but Stephanie Loranger of the Federation of American Scientists asks, “Fort Detrick is one of the premier biodefense labs, and if they have problems, what does it mean for all the others?”

 “The good news is nobody got the disease (i.e., anthrax),” says Alan Zelicoff, a biodefense expert who is now a consultant at ARES Corp., a risk analysis firm. “The bad news is that nobody got the disease because just about everybody near the BL-3 suite had been vaccinated.”

“The message here from a scientific and policy standpoint is profound,” Zelicoff says. “Facilities that are medical and microbiological may not be suitably equipped for dealing with aerosolized versions of the organisms that they otherwise deal with in great safety. These facilities probably ought not be located in a heavily populated area. How do you contain smoke?”

In 2005, the Council for Responsible Genetics released a survey it conducted of known accidents and security breakdowns at high containment biological research laboratories.  It found reports of numerous accidents and other incidents at the labs, including:

1. loss or inability to account for significant quantities of bioterrorism agents;

2. infections of laboratory personnel with dangerous pathogens, resulting in serious deaths and serious illness;

3. failure of key safeguards and security measures; and

4. use of bioterrorism agents removed from a facility to cause harm to others.

A release of a "select agent" from a high containment biological research laboratory could kill people and cause a public health crisis.  The federal government defines select agents as biological agents and toxins that "have the potential to pose a severe threat to public health and safety" and lists the select agents at 42 CFR  Part 73.  BU's laboratory will be storing, growing, and experimenting with select agents in its BSL3 and BSL4 containment areas.

An example of deaths caused by a release was the accidental anthrax release in 1979 in the Soviet Union described on the scenarios page of the MIT Security Studies Program website.

The NIH and BU not surprisingly fail to note that there is no comprehensive reporting of laboratory accidents and infections. As explained by the Council for Responsible Genetics when it released, Mistakes Happen: Accidents and Security Breaches at Biocontainment Laboratories,

Many advocates claim that accidents rarely, if ever occur at Biosafety Level 3 and 4 facilities. In fact, the frequency of such incidents is widely disputed, given a lack of comprehensive federal reporting guidelines for laboratory exposures/infections, breaches of security, or environmental releases from microbi­ological and biomedical laboratories.

The Biological Safety Program website makes clear in its historical review of laboratory- acquired infections that:

It is important to keep in mind that there is no way to determine precisely how many laboratory-acquired infections have occurred or whether all such reported infections were, indeed, laboratory-acquired. WHY IS THIS? You might ask.  The answer is quite simple. There are no regulations that require the reporting of laboratory-acquired infections.

The laboratory acquired infections website, maintained by research scientists, reaches a similar conclusion:

With the advent of bioterrorism, and the emergence of various infectious diseases (from HIV, to E. coli O157:H7 and SARS), public consciousness and the level of research activity on human pathogens are at an all time high. It is therefore logical that the frequency of laboratory-acquired infections has risen in proportion. No one really knows for sure because data are lacking. The CDC does not routinely report laboratory-acquired infections, because they are not considered outbreaks. This suits research institutions, who prefer to avoid the negative publicity of accidents in their laboratories, and federal agencies, who prefer to look the other way. The CDC does not automatically investigate lab accidents and when it does, its recommendations are not shared widely and indeed may not even be dispersed to other labs in the same institution. Too often, laboratory accidents are either entirely covered up or sanitized for public consumption. The victims of this indifference and of the cover ups are the researchers, who pay with their health, their careers, and sometimes their lives.

The establishment of a mechanism of surveillance of laboratory-acquired infections, with systematic reporting, systematic investigating, efficient dispersion of recommendations and institutional accountability is long overdue. Continued federal indifference allows continued recklessness and negligence in laboratories with potential disastrous consequences for researchers and the public they come in contact with.

 Lab Accidents Continue To Occur

Despite the failure to report, there have been recent revelations of other laboratory errors leading to infections. For example, in April 2005, the Sunshine Project reported that a leaky aerosol chamber was responsible for three laboratory-acquired infections in a Seattle, WA, BSL3 laboratory in 2004, and that such chambers are used in many high containment laboratories across the country, including those doing biodefense work. The Sunshine Project concluded that,

This case underscores how the ‘precise and neat’ public image of BSL-3 and BSL-4 facilities that is promoted by NIAID and labs is frequently at odds with messy and risky realities.  ... In this case the institutions involved apparently didn’t even inform their peers about the problems. Public Safety and an informed debate about the biodefense program require the government to mandate public disclosure of all significant lab accidents. This may be more cold water on overheated biodefense safety claims; but we frankly wonder how many more serious problems have been kept out of the public eye.

Other recent accidents in labs are discussed below on this page and in the Council For Responsible Genetics "Mistakes Happen" report discussed above.  An recent example of a shipping problem took place in a hospital near San Francisco in 2004, where researchers were exposed to anthrax, thought to be the result of a shipping error at the laboratory that supplied the anthrax.

The NIH and BU not surprisingly fail to discuss modern high containment biological laboratories outside the United States where there have been recent releases.  As reported in the Washington Post on May 29, 2004, and acknowledged by the World Health Organization and the Centers for Disease Control (CDC), SARS outbreaks in Asia have been caused by releases from high containment biological laboratories,

Scientists still do not fully understand exactly where or how SARS emerged 18 months ago. But it is now clear that the most threatening source of the deadly virus today may be places they know intimately -- their own laboratories.

Together the three SARS outbreaks have highlighted the unique hazards to public health that arise from accidental laboratory releases of germs that no longer exist -- or barely exist -- in the wild.[1]

The CDC wrote in Health Information for International Travel 2005-2006,

From July 2003 through September 2004, 17 laboratory-confirmed SARS cases were reported. Six persons were infected through laboratory exposures: one in Singapore, one in Taiwan, and four in China. One of these six cases initiated a chain of transmission that ultimately resulted in seven additional cases and one death. All the laboratory-acquired infections resulted from lapses in appropriate bio­safety procedures.

There is no comprehensive state, federal, or international law mandating and enforcing laboratory safety. Instead, there are non-binding guidelines from the U.S. National Institutes of Health that require laboratories to have an Institutional Biosafety Committee (IBC), appointed by the laboratory, to review the proposed recombinant DNA experiments to be undertaken in the lab. In 2004, the Sunshine Project undertook an eight month survey of the IBC system and published a report, Mandate for Failure, in which it found that the IBC system is broken and that a “dramatic overhaul,” including mandatory compliance and detailed disclosure requirements, is needed.

The recommendations in Mandate for Failure include:

…. the United States does not have comprehensive laboratory safety law. The system does not even have comprehensive reporting requirements for accidental releases.[2] Although there has been a rash of laboratory accidents reported in the US and abroad in 2003-4,[3] proponents of the IBC system tend to overlook this fact, even as they congratulate themselves that critics worst fears about an accidental release have not been realized. After the anthrax letters of 2001, the same glib attitude is not possible in the case of deliberate release.

The IBC system and, hence, the US approach to review of biological weapons research will only be made effective by making it a matter of law. A regulated system should establish real, enforceable consequences for noncompliance and a level playing field by making binding regulations equally apply to all institutions conducting biological weapons and biotechnology research.

The Sunshine Project pointed out in Mandate for Failure that as more scientists are attracted to the expected increase in bioweapons agents research dollars inexperienced researchers will be performing experiments with dangerous pathogens.  It quotes a biosafety officer asking, “What I want to know is who is going to tell them [inexperienced scientists] that they’re out of their league.”  (i.e., that a wave of persons with inadequate training and experience would be doing dangerous experiments with bioweapons agents.)  It also notes that,

It is generally acknowledged, most often indirectly, that too many scientists with too little training are handing agents that are too dangerous for their experience.

In a sadly ironic but perhaps telling example, the Boston University research scientist who resigned after the laboratory-acquired tularemia infections in his lab were made public in 2005 had no BSL4 experience but had been designated by BU as the head of training for the BSL4 lab in BU's application to NIAID for funding for the lab.  Also of concern is that the "Principal Investigator" of the BU lab has NO significant BSL4 lab experience.

The MIT Security Studies Program website, Level 4 Biosafety Laboratories: Public Information and Risk Analysis, makes a similar point that increased funding means increased risks:

Adding to the concern is the question of whether universities will have the resources to operate biodefense laboratories safely.  As reported in the May 4, 2004, The Scientist,

Biolab Home         Back         Next