The investigations, litigation, petitions, blog posts, and debates about how Sheri Sangji died will likely continue for quite some time. Already, however, official inquiries into the lab fire that mortally injured the 23-year-old research assistant at the University of California, Los Angeles (UCLA), make clear the need for systemic steps to prevent such events in the future. It's a need that exists at many institutions, lab-safety experts and researchers say--and not just in the lab of UCLA chemistry and biochemistry chair Patrick Harran, where Sangji was engulfed in flames as she worked without a lab coat with a pyrophoric substance on 29 December 2008.
Citations issued against UCLA on 4 May by the California Division of Occupational Health and Safety (Cal/OSHA) demonstrate some of the changes that are needed. One of the three "Serious" violations for which UCLA was cited--failure to use "appropriate personal protective clothing ... while working with hazardous chemicals"--was judged "Accident-Related" by Cal/OSHA. A "Serious" violation is "one with a high probability of causing death or serious injury to an employee," agency spokesperson Erika Monterroza explains by e-mail. Echoing lab safety experts quoted in Science Careers last month, Cal/OSHA also faulted a lack of safety training and noted "unsafe work practices identified during inspection and not corrected in a timely manner." The agency imposed fines totaling $31,875.00. In a statement issued the same day, UCLA announced it would not contest the fines.
Safety expert Neal Langerman, an officer of the American Chemical Society's (ACS's) Division of Chemical Health and Safety, wishes the investigation had gone further. In an article titled "What Went Wrong?" forthcoming in ACS's peer-reviewed Journal of Chemical Health and Safety, Langerman notes that the Cal/OSHA citations apply to specific, observed violations of particular regulations and do not delve into possible causes for the violations. The fines imposed relate to the violations of regulations and not to any consequences that might have flowed from them.
Any punitive action for those results would require separate criminal or civil proceedings. Cal/OSHA routinely reports violations resulting in fatalities to the appropriate district attorney, who decides whether criminal charges are warranted. In addition, injured individuals or their families can bring legal action of their own.
In an interview with Science Careers, Langerman objects to the word "accident," which both Cal/OSHA and Harran used. (Harran, in an e-mail to Science Careers, termed the fire a "freak accident.") Instead of being a chance occurrence, Langerman believes the incident was "totally foreseeable ... and totally preventable." But more importantly, he continues, "the big thing that isn't addressed is management responsibility. UCLA and the PI [principal investigator] failed to provide safety leadership." In his forthcoming article, Langerman states that "the safety literature clearly recognizes that individual managers must be held personally accountable for the safety and environmental protection of those they directly supervise."
The need for institutional leadership to create a "culture of safety" is mentioned prominently in a report by the National Academies' Institute of Medicine (IOM) titled To Err Is Human . But such a culture also makes safety the responsibility of each individual. The level of safety prevailing in a lab vitally affects everyone working in or near it. In 2006, Dominique Burget, a professor at France's École Nationale Supérieure de Chimie de Mulhouse, part of the Université de Haute-Alsace, died from an explosion in a lab on the floor below. The Cal/OSHA citations against UCLA, writes Langerman in his article, "speak to a laboratory which lacks a safety culture."
Finding a balance
The same goes for many academic labs, admits a senior professor at a topflight research university who also has extensive experience in industry. "My company would never have gotten away with or tolerated a lot of the conditions that we tolerate in universities," says the researcher, whom we'll call Seena Lott. Wearing protective clothing while handling hazardous substances, for example, is a basic safety principle and a requirement of OSHA regulations. (OSHA jurisdiction covers only employees in a workplace. Had Sheri Sangji been pursuing a degree or receiving a postdoctoral fellowship when she worked with t-butyl lithium minus a lab coat, Cal/OSHA could not have looked into her case.) Langerman writes, "Given the facts as we know them, it is difficult to understand why a scientist of Professor Harran's expertise would not require fire-retardant clothing or lab coats when working with pyrophorics."
Harran himself writes, in a statement sent to Science Careers, "It seems clear ... that I did not emphasize strongly enough the care and attention necessary when working with such materials." Among numerous reforms announced by UCLA in its 4 May statement, "Flame-resistant lab coats have been purchased and are being distributed to laboratory personnel in the department of chemistry and biochemistry who work with flammable materials. Other labs will be required to make coats available whenever high-risk activities are performed."
Internet chat in the wake of Sangji's death indicates that academic labs across the country routinely fail to enforce such elementary precautions. "I have worked at 5 different colleges and universities and none of them made lab coats available to the students," writes the blogger at Mad Chemist Chick. "Even when lab coats are provided, many people in academic labs don't bother to wear them. In my previous lab, a lot of people only wear them when they know an inspection is coming up," adds a poster at that site who signs "Anonymous."
"Scientists ... need to understand that they are responsible for human lives, and their carelessness has real consequences for people and their families," says Naveen Sangji, Sheri Sangji's sister and a student at Harvard Medical School in Boston. Yet in the culture of academic science, there is a widespread belief that an excessive focus on safety can work against the goal of discovery. Says Lott: "An important component of the educational process"--not to mention of scientific progress--"is uniqueness and flexibility." Researchers must constantly do things never done before, build equipment no one has ever used, [and] combine processes in unfamiliar ways. In research settings, therefore, "safety is not something that can be legislated" in the sense of prescribing every feature of a scientist's activities. Beyond that, she doubts that strict, detailed regulation works best in the long run because "people begin to depend on the procedures and they forget about common sense."
"Basically, we're looking for a balance," Lott says. "There are precautions that are reasonable." She, for example, requires new members of her lab, who routinely work with electricity, to take an electrical safety course.
Training is not enough
Merely offering or even requiring training does not suffice to create a culture of safety and responsibility, however. Also needed are conspicuous indications from the leadership of both the lab and the institution that safety is a priority. That is the opinion of an anonymous university safety officer interviewed on the Chemjobber blog. "Do people listen to me?" the officer writes. "For the most part I would say no. There's a huge push to get research done, and the grad students and postdocs will do whatever it takes to get it done. Lots of pressure from the PIs." He notes, "there are some faculty members who insist on a safe lab with things being done appropriately." But "for the most part, ... I'm told people don't ask questions because the 'safety' person will end up costing them time or money."
Scientists who find themselves in labs lacking vigorous safety cultures can take steps to protect themselves and others, Chemjobber argues. Ultimately, he believes, safety depends on the vigilance and dedication of each lab member. Two postdocs present while Sangji undertook her last experiment, for example, were wearing lab coats. It is unclear, notes a summary of a Los Angeles Fire Department investigation, "why she wasn't directed [by these senior lab members] to wear proper PPE [personal protective equipment]." " Why didn't they say something?" Chemjobber demands, in boldface italics. "If I was a postdoc (and I have been) and a coworker was starting [something dangerous] ... in the hood next to mine without their lab coat, I might have expressed some concern (understatement?)."
"I beg of you,…" he implores his fellow researchers, "don't be afraid to call out your lab mates if they're being unsafe. Yeah, it might make you the lab jerk, but it'll keep someone safe--trust me."
"I have, on numerous occasions, called people out ... for unsafe practices," attests Kyle Finchsigmate on The Chem Blog. "Safety training may not, in itself, ultimately teach people where every danger in the lab lives, but it might change the atmosphere from one of cavalier disregard to active vigilance."
Building a real culture of safety takes more than that, however, the IOM report makes clear. Institutions need to "provide strong, clear, and visible attention to safety; ... incorporate well-understood safety principles, ... [and] establish ... team training programs ... that incorporate proven methods ... such as simulation." Such training, for example, could have equipped the two postdocs, who raced to Sangji's aid and quenched the fire using a lab coat and water from a sink, with more effective means of meeting the crisis.
Lab safety is now a heightened concern in the contract negotiations currently under way between the University of California and the new union representing the postdocs on its 10 campuses, including UCLA, says UC Berkeley postdoc Matthew "Oki" O'Connor, a member of the union negotiation team.
Sensible safety practices will not limit researchers' freedom to innovate and create, Langerman argues. "If [UCLA] Chancellor Block and his counterparts throughout academia ... seize the moment and begin to make the required ... changes" Langerman writes, "the needs of the research community can be better balanced with the safety practices designed to protect those who make research possible."