Recent weeks have seen a pair of developments that could focus and intensify the national discussion of academic lab safety. On 9 May, after many delays, University of California, Los Angeles, professor Patrick Harran was arraigned on felony charges arising from the 2009 death of lab assistant Sheri Sangji. A judge entered four not guilty pleas on Harran's behalf, setting the stage for the first-ever trial of a university lab chief on criminal safety charges. The judge is expected to decide in late June when the trial will begin.

Then, in a step that could have even broader ramifications than the unprecedented trial, on 15 May the National Academies held a public fact-finding meeting in Washington, D.C., kicking off a yearlong study of lab safety in nonindustrial institutions. Together, these events could transform the academic community's understanding of faculty and institutional responsibility for safety and provide high-profile proposals for improving universities' safety performance.

Entitled "Establishing and Promoting a Culture of Safety in Academic Laboratory Research," the National Academies study plans to focus on "laboratory safety in chemical research in non-industrial settings," according to its statement of scope, but its recommendations likely will be relevant to other disciplines as well. The committee will examine, among other issues, institutions' "current hierarchy of actors responsible for laboratory safety," which is considered by many experts to be a crucial factor in any organization's safety performance. The group plans to "examine knowledge from the behavioral sciences, and experience with safety systems from other sectors (such as industrial research facilities, nuclear energy, aviation and medical) for key attributes of successful safety systems and cultures," and "Use this to draw lessons that could be applied [to] non-industrial laboratory research."

The committee is chaired by chemist H. Holden Thorp, who is chancellor of the University of North Carolina, Chapel Hill and on 1 July will become provost at Washington University in St. Louis, Missouri. The committee's 13 members bring to their task a wide range of expertise and experience in academe, industry, and government.  Along with distinguished chemists, they include specialists in risk management, human factors, and organizational change. Several come from institutions, such as the University of California and Texas Tech University (TTU), that have recently experienced catastrophic safety incidents and made significant institutional changes as a result.

A safety gap

An overriding theme at the meeting was the nature and size of the disparity between safety cultures and practices in industrial and academic settings. I have long heard from safety experts—and stated in my writing—that industry's safety record far surpasses that of academe. Typical of this view is a letter by officials of three major industrial corporations, Dow, Corning, and DuPont, that was recently published in Chemical & Engineering News and was quoted at the May meeting. "The facts are unequivocal," the letter asserts. "Occupational Safety & Health Administration statistics demonstrate that researchers are 11 times more likely to get hurt in an academic lab than in an industrial lab."

Committee discussion suggested that such definitive data on rates of safety incidents in the two settings may not actually exist. Still, industry figures argue from experience that this gap in safety practices is large and significant. Science Careers recently quoted William Banholzer, Dow's chief technology officer, saying that new hires from academe, be they bachelors or doctoral graduates, all require "remedial" safety training.

Excellence in safety and health is one of the four core values at DuPont, said Robert Krzywicki, who is the company's global practice leader in employee safety, in a presentation at the meeting in Washington, D.C. Values are "guardrails" that channel action, he explained. "We believe that all incidents and injuries can be prevented … while doing cutting edge research." Dupont, he said, holds supervisors "accountable for zero illnesses and injuries. … You will achieve the level of performance that you demonstrate you are willing to accept. … If you're willing to accept zero, it's achievable."

Evidence presented at the meeting suggests that many campuses lack forceful safety standards. Nathan Watson, president and CEO of BioRAFT, a developer of software for managing university laboratories, presented preliminary results from a 2012 survey of lab-safety culture conducted by the company, the University of California Center for Laboratory Safety, and the Nature Publishing Group. More than 90 percent of the survey's 2360 respondents work in academic research settings, 62% in the United States, 15% in the United Kingdom, 6% from the European Union, 4% in China and 2% in Japan. Although 85% of respondents agreed with this statement—"appropriate safety measures in my lab have been taken to protect employees from injury"—other answers suggest otherwise.


Courtesy of Nathan Watson

Nathan Watson

For example, 45% of respondents, and 55% of those working in labs with 20 to100 workers, expressed the opinion that "overall safety could be improved" in their workplaces. The overwhelming majority of respondents reported that their labs allow people to do experiments while alone, which is a very elementary violation. Only 7% said that this never happens in their labs, 35% called it a daily occurrence, and 80% said it occurs at least weekly. Only 46% of those who say their work requires a lab coat report wearing one at all times. And 40% of the respondents with supervisors reported that they—the supervisors—fail to "regularly check their performance in terms of safety." Almost 10 percent of workers in small labs (labs with fewer than 11 workers) reported that "no individual [is] specifically responsible for lab safety."

Evidence of wide variations in standards came from another presenter, Elizabeth Mackey, a safety program coordinator at the National Institute of Standards and Technology (NIST). Safety and safety training are core values of NIST culture, taught systematically to all employees, she said. But, NIST also welcomes visiting researchers from many other institutions, and they arrive with a wide range of knowledge, habits and attitudes regarding safety practices. The constant flow of scientists and the varying lengths of time they stay makes training people for and maintaining a "solid safety ethos" a "challenge."

Motivating change?

Another focus of discussion was how to instigate cultural change in universities and motivate new practices among faculty members accustomed to a high degree of independence. "It takes the entire institution to change the culture" in science labs, said presenter Marta Gmurczyk, staff liaison to the American Chemical Society's committee on chemical safety. Institutional efforts to instill an effective safety culture "must be both top down and bottom up," said Tobin Smith, who is vice president for policy at the Association of American Universities, noting that safety must become "part of what it means to be a scientist" but that "regulations don't always change cultures." Universities, he added, "have a lot to learn from industry." People arriving to work at DuPont immediately learn that they have to change their behavior, Krzywicki said. The point is "reinforced constantly,” he continued, that "there is no risk of injury worth taking at DuPont."

A committee member wondered aloud why such attitudes do not prevail at universities. One persuasive explanation comes from a study by the U.S. Chemical Safety Board (CSB) on an explosion that severely injured a TTU graduate student. CSB investigator Mary Beth Mulcahy, who worked on that document, addressed the meeting by telephone. (She was busy, she said, with the investigation of the April explosion in West, Texas.)

As Science Careers mentioned in previous reporting, the groundbreaking TTU study explores differences in authority hierarchies and financial incentives in the corporate and academic worlds. Basic to the structure of many research universities, the report states (quoting James C. McCroskey, an expert in communication and organizational change who died late last year), is the "fiefdom" system of professors who "are nominally subordinate to a person or persons above them, but in practice do pretty much whatever they want so long as they do not stray too far into some other fief's territory. … Each fief has an intellectual or administrative territory over which he or she reigns."

As a result, the report continues, "At academic research institutions, PIs may view laboratory inspections by an outside entity as infringing upon their academic freedom. … To combat cultural issues (such as fiefdoms) and bring a focus to safety within any given organization, it is important to ensure that the reporting structure allows for communication of safety information to those within the organizational hierarchy that have the authority and resources to implement safety change."

The report's conclusions and recommendations, expected in spring 2014, may move the nation's universities toward better safety practices—or not. Academies reports can, on occasion, strongly affect perceptions and policies. For example, Rising Above the Gathering Storm, published in 2007, crucially advanced a narrative of American scientific decline and inspired a strong legislative response.

On the other hand, most studies languish in a wonky Washington netherworld of forgotten policy pronouncements. Given the risks that students, postdocs and other lab workers now face every day in countless academic labs across the country, here's hoping that this committee's efforts prove to be bold, effective, and consequential.

Beryl Lieff Benderly writes from Washington, D.C.

10.1126/science.caredit.a1300120