Lab safety and research productivity are not at odds | Science

Prioritizing lab safety doesn’t hamper research productivity. That’s the main takeaway of a working paper published last week by the National Bureau of Economic Research and accepted for publication in Research Policy. Safety experts are cautiously optimistic these results may help support further cultural and institutional change to address poor laboratory practices in academia.

“Doing research safely doesn’t impede science. And scientists need to see that,” says Craig Merlic, an organic chemist at the University of California, Los Angeles (UCLA) and executive director of the UC Center for Laboratory Safety who wasn’t involved in the study.

In recent decades, several laboratory disasters have killed or grievously injured young scientists, sparking outrage. Still, academic science has generally remained stubbornly blasé about lab safety. Some see it as directly at odds with getting work done in the lab.

“There’s this perception that the more safety requirements you put on a laboratory, the lower the research output is going to be,” says study co-author Alberto Galasso, an economist at the University of Toronto. A 2012 survey suggested that many scientists see laboratory safety as a “hassle.” This can often lead to unsafe labor practices. For instance, a 2010 survey by the American Chemical Society found that 70% of faculty, 59% of professional staff, and 52% of graduate students work alone in laboratories “often” or “occasionally,” which safety experts advise against.

To investigate whether there is in fact such a trade-off between productivity and safety, the researchers looked into publication patterns before and after a tragedy that created a large-scale, drastic change in how people view and practice lab safety. That happened on 29 December 2008, when 23-year-old Sheharbano “Sheri” Sangji died after a reagent she was using in the UCLA chemistry lab where she worked burst into flames.

In response, UCLA and the entire UC system drastically shifted its safety policies. UCLA inspected all labs—the first time they had conducted such an institutionwide effort, as opposed to ad hoc inspections of individual labs, says Imke Schroeder, a research project manager at the UC Center for Laboratory Safety, which was established after Sangji’s death. The university system also enforced more stringent rules for protective equipment and chemical use. Some of the shift was cultural, Schroeder adds. When Sangji’s lab head was criminally prosecuted for the incident, it indicated that professors might be personally responsible for accidents in their labs and they began to see laboratory safety as more important.

Amid those changes, publication productivity for 600 UCLA chemistry labs wasn’t affected, the authors of the new study found based on records from the Web of Science, a bibliographic database of scholarly articles. To control for universitywide factors that might have affected research output, the team looked at both experimental “wet” labs and computational “dry” labs, reasoning that dry labs wouldn’t be as affected by changes in safety policies. Using the preshock publication rates from 2004 to 2008 as a baseline, the authors found that relative to dry labs, wet labs on average published studies at the same rate between 2009 and 2017 as they had before the shock.

UCLA’s stricter policies made a difference in determining which chemicals labs used in their studies, however. Some labs that were previously using hazardous chemicals shifted to using safer ones after 2008, the authors determined by analyzing data from SciFinder, a proprietary chemical database that documents all of the chemical compounds associated with a journal article. But only a minority of labs use such reagents in the first place, the authors point out, and only a tiny fraction of all labs showed this effect.

The study could help build the case for increasing safety and academic labs, says chemical safety expert Frankie Wood-Black, division chair of engineering, physical science, and process technology at Northern Oklahoma College. “In industry settings, we heard these exact same arguments in the ’80s and early ’90s” regarding productivity, she says, but studies like this helped persuade industry executives to enact safety measurements that are now more stringent than those in academia.

Not everyone is so optimistic. The study is useful, says James Kaufman, founder and president emeritus of the Laboratory Safety Institute, a nonprofit educational organization for safety and science, but he’s not sure it will persuade researchers to change the way they act. “I think in general, people might be unmoved,” he says. In his experience, when it comes to changing risky behaviors, “people tend not to be swayed by facts.”

Regardless of whether the paper changes minds, safety experts agree that it’s only one part of creating structural change in academic laboratory settings. Grant funding agencies, professional associations, and universities need to start considering safety and their hiring and awarding decisions, Kaufman says. “Working safely has to be an essential condition of employment and learning.” He adds that change must come from educating young students, so they progress through academia in a culture of safety.

“At the end of the day, scientists need to understand that good science is science done safely,” Merlic says. “Safety isn’t going to impede research.”

Prioritizing lab safety doesn’t hamper research productivity. That’s the main takeaway of a working paper published last week by the National Bureau of Economic Research and accepted for publication in Research Policy. Safety experts are cautiously optimistic these results may help support further cultural and institutional change to address poor laboratory practices in academia.

“Doing research safely doesn’t impede science. And scientists need to see that,” says Craig Merlic, an organic chemist at the University of California, Los Angeles (UCLA) and executive director of the UC Center for Laboratory Safety who wasn’t involved in the study.

In recent decades, several laboratory disasters have killed or grievously injured young scientists, sparking outrage. Still, academic science has generally remained stubbornly blasé about lab safety. Some see it as directly at odds with getting work done in the lab.

“There’s this perception that the more safety requirements you put on a laboratory, the lower the research output is going to be,” says study co-author Alberto Galasso, an economist at the University of Toronto. A 2012 survey suggested that many scientists see laboratory safety as a “hassle.” This can often lead to unsafe labor practices. For instance, a 2010 survey by the American Chemical Society found that 70% of faculty, 59% of professional staff, and 52% of graduate students work alone in laboratories “often” or “occasionally,” which safety experts advise against.

To investigate whether there is in fact such a trade-off between productivity and safety, the researchers looked into publication patterns before and after a tragedy that created a large-scale, drastic change in how people view and practice lab safety. That happened on 29 December 2008, when 23-year-old Sheharbano “Sheri” Sangji died after a reagent she was using in the UCLA chemistry lab where she worked burst into flames.

In response, UCLA and the entire UC system drastically shifted its safety policies. UCLA inspected all labs—the first time they had conducted such an institutionwide effort, as opposed to ad hoc inspections of individual labs, says Imke Schroeder, a research project manager at the UC Center for Laboratory Safety, which was established after Sangji’s death. The university system also enforced more stringent rules for protective equipment and chemical use. Some of the shift was cultural, Schroeder adds. When Sangji’s lab head was criminally prosecuted for the incident, it indicated that professors might be personally responsible for accidents in their labs and they began to see laboratory safety as more important.

Amid those changes, publication productivity for 600 UCLA chemistry labs wasn’t affected, the authors of the new study found based on records from the Web of Science, a bibliographic database of scholarly articles. To control for universitywide factors that might have affected research output, the team looked at both experimental “wet” labs and computational “dry” labs, reasoning that dry labs wouldn’t be as affected by changes in safety policies. Using the preshock publication rates from 2004 to 2008 as a baseline, the authors found that relative to dry labs, wet labs on average published studies at the same rate between 2009 and 2017 as they had before the shock.

UCLA’s stricter policies made a difference in determining which chemicals labs used in their studies, however. Some labs that were previously using hazardous chemicals shifted to using safer ones after 2008, the authors determined by analyzing data from SciFinder, a proprietary chemical database that documents all of the chemical compounds associated with a journal article. But only a minority of labs use such reagents in the first place, the authors point out, and only a tiny fraction of all labs showed this effect.

The study could help build the case for increasing safety and academic labs, says chemical safety expert Frankie Wood-Black, division chair of engineering, physical science, and process technology at Northern Oklahoma College. “In industry settings, we heard these exact same arguments in the ’80s and early ’90s” regarding productivity, she says, but studies like this helped persuade industry executives to enact safety measurements that are now more stringent than those in academia.

Not everyone is so optimistic. The study is useful, says James Kaufman, founder and president emeritus of the Laboratory Safety Institute, a nonprofit educational organization for safety and science, but he’s not sure it will persuade researchers to change the way they act. “I think in general, people might be unmoved,” he says. In his experience, when it comes to changing risky behaviors, “people tend not to be swayed by facts.”

Regardless of whether the paper changes minds, safety experts agree that it’s only one part of creating structural change in academic laboratory settings. Grant funding agencies, professional associations, and universities need to start considering safety and their hiring and awarding decisions, Kaufman says. “Working safely has to be an essential condition of employment and learning.” He adds that change must come from educating young students, so they progress through academia in a culture of safety.

“At the end of the day, scientists need to understand that good science is science done safely,” Merlic says. “Safety isn’t going to impede research.”