At Yale University this past week, Michele Dufault, a very bright and vivacious 4th year physics/astronomy major, was working alone at night in the machine shop for Chemistry students when her hair became entangled in a powered lathe. She was found around 2:30 in the morning, still sitting at the machine, but not breathing. She had died of asphyxiation produced by neck compression.
Accidents such as this that lead to the death of a student, are rare—and in many ways that’s surprising. The practical work of students, whether in the sciences, engineering, or the arts, often involves dangerous tools and materials. Something as simple as a propane line for a bunsen burner can provide the means to blow up a wing of a building. The same goes for hydrogen, which blew up a biochemistry lab at the University of Missouri last June, injuring a graduate student, two post-docs, and a staff scientist. When I was a post-doc at the University of Toronto, they evacuated a wing of the Metallurgy department while the bomb squad took out a container of benzoyl peroxide that had been left for way too long on a shelf immediately over a PhD student’s desk. And now, in my Biomedical Engineering lab, our students routinely use servo-hydraulic testing machines driven by 200 atmosphere oil pressure and capable of exerting 10 tonnes of force. Physics, biochemistry, and medical imaging all use radioisotopes and other radiation sources capable of causing significant pathologies or death. Even students in fine arts, theatre, and architecture, can find themselves working long hours with power tools, solvents, ladders and ropes, or molten glass and metals.
Universities do a variable job of training students in safe practices for the techniques that they will employ. Training can be very good and readily available around common laboratory practices—and departments and supervisors do well to make attendance mandatory. (For instance, my two post-docs are spending all this coming Tuesday in a Dalhousie-sponsored laboratory safety course.) In such courses, one can be taught the ins and outs of dealing with hazardous chemicals, biohazards, radioisotopes, experimental animals, electrical hazards, etc. And, where necessary, trainees can receive more specialized training in the operation of dangerous equipment, preventative measures, and emergency procedures. On this sort of front, it seems to me that universities do a pretty good job.
There’s another sort of problem, though, that is especially acute for research trainees working semi-independently at the leading edge of their fields. Students get kudos for trouble-shooting problems and for finding workarounds when existing methods don’t serve their needs. Unfortunately, there’s a fine line between innovating in development of a new technique and kluging together new apparatus or procedures that are under-engineered for safety of the operator—if properly planned out at all. Generalized lab safety training can’t be expected to anticipate something that will be done for the first time. What’s more, the stories of great innovators are rife with examples of new approaches that went badly wrong but gave serendipitous results. For example, polyethylene was discovered when a reaction failed in a chamber under 1700 atmospheres of pressure! When universities are in the business of innovating, doing things for the very first time, can they truly be expected to anticipate the risks that may ensue for their faculty, staff or trainees?
One place to look for an answer is in industry where the record of safety is remarkably good given the magnitude of the enterprise. Industrial safety experts argue that universities do a poor job of creating a safety-first culture. This from the Connecticut Post today, with reference to Ms. Dufault’s death:
"'If that very same lathe was in an industrial setting, she would have been required to have much shorter hair,' said safety expert Russ Phifer, chairman of both the American Chemical Society's Division of Chemical Health and Safety and the ACS Committee on Chemical Safety.
'It was a sad learning experience,' Phifer said. 'The corporate safety culture is 10 times what it is in universities, but part of that is the nature of the university lab work — you have a variety of work being done in university labs by a variety of different people.'
Still, serious laboratory accidents at universities are "few and far between" according to Ted Fitzgerald, a spokesman for OSHA's Boston regional office. "This is an unusual event."
On Dec. 29, 2008, UCLA research assistant Sheri Sangii was extracting a tiny amount of t-butyl lithium, a "pyrophoric" compound, meaning that it ignites instantly when exposed to air.
The plastic syringe she was using ruptured, and the resulting flash fire left her with severe burns over 43 percent of her body. She died 18 painful days later, and the subsequent investigation revealed that, among other safety lapses, she wasn't wearing a flame-resistant lab coat, but rather a highly flammable acrylic sweater. Two months earlier, a UCLA safety inspection cited the lab for its lack of protective clothing."
In the case of Ms. Dufault, how good was the safety training? Apparently, very good indeed. In the Yale News from yesterday, guest columnist John Scurdato who was a classmate of Ms. Dufault in her machine shop safety course wrote:
"The training program for the Sterling Chemistry shop is actually a full-semester course. It is the same course for both undergraduates and graduate students. When I took the course, Dave Johnson, the shop manager, trained students for four hours a week in basic shop safety and machining techniques. Students built working steam engines from blocks of brass, aluminum and steel using lathes and mills. Throughout the process, Dave always stressed safety, and safety goggles were an absolute must. He repeatedly told long-haired individuals they must tie up their hair and warned of the danger of getting one’s hair caught in a lathe. Dave secured the shop door when he left at 4 p.m., and after-hours access was only provided to the graduates of the shop training class. These graduates were told not to work alone after hours."
So what happened? If the training was excellent, was the accident the result of carelessness, willful ignoring of the rules, lack of supervision? Perhaps. No doubt a full investigation will yield some answers. However, given that the accident occurred at night, and it was the end of term, another possibility is fatigue. In a piece this past Thursday for the Channel 4 News in Belfast, Tom Clarke argues that a great deal of university research work has gone nocturnal with adverse consequences for the trainees. He notes that “(i)n the strange and often lonely world of the graduate science student it is often common practice to be working late and alone.” That in itself may be an invitation for trouble. Graduate students and post-docs often work alone, late at night, on just the sort of untested, leading-edge developments that I mentioned above. In my own lab I’ve insisted on a buddy system when working with dangerous hardware or materials out of hours but how can one fully police compliance? In some really big labs that I know of, bench space can be so valuable a commodity that research trainees do the equivalent of hot-bunking: working alternate daytime/nighttime shifts at the same bench space. Given the pressures that already exist in high output labs, this doesn’t make for well-rested trainees. It’s well understood how fatigue can lead to poor judgement and treatment by medical interns and residents, but the equivalent phenomenon in research trainees hasn’t to my knowledge gotten the same attention.
When accidents like that of Ms. Dufault’s happen in places like university machine shops—especially after hours—some unsurprising institutional responses occur. Administrators, faculty, and students all begin to ask questions about how it could have happened. The shop is closed (at the least temporarily), the safety training and supervisory measures are reassessed, and everyone begins to ask questions about their insurance status. One reaction can be the urge to reduce or eliminate student access to facilities like machine shops on the premise that, had the victim not been in there, this couldn’t have happened. Should such bans become commonplace, it would be a great pity. Over the years, universities have lost more and more of facilities like machine shops and our graduates have emerged poorer in the practical skills that enable the art and craft of high-level endeavour. It’s my personal experience that science and engineering graduate students are becoming weaker year-over-year in the ability to design and fabricate experimental apparatus. And where students need support to develop new gear, the university shops are fewer, busier, and less available for student training and hands-on experience. [Don’t even get me started on the need for a new generation of skilled machinists and other technologists!] In my own lab these days, we have most of our apparatus fabricated by the son of a local surgeon who abandoned computer science at university to set up a machine shop in his basement.
In the Yale Times, John Scurdano makes a cogent case for saving the very machine shop where Ms. Dufault died:
"Many have asked why we have (the) machine shop, given the danger. But for a little-known community of Yale scientists and engineers, building new designs is our theater, our campus publication, our YPU. Imagine a Yale where students did not apply knowledge outside of the classroom. Think of art programs where no one made sculptures or a theater program that did not build sets or put on plays. There is nothing so rewarding as seeing an idea or a lesson translate into a tangible object. The Yale shops provide that experience.
Shops are the lifeblood of many student groups. Michele was a member of the Yale Drop Team, a multidisciplinary team that has run multiple experiments in microgravity aboard NASA airplanes. Most people would never have the opportunity to experience near-weightlessness, but Michele and her team did because of their work in the shop. Other student groups like the Society of Automotive Engineers fabricated a hybrid racecar from scratch with Yale’s facilities. Every student involved in these projects puts dozens of hours a week into them. They are the invaluable, often once-in-a-lifetime experiences for which so many of us came to Yale."
For the sake of our students and research programs, I hope that we don’t collectively run scared in the face of tragedies like that at Yale last week. Universities need to make every effort to have safety training as an integral part of their operations. But innovation will always mean doing things for the first time. That’ll be dangerous at times. Let’s manage the danger, not hide from it.
—Mike