Staying healthy and safe in the laboratory can be tedious, inconvenient, uncomfortable--even unfashionable. Forms must be filled out. Bottles must be properly labelled. You have to wear ugly lab coat and goofy glasses--and you may not be permitted to wear those nifty open-toe sandals you bought last week.

But as anyone who has ever had a laboratory accident can attest, it's more than worth the trouble. Accidents happen. Chemical reactions explode. Lasers are used irresponsibly. Disease organisms escape. When they do, you have to be ready. "You have to make common sense judgements about what could happen," says Jeremy White, deputy head of safety at University College, London.

No regulation can guarantee your safety; the only way to stay safe is to be vigilant. Carry out regular safety checks to see if your chemicals, equipment, or surroundings could be hazardous and fill in a risk assessment each time you do something new. Don't work alone--or if you must, check in on a regular schedule with friends, family, or colleagues. It sounds like a lot of trouble, but if you don’t work hard at staying safe, someone will get hurt.

White speaks from experience. In the late 1970s, when he was already his department's safety officer, a colleague knocked over a large glass vessel that was stored on the floor as they were making up cell cultures together. His colleague slipped and found herself sitting on the floor in a pool of spilled liquid. It was glacial acetic acid, corrosive enough to send her to hospital with skin burns, but not a serious, long-term health risk. On the same floor at the time were two other bottles, one containing harmless glycerol, the other concentrated sulphuric acid. "Her injuries would have been much more severe if she had knocked over the sulphuric acid," White observes.

White was horrified by the incident because he had placed the bottles on the floor deliberately, thinking they would be safer there. "If I had looked around the lab systemically, spotting hazards, these bottles would have been deemed inappropriately stored," he says. "They should have been in a metal storage bin. Or we should not have purchased hazardous materials in such large quantities." White had decided, however, that they were better off on the floor than on some high, open shelf. He was probably right, but his attempt at making things safer was nonetheless inadequate.

Formal risk assessments are required in Europe since the 1989 Framework Directive for Health and Safety at Work , and are now implemented in all the member states. The law also requires that all staff be trained to avoid the risks identified in those formal risk assessments. For example, every laboratory utilising radioactive substances must adopt one code of practice for routine handling and another for what to do in case of an accident. Other serious risks, such as infectious agents and toxic chemicals, are governed by separate, more specific European Directives.

Be aware

But regulations are no substitute for vigilance; even with regulations in place, hazards can be overlooked. "Most accidents involve people doing something they don’t normally do," says Dave Kinnison, chemistry-safety adviser at Southampton University. His mantra is 'never do anything new without a risk assessment.' Last year, a chemistry Ph.D. student at Southampton was trying a new reaction when the glass tube exploded, blowing out the glass sash of the fume cupboard and scattering debris metres across the lab. Fortunately, the student was not standing nearby at the time and no one was hurt, but the damage was extensive.

"This student hadn’t followed procedures," says Kinnison. "If he had looked at the Materials Safety Data Sheet, he would have seen that the chemicals he was using, ethyl formate and a strong base, are incompatible and potentially explosive when combined in an enclosed system." The procedures he should have followed, described in a U.K. document known as Control of Substances Hazardous to Health (COSHH), have been in place for nearly 20 years. "We have about 160 postgraduate students in this department, many of them non-native speakers, and 50 new ones arriving every year," says Kinnison. "Our challenge is to train them well enough in safety procedures that accidents don’t happen."

Don’t cut corners

Shortcuts in the lab are often tempting but they're rarely a good idea. A student in a physics and astronomy department at another British university paid dearly for his impatience in 2004. "This student was near the end of his Ph.D., in a rush to get results," says the safety officer who reported the incident. "He was using the laser to ionise a target in a vacuum chamber." But the vacuum chamber was not designed to allow lasers in; it was surrounded by metal mesh and intended for thermal ionization, in an oven. Thermal ionization wasn't working, so the student decided to use a laser; just a quick extra step to be sure that ionization was occurring. He cut a hole in the metal mesh and tried to direct the laser beam through it by moving the laser itself. "Normally, you would turn the power down and align the beam using two mirrors, with the laser in a permanent position," says the safety officer. "And of course you would be wearing protective eye wear." The student did none of these things. A full power, pulsed laser beam hit the metal mesh, bounced back, and struck his retina.

The accident was even more unfortunate in context. "We had recently had a laser safety inspection from the Health and Safety Executive, which caused a shake-up in our procedures," the safety officer says. "We had written risk assessments for nearly all our activities, and run a training scheme for all laser users. The whole department [was] shocked and disappointed that this could happen after all our effort."

Never assume you’re protected

Believing that you’re safe may itself be a safety risk. In 2002, in Munich, a technician who had been working with a genetically modified version of the Vaccinia virus for many years became infected. The man had injured the skin of his hands changing car tyres in cold weather and "should have been wearing plastic gloves," says Heidelore Hoffman, the dermatologist who reported the case, "but he thought he was protected because he had been vaccinated against smallpox as a child." Smallpox vaccines, which were administrated until the 1970s, used to be made from the Vaccinia virus, but the vaccination didn't protect him from infection. Hoffman and colleagues speculate that the virus overrode the man’s immunity because it had been genetically modified.

In any case, Hoffman argues that all scientists working with Vaccinia should be vaccinated, as they are in America. In non-vaccinated people, Vaccinia can cause generalised infection with a high fever. In this patient, fortunately, the infection was localised and he recovered. "My patient was very worried by the infection," says Hoffman. "Since then he has strictly followed the lab rules, and warned his colleagues to do the same."

When your experiments are outdoors

In a laboratory environment, risks can be well defined and prominent health-and-safety labels decorate chemicals and equipment. But outside the lab, all manner of unexpected dangers lurk. The best way to minimise the risk is never to work alone, even for a moment.

Clarisse T*, an agricultural researcher, suffered a bad fall that she believes was avoidable. "We were aerating pig slurry in very large tanks, on a farm out in North Norfolk," she explains. While T was working at the top of the tanks, perched up on a ladder, her supervisor left to use the toilet. As she went to descend, the ladder fell and T was left hanging on the rail, 4 metres up, shouting for help. "In the end I had to let go and fall," she says. T fractured her heel bone. "It hurt like hell." The incident was investigated by the health and safety executive and the supervisor was not blamed; the verdict was that the ladder should have been fixed. But T believes that she should not have been left alone. "My supervisor should have told me he was going and I should have come down and waited."

Europe has no specific regulations about working alone, unless being alone is identified as a risk in a formal risk assessment. But many organisations, including universities and research councils, treat lone fieldwork as a substantial risk that should be avoided if possible. Nevertheless, "In Spain, it is common practice for researchers to work alone in the field," says Pedro Rey, a plant ecologist from the University of Jaén, "because we do not usually get public funds for a field assistant." Last July, a colleague of Rey’s, José Luis Garrido, suffered a severe accident that illustrated just how much of a risk working alone can be.

More resources:

Guide to European health and safety legislation

Garrido was working in the mountains of El Bierzo in northern Spain on a Saturday afternoon, collecting fruit to measure pollination success. "I took a break to have my packed lunch," he describes, "and then I can’t remember what happened." Somehow, Garrido fell several metres down the mountain. He came round, covered in blood, embedded with stones, and with severe fracture injuries to both legs. "I could hear my mobile phone ringing up the mountain all through the first night, until the battery went dead." Garrido was not found for 3 days and the open fractures became infected. Now, a year and multiple surgeries later, Garrido is in a wheelchair with one leg paralysed from the knee down. His other ankle has a chronic bone infection and is no longer flexible because the shattered bones had to be repaired with transplants.

Garrido's accident rattled the Spanish ecology community. "We feel now that our fieldwork is much more dangerous than we believed," says Rey, "and we do not receive appropriate support from universities or public funds." So the alternative, according to Carlos Herrera, an ecologist at the Doñana Biological Research Station in Spain, is to make remote, lone fieldwork safer. Members of his team, he says, are required to tell other staff exactly where they will be and to phone the field station at the end of each day’s work to let others know that they are safe.

In all these cases, the outcome would have been better if the scientists and their colleagues had taken the time to fully appreciate the risks. "The modern approach to health and safety is that you only need to act when there is a risk," says Mark Dickinson, a safety officer from Manchester University. But you must regularly remind yourself of the risks. There is peril in activities you are unaccustomed too, but there is also risk--complacency--in the familiar. So be on your guard. Take safety seriously, and you will stay safe.

* names have been changed to protect privacy

Comments, suggestions? Please send your feedback to our editor.

Lynn Dicks is a science writer and editor based in the U.K.