EU legislation would have stopped workers getting close to MRI scanners during operation. Credit: Corbis/Photolibrary

A controversial piece of European Union (EU) legislation that outlaws much routine brain and body imaging in research laboratories and clinics is on course for a revamp.

The EU directive was designed to limit workers' exposure to electromagnetic fields (EMF) and was approved in 2004. But it was put on hold in 2007 after scientists and others said that it would stop them from carrying out important work without improving worker safety.

The European Commission now says that it will propose new legislation by the end of this year that addresses these issues, and will allow magnetic resonance imaging (MRI) use to be exempted from binding exposure-limit values. "We know this is only a proposal," says Stephen Keevil, a medical physicist at King's College London. "But if the final directive turns out to be like this, then everything will be fine."

The 2004 Physical Agents (EMF) Directive was originally designed to address health risks for those working in the electrical power and telecommunication industries, but it turned out to have a range of unintended consequences. Many sectors — medical, industrial and even military — claimed that the directive set occupational exposure limits that were overcautious and that would interfere with normal activities, including routine imaging, which is vital in both the clinic and the research lab. They also disputed the scientific data on which the limits had been based.

The protests earned the directive the dubious distinction of being the only piece of EU legislation ever to have been put on ice. The original directive was scheduled to be incorporated into national legislation by 30 April 2008 — member states are normally obliged to implement EU directives within four years of their approval — but after the complaints this was delayed until April 2012 to allow time to consult stakeholders and, potentially, to propose revisions. The European Parliament and Council must sign up to any changes the commission proposes to make.

If the final directive turns out to be like this, then everything will be fine. ,

Even with those extra years, the timetable for repair is tight. "We hoped to have our proposal for new legislation much earlier, but it all proved complicated," admits Georges Herbillon, the commission official responsible for drafting the new rules. "Our current timetable gives 16 months for the political approval and that should be enough." But the commission is already behind schedule and, he says, much further slippage would mean that the original directive would automatically come into effect on 1 May 2012.

The directive covers the electromagnetic spectrum up to 300 GHz, the highest frequency of radio waves, but the disputes focus on the low-frequency range under 100 kHz, typically found in the vicinity of standard MRI machines. Many procedures require health or research workers to remain close to the scanner when it is running, to carry out procedures on people inside the scanner, for example. Decades of experience have produced no evidence of permanent adverse health effects, MRI scientists argue. Physical effects, such as a feeling of vertigo, or peripheral nerve stimulation resulting in a harmless twitch, may occasionally occur, but they are temporary.

The original directive banned workers from staying close to a running scanner. The calculations used to estimate safe levels of exposure were based on recommendations of the International Commission on Non-Ionizing Radiation Protection (ICNIRP), a non-governmental organization recognized by the World Health Organization.

But a report released this week by the European Science Foundation says that those recommendations are extrapolated from very limited experimental data with excessive caution. "It is nonsense — taking the precautionary principle too far," says Denis Le Bihan, director of the NeuroSpin neuroimaging project at the CEA-Saclay Centre near Paris, and a member of the expert group that prepared the report.

In its latest consultation document, the commission says that when defining adverse health and safety effects it will take the specific circumstances into account — a transient feeling of vertigo may be dangerous only if the worker is standing and at risk of falling. It also plans to consider safety calculations from agencies other than the ICNIRP, such as the international Institute of Electrical and Electronics Engineers (IEEE) and Germany's Federal Ministry of Labour and Social Affairs (BMAS), which some researchers say are more strongly grounded in science. It suggests grading the disputed sub-100 kHz frequency range into three categories on the basis of a risk assessment that acknowledges the higher safety limits recommended by the IEEE and BMAS, banning access only when exposure is high.

MRI would be exempt from these binding exposure limits. Instead, a Europe-wide agreement on safe working practices would be developed and workers would be given appropriate training. This, Keevil says, will avoid the problem of impeding research or medicine, but "will achieve the same aim of keeping people safe".

figure a