Every day hundreds of thousands of doctors and patients around the world discuss the benefits and risks of drugs. You might think therefore that we know how to communicate the information well, but the European Medicines Agency (EMA) and the Food and Drug Administration agree that we don’t.

Indeed, the EMA logically thinks that before we can communicate well we need to reach an agreed definition on what we mean by benefits and risks and determine the best methodology for measuring and describing them.Research among 42 assessors of drugs showed that there was no agreed definition of benefits and risks but rather multiple definitions. There was no agreed systematic approach, but there was agreement that it was difficult to define and measure benefits and risks.

Should definitions be qualitative, quantitative, or both? Should there be explicit criteria for both benefits and risks? How can patients’ values best be incorporated? How might definitions relate to the severity of the conditions being treated? Patients will accept much more risk in treatments for serious, life threatening illnesses. How can definitions deal with the gap between efficacy (as measured in clinical trials) and effectiveness (what happens in the real world)? And what about drugs being used off label, when there may be risks but no benefits? What is the best way to acknowledge uncertainties? How can doctors explain to patients that only 20% of them may benefit from a treatment and that we have no way of knowing whether any particular patient is in the 20% or the 80%? How can we best describe the huge uncertainties around the benefits and risks in the many patients who have multiple comorbidities?

The EMA is tackling these questions with a series of activities and hopes to have some answers within about 18 months. But what might be the role of drug companies in communicating benefits and risks? This question was discussed at a meeting I attended at the Royal Society last week.

Drug companies are understandably perplexed about their role in communication in Europe. They know more about their drugs than anybody else, but they are not allowed to communicate with patients—even if a patient contacts them with a specific question. Politicians in Europe are unyielding in their refusal to allow drug companies to communicate with patients—fearing bias or inflation of their drug budgets.

A particularly tricky issue arises over periodic safety update reports (PSURs), which drug companies are required to present to the EMA. These may be huge documents and give detailed, patient level information of new reports they have received on adverse drug reactions and any other new information that bears on the safety of their product. Most doctors and patients don’t know of the existence of these PSURs, but some groups have found out about them and want to know more.

The motivation to know more might be either a hunger from patients for all relevant information or a fear that something is being hidden. Various groups in Europe have asked to see these reports under freedom of information acts, and some redacted versions have appeared on websites in the Netherlands. (We in Britain learnt the word “redacted” through the scandal of MPs’ expenses: it means that some information, in this case patient details, is blacked out.)

As they stand these PSURs are not good for communicating benefits and risks for patients, not least in that they concentrate on risks and say little about benefits. But could they be made useful? This remains an open question, but most people at the meeting thought that better ways could be found to communicate benefits and risks—particularly after the EMA has completed its work on definitions.

There were disagreements at the meeting, but everybody seemed to agree that many medicines are not used rationally and that it’s in the interests of everybody—patients, clinicians, regulatory authorities, health care funders, and drug companies—that they come to be used more rationally. Better ways of communicating the benefits and risks of medicine will be one way to help achieve that aim.

Competing interest: The meeting was funded by Roche, but I received no fee or expenses. I did, however, have two cups of coffee, two cups of tea, and a tasty lunch, for which I thank Roche. If I was an employee of the US government I wouldn’t have been able to accept the lunch, but I can’t say I feel corrupted.