EFSA confronts the tricky questions of transparency

EFSA’s stakeholder conference on transparency made one thing clear: there are no simple answers to questions about openness. The meeting – attended by senior EFSA staff, scientists, industry representatives, non-governmental organisations, international lawyers, and other stakeholders – debated a range of themes and questions, such as: does openness mean publishing all data? How should uncertainty be reflected in EFSA’s conclusions? And should more of EFSA’s scientific meetings be open to observers? Get a taste of the discussion by reading our report of the day’s main talking points.

Tim Radford, freelance journalist and former science editor of The Guardian
There is nothing clear about transparency. In governmental terms, it means openness. Transparency in risk assessment sounds like a straightforward ambition for the European Food Safety Authority: if the public understands how scientists assess the risks that may be involved in additives to food, or to animal feed, or in novel foods, then there will be greater trust in EFSA and clearer guidelines for industry to follow. Everybody wins.

But a one-day meeting in Parma, between EFSA chiefs, scientists, industry representatives, non-governmental organisations, international lawyers, journalist commentators and other stakeholders has made one thing clear: there are no simple answers to questions about openness.

There is already information about food safety investigations in the public domain, but how much more should there be, and in what form, and how quickly? Does openness mean summaries of key decisions, or full minutes of all meetings? Should more meetings be open to observers (15 such meetings have been open to stakeholders so far in 2013)? Should EFSA live-stream all its meetings on the Web, or record discussions for the public to download? Should EFSA explain in all cases how it arrives at particular risk assessments? Does openness mean publishing all data, presented in a way that would allow anyone to replicate the study, or re-interpret the findings, or use them for another purpose? Would such public disclosure actually make things more difficult for EFSA, or for the food industry? And would later challenges to EFSA conclusions, by third parties, reassure the public or just confuse the consumer?

There are direct problems even with simple disclosure. Industry may wish to demonstrate to the public, as openly as possible, that food products or additives or flavourings or colourings under examination are safe, but companies would be less willing to open their data and reveal confidential commercial information to competitors. Would the equivalent of a private “reading room” suffice for sensitive information and, if so, would that provide enough data for the scientific community to examine the tests carried out? And should EFSA’s risk assessors declare all their interests and should these be available for at least five years after they have left the organisation?

There are wider questions, about the mandates from the European Commission’s DG-SANCO to EFSA and about how risk assessment policy is established before individual risks are assessed. Should EFSA more often decide for itself which issues or topics it should address? There are problems about how industry approaches EFSA in the first place: industry needs to know what and which data must be produced before a submission can be placed, but should these pre-submission meetings be recorded and made public? If such meetings are not transparent, could there be a diminution of public trust? Should there be open workshops after submissions, involving NGOs as well as consumer groups and industry? Is the EFSA Applications Helpdesk an open office to help applicants and speed the progress of approvals, or a kind of firewall? Should the Applications Helpdesk’s role be transparent and its rules open to all scrutiny? It may tell applicants what kinds of studies and data may be necessary but it must not indicate that these would be sufficient, so perhaps this front line operation itself requires an explicit protocol to determine its scope? Should applicants meet EFSA staff at a personal level at all, on a one-to-one basis, or should such questions and answers be submitted in writing, and placed on public record for comment by other stakeholders? Should scientific meetings be closed and confidential? Or should differences of opinion and arguments about interpretation be live-streamed or recorded, to help the public and other scientists understand how decisions were reached?

Having obtained and published all the data and secured the full cooperation of industry, the Authority still has the problem of getting the information to the public in a widely understandable form: to be credible such information has to be comprehensible and there are, of course, difficulties with the precise language of law or the specialised lexicon of science. The applicants need to frame their information in one form, but civil society needs it in another. Manufacturers and consumers, scientists and legal advisers, legislators and pressure groups need to have some understanding of what data were included in making a decision, what were not included, and why. There needs to be some mechanism for dialogue, because people may have questions, or complaints. There needs to be clarity about uncertainty, and how any uncertainty is reflected in EFSA’s conclusions, and there would be a good case for summaries specifically designed for a lay audience, perhaps in question and answer form, but the message of these summaries would have to be consistent with informed science.

There would be a good case for publishing a detailed glossary of the acronyms, abbreviations and specialist terms used by nutrition and food science professionals, lawyers and EU authorities; there would also be a strong case for making clear, in all public discussion, the difference between risk assessment and risk management; and there are would a good argument for special video presentations to help wider public understanding of topical issues. Finally, openness or transparency is not just a passive condition: EFSA might need to reach out to some professional communities not yet aware of the Authority’s importance, and it should actively engage with the EU’s member states.

All these are questions and suggestions framed by three break-out groups set up to consider submissions outlined and challenges posed at the beginning of the one day conference.

In the next year, EFSA plans to finalise a new policy on openness and transparency: Bernhard Url, Deputising Executive Director of EFSA, reminded the conference that transparency was intimately linked to trust in the risk assessment process and this had been recognised at EFSA’s creation. “But of course, the operating context has evolved and so has EFSA,” he warned. In the coming years the volume of scientific data would increase, together with increasing demand for access. It was important that risk assessment policies were fully understood, and to be effective EFSA needed a full understanding of the information needs of its stakeholders, which was why the Authority was redoubling its efforts to make its communications more meaningful and more accessible to all parties.

Sue Davies, Chair of the EFSA Management Board, listed what she described as areas for improvement: most meetings of Scientific Panels were not open to public observers; opinions needed to be updated with the arrival of new evidence (or fresh criticism). There had to be greater clarity about the sources of data, and the handling of conflicting data, and about how data were collected, who “owned” the data and who had access to it. The whole point of the meeting was to invite stakeholders to help identify the priority areas for action.

Transparency has become embedded in US and European democratic ideals, according to Alberto Alemanno, a professor of EU law at HEC Paris and at New York University. “Sunlight is the best disinfectant,” he quoted, and then added: “How could anybody be against transparency?” But the principle of openness – the right of access to documents – and its rationale (to enhance the legitimacy of EU action by promoting active engagement) had to be reconciled with EFSA’s founding regulations, questions of commercial sensitivity, and with existing procedures. There were constraints to overcome: but transparency and in particular public participation could also help EFSA solve allegations of conflict of interest; open up new avenues of accountability; contribute to open, efficient and independent public administration and turn EFSA (and other EU agencies) from mere providers of technical information into managers of information platforms.

To a certain extent, some of this is already happening. Per Bergman, Director of Scientific Evaluation of Regulated Products, saw transparency as a “game changer” for EFSA, and revealed there had been a seven-fold increase in data requests; page views of the EFSA Journal were running at the rate of 200 an hour, or 5,000 a day, and opened the way to scientific advice at every stage of the food chain, on matters such as plant health, plant protection, genetically modified organisms, animal feed, animal welfare, biological food chain hazards and contaminants, food additives, flavours, processing and packaging, and issues around health claims and novel foods. There was, he said, a timetable for action, and the stakeholder conference would be followed by a draft policy, a period of public consultation some time during April and May 2014, and submission of a draft policy document to the EFSA management board by December 2014.

That was the view from within EFSA. Ruth Veale of the European consumer organisation BEUC was keen on transparency and argued that information about food-related risks was already far more accessible and understandable and provided the basis for wider communication to Europe’s citizens through the member states. But there was a clear need for the risk assessment process to be independent, transparent and free from the influence of industry. She also thought the distinction between risk assessment and risk management now meant that there was a much clearer view of where the science stopped, and political or economic interests began.

Eric Meunier of Inf’OGM spoke for the NGO stakeholders and argued that complete transparency in the decision-making process and public access was the minimum required of EFSA to restore public trust. He called for “complete, unrestricted and proactive online publication of applicants’ files when these reach EFSA” and the data contained would have to be accessible to all without justification or identification.

Euros Jones of the European Crop Protection Association put the cards on the table for the food industry. Transparency was fine by industry if it genuinely served the public interest, but it must do so without undermining legitimate private interests. So the trick would be to protect confidential business information and intellectual property rights and thus preserve the competitiveness of EU industry, and at the same time ensure a positive investment climate within the EU and maintain the integrity and effectiveness of the decision-making process. There should be consistency – that is, the same level of scientific scrutiny was applied to all applications – and different EFSA panels should co-operate to ensure consistency when the same product was evaluated by different panels. More pointedly, public access to reports should be on a request basis only; applicants should be informed about data requests; and at the same time applicants should have a chance to review data about to be released; he also wanted EFSA to take “reasonable precautions” to ensure that released data were not then misused by other people.

These stakeholder standpoints are likely to leave brows furrowed and heads scratched for some time to come. The challenge now for EFSA and its advisers is to find a way to reconcile two very different sets of needs and at the same time to extend the principles of openness in a way that is fair to industry and clear to the public.

Tony Hardy, Chair of the Scientific Committee, could hardly resolve such a clash of issues in the course of a late afternoon summary. Instead he reminded the conference that transparency mattered to EFSA because it generated trust in the independence and soundness and rigour of its advice; it underpinned scientific clarity and excellence, and it brought consumers and the wider scientific community together. The devil as usual, lies in the details: the clash of demands from industry and NGOs was discernible in reports from all the breakout groups, and EFSA faced a challenge in communicating unambiguously with a wide range of different stakeholders, while at the same time clearly identifying the risk assessment process. But there would be further meetings on process transparency and information access and by 2014, EFSA’s new policy on transparency would be “informed” by such interaction with the stakeholders. EFSA’s journey to a new level of transparency has begun. The road ahead, however, is not yet clear.