Unhappy meal. The European Food Safety Authority's independence problem

Download full report - Read press release

In recent years, the European Food Safety Authority (EFSA) has come under sustained criticism from the European Parliament, NGOs and the media over conflicts of interest of those sitting on its scientific panels. These experts play a crucial role in decisions key to the health and safety of Europe's food supply chain. Yet some were shown to have commercial ties with the industries whose profits depend on these products, undermining the credibility of the organisation's scientific output on issues such as food additives1 and GMOs.2 After an initial phase of denying there was a problem, EFSA has developed – in its own words – “a new, comprehensive and sophisticated” policy on independence.3 The renewal of eight of its ten scientific panels in the course of Spring 2012 was the opportunity for the agency to start implementing its new policy to vet the participants for conflicts of interest – and regain credibility.

Having published a significant amount of research into industry's influence at the agency over the past few years,4 Corporate Europe Observatory (CEO) felt it important to assess this new policy, combining the experience we have acquired through previous investigations with a more systematic method of data analysis. Replicating EFSA's work and using only the information declared by the scientists themselves to EFSA – a conservative approach – we screened all interests of the scientific panel members as well as members of the Scientific Committee. As a result of its new independence policy, EFSA banned 85 experts from joining its panels, so there was room for optimism that some of the major problems had been fixed.

Unfortunately, that optimism is not borne out by our analysis. The results of our screening are dismaying. While we were still expecting to find conflicts of interest, we were surprised to find so many: 122 experts out of 209 (58.37%) have at least one conflict of interest with the commercial sector. Experts with conflicts of interest dominate all panels but one. All but two panel chairs and 14 vice- chairs among 21 have conflicts of interest. The “worst record” is held by EFSA’s panel on Dietetic Products, Nutrition and Allergies (NDA), with 17 of its 20 members totalling 113 conflicts of interest between them. In all panels, ten experts have more than 10 conflicts of interests each. One member of the panel on Additives and Products or Substances used in Animal Feed (FEEDAP) has 24 on his own. Among 849 interests screened, CEO counted an overall sum of 460 conflicts of interest. On top of that, there is no visible difference in the proportion of conflicts of interests in the eight renewed panels compared to the two that are yet to be renewed, which poses further questions about the new policy's efficacy.

Why such terrible results? Complacency at EFSA? Excessive rigour on our part?

To repeat, our methodology for this screening was very conservative: we did not check for undeclared interests by experts, we did not add omitted interests that we found in the course of the research. Fortunately, lack of political will at EFSA seems to be less of an issue than in the past. The agency now seems seriously concerned, dedicating significant resources to the problem. Indeed, as soon as EFSA was informed of CEO’s intention to scrutinize its new policy, we were invited to the agency's headquarters in Parma, Italy, in June 2013. Thanks to this unprecedented gesture, we had the opportunity to converse with several high up officials from the Authority in the course of an intense full-day meeting, during which the agency's handling of conflicts of interests was extensively discussed. Showing dedication and good will was clearly part of the picture, but this exercise was also instructive for us in terms of what remains to be done at the agency: EFSA's independence policy has many flaws.

A first one has to do with the nature of the rules themselves, as they are insufficiently rigorous. Another important weakness is the reliance on experts to declare their own interests. But there is also a cultural issue, different in nature from problem of rule implementation, which we shall also examine.

Firstly, EFSA's independence rules themselves are insufficient in the sense that they are too narrow. The main criterion the agency uses to assess an expert's given interest is to consider whether it falls inside the thematic mandate of the panel the expert is applying to. In other words, any scientist with ties to the commercial sector can still be accepted as long as the interest is not related to the panel's topic. This is in our view the biggest loophole in the rules, and probably the one main factor explaining our results. We considered that the relevant criterion was not the panel's mandate but the remit of EFSA itself. Furthermore, this thematic specialisation forces the declarations of interest to be assessed individually – at a considerable expense of time and energy – by all the agency's heads of units.

As a consequence, some problematic interests are not considered as such. To cite only one example, a few ongoing collaborations with the industry think-tank International Life Sciences Institute (ILSI) are still tolerated whereas this particular organisation has been at the core of past controversies with EFSA. Moreover, there is no cooling-off period: recent activities with industry-affiliated bodies are not considered a problem by the agency insofar as they are terminated, meaning scientists can go straight from these to sit on an agency panel. As a result, more than 30 experts with a history – even in the very recent past – at ILSI are still members of EFSA's scientific panels.5

Secondly, while EFSA as an institution should safeguard its independence by taking charge of the checks for conflicts of interest proactively, it relies instead on the experts’ own self-assessments. Information recorded by the scientists themselves in their declarations of interests or CVs is the only source used by EFSA. Their accuracy is taken for granted, and not even a basic check on publicly available information, on the internet for example, is performed – almost an incentive for abuse. The whole system will remain flawed as long as it only relies on experts' self-assessment.

Many cases of conflicts of interest remain undetected by EFSA's current system because the rules are clearly insufficient. What is worse, there are problems of implementation of EFSA's already lenient existing rules: some conflicts of interests should have been detected according to EFSA's own rules but haven't been. Had EFSA thoroughly applied its new policy by the rulebook, we think, seven chairs and three vice-chairs of the scientific panels would not have been appointed.

Finally, and crucially, an insufficient understanding of what conflicts of interest entail in practice undermines the screening process performed by EFSA’s staff. The agency's idea of a conflict of interest revolves around a dramatic picture of corruption and infiltration by industry “moles” with evil intent. Even though this cannot be ruled out, the reality is usually more subtle. Industry influence tends to be exerted through long-term and structural processes of relationship- building within the scientific community itself, through culture, collective dynamics, accepted paradigms and group thinking – where it becomes natural to “think industry” – rather than through some kind of manipulation at the individual scientist level only. As we remind our readers in this report, science itself is nowadays an open battlefield for corporate interests. That should be cause for extra vigilance and scrupulousness when creating and implementing rules governing conflicts of interest. But EFSA seems unconcerned by this reality.

Based on our research, the numerous discussions we had and our previous knowledge of the field, we have come to a series of recommendations that can also be seen as a more general contribution to the EU's initiative to deal with conflicts of interest in the agencies in a more rigorous and informed manner. In the short term, EFSA could upgrade its rules by banning commercial interests entirely and improving its screening system. In the medium term, EFSA could outsource the screening of the declarations of interests from heads of units to specialised personnel, for instance magistrates from the European Court of Auditors. In the longer term, expertise could be in-sourced in order to give them all the means to do their work properly and be independent from the sectors they are regulating. Another long-term recommendation would consist of having the studies on regulated products conducted by independent/public laboratories on the basis of very strict rules including blinds (these could still be paid for by industry). Our recommendations are detailed at the end of this report.

It is important to keep the bigger picture in mind. While our recommendations for better rule implementation might improve the quality and credibility of EFSA expert panels, there are also some larger structural issues that are beyond the scope of this report to properly address. It is crucial to note that EFSA experts are unpaid (only expenses), for one. For another, there is a structural conflict of interest built into the system, as the experts only assess studies performed by the producers of the products at stake (they do not perform research themselves). Combine this with excessive workloads, and we can see that to do this job properly is a daunting task. Moreover, parts of these studies are usually kept secret for commercial confidentiality reasons, preventing their integration in the normal work of the scientific community. As a result, it seems that serving on an EFSA panel is neither beneficial nor attractive to build a scientific career, making it harder to find young and independent experts working disinterestedly for the public good. It is unacceptable that such a crucial task for public health is rendered so unrewarding.

The full report can be downloaded here.

Disclaimer

The aim of this report is to review EFSA's new independence policy and particularly its handling of conflicts of interests among its scientific panel members. It is therefore important to point out that what is being assessed here is EFSA's decision-making process in whether to accept or reject given experts on its panels for a public interest mission. Having a conflict of interest with the commercial sector does not mean that an expert is criticised for his/her ethics or intellectual honesty, but that he/she cannot be considered independent from industry's influence. Therefore, we think, the expert is not in a position to participate in the work of an agency whose workload consists primarily in assessing the safety of industrial products to be commercialised on the EU market.

All unreferenced interests mentioned in the report come from the experts' declarations of interests, downloaded from EFSA's website on April 29, 2013.
 

Erratum

The following sentence (p.13, box "crash course on regulatory capture")

Although it claims to be seeking “a balanced approach to solving problems of common concern for the well being of the general public”, ILSI, co-founded by Philip Morris, is financed by food, chemical, pesticides, GMO and pharmaceutical multinational corporations such as Coca-Cola, BASF, Unilever, Syngenta, Pfizer and so on. (endnote)

contained a mistake (ILSI was not co-founded by Philip Morris) and has been replaced on Wednesday 30 October with the following:

Although it claims to be seeking “a balanced approach to solving problems of common concern for the well being of the general public”, ILSI, co-founded by Coca-Cola, Heinz, Kraft, General Foods and Procter & Gamble, is financed by food, chemical, pesticides, biotechnology and pharmaceutical multinational corporations and, between 1983 and 1998, “provided assistance to the tobacco industry in its attempts to subvert many attempts at legislative control over the industry's activities”. (endnote)

• 1. Exposed: conflicts of interest among EFSA’s experts on food additives, Corporate Europe Observatory, 15 June 2011, http://corporateeurope.org/efsa/2011/06/exposed-conflicts-interest-among...
• 2. Approving the GM potato: conflicts of interest, flawed science and fierce lobbying, Corporate Europe Observatory, 7 November 2011, http://corporateeurope.org/efsa/2011/11/approving-gm-potato-conflicts-in...
• 3. Decision of the European Ombudsman closing his inquiry into complaint 775/2010/ANA against the European Food Safety Authority (EFSA). 23 May 2013. http://www.ombudsman.europa.eu/fr/cases/decision.faces/en/50246/html.boo...
• 4. Conflicts on the menu, a decade of industry influence at the European Food Safety Authority (EFSA), Corporate Europe Observatory and Earth Open Source, February 2012, http://corporateeurope.org/efsa/2012/02/conflicts-menu
• 5. The agency's recently appointed Director of Science Strategy & Coordination herself worked for ILSI until 2004.