Since the new definition of sepsis was published,1 it is worth reflecting on the salience of conflicts of interest in this type of document.

Consensus documents have a large impact on practice, resulting in changes to diagnostic criteria and introducing new interventions, as well as determining the allocation of resources. In recent years there has been a growing interest regarding the impact that conflicts of interest have on these publications.2,3 Traditionally, economic conflicts of interest have held the spotlight. Nevertheless, in the scientific arena there are other elements capable of creating a secondary interest in the signing authors. Thus, the pressure to publish, or the research group's own pathway, can create intellectual conflicts of interest.4

In 2015, we reviewed the main clinical practice guidelines for our specialty. The guidelines published between December 2009 and December 2014 were reviewed. A total of 96 guidelines were included. Conflicts of interest were not reported in 32.6% of the guidelines. Moreover, in 54% of the guidelines that did report them, all the authors presented conflicts. In the main document of the new definition, of the 19 authors, 10 present some economic conflict. Despite this, the data provided are limited and confusing. Only the company or institution with which the conflict is established was reported, without specifying the type of relationship, the amount, or the area of knowledge affected.

In this context, consensus documents are highly vulnerable to the effect of conflicts for several reasons. Firstly, the recommendations issued are not always based on solid tests (evidence), resorting to the opinion of experts. They usually use a mixed methodology that includes consensus strategies (Delphi), systematic reviews, and GRADE evaluation, among others. This all creates a confusing result, which makes it difficult for other scientists to establish the traceability of the process, perform an external validation or make an assessment of the possible sources of bias.

Secondly, the participation of opinion leaders, though useful (clinical perspective, methodological abilities and community acceptance), can add new risks to the process.5 Because of their career path, these authors may be more exposed to the pharmaceutical industry. The conflicts will be reflected in the document, but they are not characterised or quantified, further clouding the issue. Lastly, it is clear that the career path of each author remains linked to specific lines of research. In this group of authors, this fact could be associated with a higher risk of intellectual conflicts and pathway biases.

No one questions that the aim of these documents is to improve patient prognosis, decrease variability in practice, and produce more efficient research. However, attaining these goals requires a critical reading of these documents and the setting in which they were created. Before taking a new path, we have to ask ourselves where it comes from to know where it is taking us.