[Impact of conflicts of interest on guideline recommendations : Empirical study within the second update of the German interdisciplinary S3 guidelines on fibromyalgia syndrome].

BACKGROUND: The impact of conflicts of interest (COI) in general and of academic COI in particular on guideline recommendations in pain medicine has not yet been studied. Whether the inclusion of patients and of representatives of all relevant healthcare professions into a guidelines group is protective against a systematic bias of decisions of a guidelines group is currently unknown.

METHODS: All members of the guidelines group declared their COI before the consensus conferences by a standard form according to the rules and standards of the Association of the German Medical and Scientific Societies. The acceptance or rejection and the strength of consensus of recommendations of the second update of the interdisciplinary guidelines on fibromyalgia syndrome was analyzed twice by first including and then excluding the votes of the guideline group members with COI related to a recommendation from the results of anonymous voting via an internet platform.

RESULTS: A total of 42 persons from different healthcare professions and patients participated in the online voting on recommendations. Of the participants 29% had no COI according to the predefined criteria, 53% met the criteria of academic and 33% the criteria of financial COI. In the case of exclusion of participants with a COI related to a specific recommendation, 2 out of 23 recommendations (homeopathy, tramadol) were not accepted. In all votes, there were more participants without COI than with COI.

CONCLUSION: Academic COI were more frequent than financial COI in the second update of the German interdisciplinary guidelines group on fibromyalgia syndrome. The impact of COI on guideline recommendations was low. The inclusion of patients and of all relevant healthcare professionals into a guidelines group is a protective factor against the influence of COI on guideline recommendations.