Signals of Adverse Reactions to Herbal Medicines: Evidence and Document Analysis Based on a Scoping Review

Background: To date, signals of adverse reactions to herbal medicines have not been systematically reviewed, limiting pharmacovigilance of herbal medicines because of a lack of data. Objectives: We sought to analyse the available evidence on signals involving herbal medicines and to determine to what extent they had been documented at the European Union (EU) level and in the USA. Methods: We used the results of a published scoping review of interventional and non-interventional studies that reported signals of adverse reactions to drugs. We assigned Anatomical Therapeutic Chemical classification to all drugs, and identified herbal medicines when they fell under the Anatomical Therapeutic Chemical V90. We ascertained the presence of the adverse reaction, or related adverse reactions, for each signal in reference documents for healthcare professionals: the US Botanical Safety Handbook and the EU monographs and US Dietary Supplement Fact Sheets; and in those for consumers: the US Dietary Supplement Label Database. We summarised the data descriptively, treating US documents as one and comparing harms across pairs of US and EU documents by signal. Documents were deemed concordant if they both included the same or related adverse reactions, or if neither did. We also compared adverse reactions across US documents for healthcare professionals with those for consumers. Results: Of the 10,861 signals covered by the scoping review, 53 (0.49%) concerned herbal medicines, all based on case reports. Reference documents from both the US and EU were available for 37 signals. Most of the documents were concordant (73%), and ten (27%) were discordant: six adverse reactions were mentioned only in US documents, three only in EU monographs, and one was warned against in US documents but not in EU documents. Twenty-one signals could be followed up in the Botanical Safety Handbook and Dietary Supplement Fact Sheets. Most (68%) US documents for healthcare professionals were concordant. When the Botanical Safety Handbook and Dietary Supplement Fact Sheets did not include an adverse reaction, neither did the Dietary Supplement Label Database. However, when they did, only 20% of the labels for consumers did too. The proportion of labels mentioning adverse reactions otherwise available in documents intended for healthcare professionals ranged widely, reflecting differences across multiple labels for the same products. Conclusions: Very few signals of adverse reactions from the wider scoping review concerned herbal medicines, and were all based on case reports. Information was mostly concordant across documents in the EU and USA. As manufacturers are solely responsible for the contents of the Dietary Supplement Label Database, regulatory oversight may be required to ensure that consistent and comprehensive information on the harms of herbal medicines is made available to consumers in the USA.