National regulatory agencies’ decisions to approve new drugs are based on limited safety evidence collected during clinical development. Often, only when a drug enters general use do rarer or longer-term adverse events become known or better understood, prompting regulators to issue safety advisories.1 We examined how often medicines regulators in 4 countries with similar medical traditions, population health, and demographics—Australia, Canada, the United Kingdom, and the United States—were concordant in their decisions to issue safety advisories on approved prescription medicines.
We undertook a retrospective analysis of safety advisories issued by the US Food and Drug Administration (FDA), Health Canada (HC), the UK Medicines and Healthcare products Regulatory Agency (MHRA), and the Australian Therapeutic Goods Administration (TGA) from January 1, 2007, to December 31, 2016. Safety advisories were defined as notifications to prescribers and/or the public about a potential or confirmed drug risk, excluding issues of production quality, shortages, or overdoses. We obtained advisories from regulators’ websites (current and archived). This project did not fall within the scope of the National Statement on Ethical Conduct in Human Research because all included data are from publicly available documents. Therefore, according to University of Sydney Research Code of Conduct, ethics approval was not required.
Each identified drug-risk issue was reviewed for concordance or discordance among regulators. Overall concordance rates were calculated for drug-risk issues for which at least 2 countries had issued an advisory.
Statistical analysis was performed from February 1, 2017, to October 1, 2018. Regulators’ discordancy rates, defined as the number of drug-risk issues for which each regulator did not issue an advisory for an approved drug, were compared using χ2 tests for independence. All P values were from 2-sided tests, and the results were deemed statistically significant at P < .05, and Bonferroni corrections were used for multiple comparisons.
We identified 1441 advisories in the 4 countries, covering 680 drug-risk issues. The MHRA issued advisories for 344 of 657 drug-risk issues (52.4%) for medicines approved in the United Kingdom, HC issued advisories for 317 of 635 drug-risk issues (49.9%), the FDA issued advisories for 265 of 647 drug-risk issues (41.0%), and the TGA issued advisories for 183 of 619 drug-risk issues (29.6%) (Table).
The overall frequency with which the 4 regulators issued safety advisories differed significantly (χ23 = 82.3; P < .001). The MHRA was more likely than other regulators to issue an advisory (136 of 657 [20.7%] vs FDA, 89 of 647 [13.8%]; HC, 122 of 635 [19.2%]; and TGA, 69 of 619 [11.1%]; P = .001), whereas the TGA was least likely to issue an advisory (P < .001; Bonferroni-adjusted significance, P = .006). The FDA and HC did not differ significantly from expected distributions.
The Table shows a low rate of agreement in decisions to inform professionals and the public of emergent safety concerns. The TGA had the highest discordance rate, providing no warning for 436 of 619 drug-risk issues (70.4%) for drugs approved in Australia. The MHRA issued more warnings, but still failed to provide warnings on 313 of 657 relevant drug-risk issues (47.6%). For 70 of the 680 identified drug-risk issues (10.3%), regulators issued advisories in every country where the drug was marketed. For 40 of 573 drug-risk issues, all 4 countries had approved the drug and also issued advisories (Figure).
Overall, we found a low level of concordance (10.3%) between regulators in the decision to warn clinicians and the public about risks of approved prescription medicines. These results suggest widespread differences in communication patterns among included regulators.
Our findings expand on and confirm the results of a comparative analysis of direct health professional communications issued by 4 European countries.2 That analysis reported inconsistency in decisions to issue advisories, despite national regulators’ reliance on the same information from the European Medicines Agency.
This study does have some limitations. We did not consider advisories outside this 10-year window, and some warnings may have been issued before or after this period. We also did not consider other risk management tools that may have been used to communicate harm, such as the US Risk Evaluation and Mitigation Strategies. However, a sensitivity analysis including safety letters sent out within the US Risk Evaluation and Mitigation Strategies indicated little effect on results (57% vs 59% FDA discordance rate).
Our results likely reflect differences in national approaches to pharmacovigilance. National medicines policy determines the activities of the medicines regulator and resource availability, including the capacity to undertake postmarket monitoring and its administrative burden.3 Further study into the regulators’ decision making and action thresholds4 (based on seriousness of harm or strength of evidence) and the follow-on effects of these decisions and actions is required to fully elucidate the public health implications of these policies.
Accepted for Publication: January 26, 2019.
Corresponding Author: Barbara Mintzes, PhD, Charles Perkins Centre and School of Pharmacy, Faculty of Medicine and Health, The University of Sydney, D17, John Hopkins Drive, Camperdown, NSW 2006, Australia (barbara.mintzes@sydney.edu.au).
Published Online: April 29, 2019. doi:10.1001/jamainternmed.2019.0294
Author Contributions: Ms Perry had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Perry, Bhasale, Puil, Mintzes.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Perry, Lexchin.
Critical revision of the manuscript for important intellectual content: Bhasale, Fabbri, Puil, Joarder, Mintzes.
Statistical analysis: Perry.
Obtained funding: Puil, Mintzes.
Administrative, technical, or material support: Bhasale, Joarder, Mintzes.
Supervision: Lexchin, Mintzes.
Funding/Support: This research was funded by grants from the National Health and Medical Research Council of Australia (APP1122332) and the Canadian Institutes of Health Research (CIHR PJT-153275).
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Conflict of Interest Disclosures: Ms Perry reported being employed by George Clinical Pty Ltd outside the submitted work. Ms Bhasale reported part-time employment with the Australian Commission on Safety and Quality in Health Care. Dr Lexchin reported serving as a member of the Foundation Board of Health Action International and the Board of Canadian Doctors for Medicare. Dr Mintzes reported receiving personal fees from the legal team representing plaintiffs in a Canadian application for a class action suit on cardiovascular risks of testosterone outside the submitted work. No other disclosures were reported.