1 Klungel OH Pottegård A Strengthening international surveillance of vaccine safety. The comparison of the risk–benefit balance of COVID-19 vaccines in real-world populations has presented new challenges over the past few months. For instance, the detection of rare adverse events is unresolved by clinical trials but mandatory to better inform clinical decision-making in countries where several vaccines are available.However, heterogeneity in the distribution of different COVID-19 vaccines among countries and populations makes such comparisons difficult. 2 Wan EYF Chui CSL Lai FTT et al. Bell's palsy following vaccination with mRNA (BNT162b2) and inactivated (CoronaVac) SARS-CoV-2 vaccines: a case series and nested case-control study. 2 Wan EYF Chui CSL Lai FTT et al. Bell's palsy following vaccination with mRNA (BNT162b2) and inactivated (CoronaVac) SARS-CoV-2 vaccines: a case series and nested case-control study. We read with interest the Article by Eric Yuk Fai Wan and colleagues,which explored the association between Bell's palsy, the mRNA-based BNT162b2 vaccine, and the inactivated CoronaVac vaccine in Hong Kong.Hong Kong is one of the rare places where both types of vaccines are available, allowing direct comparison of their efficacy and safety from a unique database. The results of this study suggest a higher risk of developing Bell's palsy in individuals who received CoronaVac than in those who were unvaccinated, and also in those who received the BNT162b2 vaccine. Surprisingly, this safety signal has not been detected by global pharmacovigilance systems. 3 Harpaz R DuMouchel W Shah NH Madigan D Ryan P Friedman C Novel data-mining methodologies for adverse drug event discovery and analysis. 4 Raschi E Moretti U Salvo F et al. Evolving roles of spontaneous reporting systems to assess and monitor drug safety. Since 1968, the WHO pharmacovigilance database has collected and aggregated suspected adverse drug reactions that are spontaneously reported by patients and health professionals from more than 150 countries. This system allows early detection of rare adverse drug reactions by identifying unexpectedly increased proportions of adverse drug reactions reported with a particular drug compared with other drugs in the database (ie, disproportionality signals).Comparison of signals between drugs is challenging in global pharmacovigilance databases because of the heterogeneity in pharmacovigilance systems, unmeasured confounding, change in the rate of adverse events reported to pharmacovigilance systems with drug time on the market, and media coverage of drugs or adverse events, or both. However, COVID-19 vaccines are a rare case for which comparison of signals could be relevant because of similar therapeutic indication, use, and time on the market. 5 Renoud L Khouri C Revol B et al. Association of facial paralysis with mRNA COVID-19 vaccines: a disproportionality analysis using the World Health Organization pharmacovigilance database. As of Aug 31, 2021, the WHO pharmacovigilance database contained 770 343 reports of adverse events with the BNT162b2 vaccine, of which 7892 were reports of facial paralysis (the method for case identification has been described elsewhere).However, only 30 091 reports of adverse events had been made for CoronaVacc, of which 38 were reports of facial paralyses. Therefore, the disproportionality signals of facial paralysis are lower for Coronavac than for BNT162b2, which differs from the findings of Wan and colleagues’ study. This discrepancy could be due to several reasons. First, the broad media coverage of this potential adverse event might have stimulated its reporting with mRNA vaccines. Second, the low number of adverse events reported with CoronaVac vaccines does not allow for an accurate estimate of the proportion of rare adverse drug reactions. Several large countries that used CoronaVac did not report any adverse events as part of the WHO Program for International Drug Monitoring ( appendix pp 1–2 ). The results of Wan and colleagues’ study highlight the risk of false conclusions being drawn from comparison of disproportionality signals in an international pharmacovigilance database when reporting is incomplete. Therefore, we urge all countries to report all suspected cases of severe and unexpected adverse drug reactions to international pharmacovigilance systems in a transparent and timely manner to improve the collective knowledge on the safety of these vaccines. We declare no competing interests. Supplementary Material Supplementary appendix Article Info Publication History Identification DOI: https://doi.org/10.1016/S1473-3099(21)00646-0 Copyright © 2021 Elsevier Ltd. All rights reserved. ScienceDirect Access this article on ScienceDirect