BACKGROUND: Anaphylaxis (ANA) related to intravenous (IV) iron treatment is a poorly characterized safety concern in Europe. A postauthorization safety study requested by the European Medicines Agency assessed the risk of ANA in IV iron users in Europe.
OBJECTIVES: To assess the risk of ANA overall, by IV iron groups (iron non-dextrans and iron dextran), and by individual types.
METHODS: A cohort study of IV iron users was conducted in 5 European countries covered by 6 health databases, including data spanning from 1999 through 2017: Denmark (Health Services Database of the Central Region), France (SNDS), Germany (GePaRD and Quality in Nephrology programme), the Netherlands (PHARMO), and Sweden (Swedish national registers). IV iron treatment was captured mainly from ambulatory drug dispensing data. Data source-specific analyses followed a common protocol with a new-user design. Risk of ANA occurring the day of or day after first, second, and third or subsequent iron treatment was estimated using beta-binomial derived combined incidence proportions (IPs) and 95% confidence intervals. To provide context, the risk of ANA was also assessed among users of IV penicillins, if available. Validation of potential ANA events was conducted in the Health Services Database of the Central Denmark Region and PHARMO by review of medical records. For GePaRD, indirect validation of the case-identification algorithm was performed in the Oldenburg Hospital.
RESULTS: Of 304,210 first-recorded IV iron treatments, 2.1% were iron dextran. There were 231,294 first-recorded IV penicillin treatments. The overall IP of ANA at first IV iron treatment ranged from 0.41 (0.13-1.28) to 1.16 (0.20-6.68) per 10,000 treatments; at first iron non-dextrans treatment, the IP ranged from 0.43 (0.13-1.37) to 1.2 (0.21-6.72) per 10,000 treatments (IPs reported as ranges to comply with data protection rules). There were no ANA events among first iron dextran treatments. Risk by groups and types of IV iron were estimated but were limited by very low numbers. No adjusted analyses could be performed either. The IP of ANA at first IV penicillin treatment was 1.29 (0.54-3.11) per 10,000 treatments.
CONCLUSIONS: We found an IP of ANA lower than the estimated 2 to 6.8 per 10,000 first treatments (IV iron non-dextrans and iron dextrans, respectively) reported by two recent United States studies. The IP of ANA in users of penicillins was consistent with the incidences reported in the literature. Most data sources do not capture in-hospital use of IV iron, likely resulting in misclassification of repeated users as first users. This may have resulted in an underestimation of the IPs of ANA.
