BACKGROUND: Severe hypersensitivity reactions (SHRs) in intravenous (IV) iron treatment are a known but rare, and therefore poorly characterized, safety concern. A regulatory-mandated postauthorization safety study (PASS) with multiple sponsors will assess the risk of SHRs in IV iron users in Europe. A multidatabase study approach is required to evaluate this rare safety outcome. Results will be available in 2020.
OBJECTIVES: To describe cohort attrition of IV iron users and challenges encountered in conducting this PASS. Methods: Cohorts of new and prevalent adult IV iron users with at least 12 months of available information were identified from seven data sources in five countries: Denmark (national and regional registers), France (SNDS), Germany (GePaRD, QiN, DIMDI-DaTraV), the Netherlands (PHARMO), and Sweden (national registers). The algorithms used to identify SHRs rely both on diagnostic codes and markers of SHRs (e.g., symptoms and treatments). A cohort of IV penicillin users was identified, where feasible, to assess the performance of the algorithm. Code lists for the algorithm and covariates were harmonized across data sources. Validation of cases through source record review is planned in Denmark and the Netherlands. Indirect external validation of the SHR algorithm will be conducted in the Oldenburg University Hospital in Germany. Data source-specific analyses and pooled analysis of aggregate data will follow a common protocol.
RESULTS: Heterogeneity in the data available across data sources was addressed through collaborative work across all research centers. Drug exposure is captured through prescription, dispensing, or administration records from hospital or outpatient settings, thus requiring tailored definitions of “time at risk” windows. As of January 2019, based on data from three data sources after cohort attrition, approximately 180,000 IV iron users with a first use, 102,000 IV iron users with a second use, and 80,000 IV iron users with a third or subsequent use were eligible for the study. The IV iron cohort sizes vary greatly across data sources and countries. Data for all IV iron types are not captured in any single data source.
CONCLUSIONS: Initial data suggest sizeable numbers of IV iron users. However, the informativeness of results for individual drugs will depend on the relative use of each IV iron subtype and the frequency of SHRs. Understanding the commonalities and differences of the data available in the collaborating centers and standardizing data definitions are critical to conducting a multidatabase study.
