Severe abnormal behavior incidence after administration of neuraminidase inhibitors using the national database of medical claims

An earlier study using the number of abnormal behaviors reported to the study group as the numerator and the number of influenza patient prescribed each neuraminidase inhibitor (NI) estimated by respective pharmaceutical companies found no significant difference among incidence rates of the most severe abnormal behaviors by type of NI throughout Japan. However, the dataset for the denominator used in that earlier study was the estimated number of prescriptions. In the present study, to compare the incidence rates of abnormal behavior more precisely among influenza patients administered several sorts of NI or administered no NI, we used data obtained from the National Database of Electronic Medical Claims (NDBEMC) as the denominator to reach a definitive conclusion. Results show that patients not administered any NI (hereinafter un-administered) or those administered peramivir sometimes showed higher risk of abnormal behavior than those administered oseltamivir, zanamivir, or laninamivir. However, the un-administered or peramivir patients were fewer than those taking other NI. Therefore, accumulation of data through continued research is expected to be necessary to reach a definitive conclusion about the relation between abnormal behavior and NI in influenza patients. Since severe abnormal behaviors with all types of NI or of un-administered patients have been reported, there are some risks in the administration of NI or even in un-administered cases. Therefore, we infer that the policy mandating package inserts in all types of NI.