“…Underreporting (Martin et al, 1998), the phenomenon of insufficient awareness of a possible association between drug treatment and occurrence of an ADR or unwillingness to report an ADR, might be the main reason. Secondly, considering that the risk of bleeding is presumably only slightly increased under treatment with SSRI (Andrade et al, 2010;Aarts et al, 2014), methods of pharmacovigilance (i.e. signal detection analyses using pharmacovigilance databases) might be insufficiently sensitive for detecting rarely occurring ADR such as drug-related haemorrhages.…”