ABSTRACT This study uses the publicly available International Medical Devices Database in three-phase research to present empirically grounded multi-dimensional typologies of events involving faulty medical devices. In the first phase, considering three major categorical data, i.e. event actions, event types, and countries, and, by employing descriptive analysis, we identify that perhaps ‘Field safety notice’ is the most dominant event type under which 66 event actions exist. In the second phase, employing the two-step clustering analysis technique, we identify that four distinct clusters exist. Finally, in the last phase, employing the focus group-driven content analysis approach using event reason description in conjunction with the SHELL (software-hardware-equipment-liveware) model, we classify the top 10 event actions. We also identify corresponding risk levels considering the top 10 sources of device problems. Further, the focus group employing the interest-influence framework in conjunction with pertinent self-addressed questions ascertained that the greatest number of top 10 event actions lies at the quadrant wherein the major stakeholder has a high degree of interest and influence in mitigating those medical events. Our study bridges the extant research gap about the lack of comprehensive, data-driven, and evidence-based themes around medical device events, the associated risk characteristics, and stakeholder mapping.