EDITORIAL: Why should we report adverse incidents?

Three papers in this issue of the Journal illustrate the challenges of measuring and reporting therapeutic mishaps. Using a broad definition of ‘adverse incidents’ that embraces injuries (including death) from any cause, drug errors, unexpected events (e.g. convulsions) or findings (e.g. Hb8gm/dl), and worrisome indicators (e.g. LOS5 days), Stanhope et al. (1998) found 196 incidents identified from 500 consecutive deliveries in two obstetric units. While many (most?) of these ‘incidents’ were not actual injuries, the fact that 39% of patients had something other than a perfectly smooth course is very disturbing. It is an unpleasant reminder that, even in 1996, the most normal of biological events, childbirth, is still fraught with hazard. Also disturbing, but less surprising, is their finding that of incidents that were discoverable on later record review only 23% were reported (as required) when they occurred. In their search for explanations for non‐reporting by survey of the staffs of these same two units, Vincent et al. (1998) found that one‐third of physicians did not know how to find the list of reportable incidents and 15% did not know how to file a report. No single reason for non‐reporting among 12 reasons suggested by the investigators was accepted by a majority of respondents, suggesting that the problem is multi‐factorial or, more probably, denial of reality by the respondents. The authors also note the role of perceived lack of necessity and concern about disciplinary action as factors in non‐reporting.