• Patient safety
• Incident reporting
• Risk management

Since its inception, the patient safety movement has been obsessed with reporting systems; roughly 20% of the To Err Is Human report1 dealt with some aspect of reporting; similarly, about 10% of the articles in BMJ Quality & Safety (and its predecessors) mention ‘reporting’ in their title, abstract or keywords. This interest sprang from an unholy trinity—an infortuitous combination of the epidemiological bent of many health professionals interested in safety, the epistemological blinders that a medical education produces2 and a kind of ‘aviation envy’—a fascination with aviation safety as an exemplar which healthcare would do well to emulate.3 But, in an all-too-common pattern in patient safety, fundamental aspects of successful safety practices in other domains were misunderstood, misapplied, mistranslated or missed altogether as health professionals encountered work whose basic assumptions are far removed from the realism and positivism of biomedicine.2 ,4 Thus, the attempts to plant practices such as incident reporting as new colonies in the world of clinical care have foundered and fallen short of the cheers and high hopes that accompanied their launching. Tangible improvements from reporting are hard to find,5 ,6 while reporting systems languish and curdle into objects of satire.7

Social psychologist Macrae5 recently summarised the problems with incident reporting in the pages of this journal, noting that healthcare has focused far too much on the technical infrastructure of reporting …