The Field Guide › Paper
Sutcliffe, Lewton and Rosenthal interviewed twenty-six residents at a large American teaching hospital about the mishaps they had recently been involved in, and surfaced around seventy of them. Communication failure was implicated in most, which is the expected finding. The unexpected and far more useful finding is what those failures actually consisted of. Poor communication, the authors conclude, is not simply a matter of information failing to be transmitted or exchanged; it is produced by the social, relational and organisational structures of medicine itself. Residents described hesitating to ask a question or raise a doubt because doing so risked appearing incompetent in front of a senior; they described the vertical hierarchy of the profession making it awkward or unsafe to challenge someone above them, and the horizontal divisions between specialties and professions leaving no one clearly responsible for saying the difficult thing. Communication broke down not because the words were unavailable but because the conditions for speaking were absent. For a corpus about psychological safety this is a striking document, because it arrives at Edmondson's mechanism from the clinical side and largely without her vocabulary: the barrier to speaking is the image cost of speaking, and the hierarchy determines who pays it. It also supplies the everyday texture that the disaster literature cannot, since these were not catastrophes but ordinary mishaps of the kind that happen in every hospital every week, which is precisely what makes them insidious. The paper's implication is that interventions aimed at communication as a skill, at handover checklists and read-backs and SBAR, will only travel so far, because the failure is structural and interpersonal rather than technical. Its limits are those of a single-site qualitative study with a small sample, relying on residents' retrospective accounts of events in which they were participants, so it maps the mechanisms richly without establishing their prevalence. (Text drawn from the 2004 Academic Medicine paper, 79(2), pp. 186-194.)