Maintaining the link between methodology and method in ethnographic health research

We read with interest the debate around the use, and representation, of ethnography in the journal following the publication of Lamba et al's article on the identification of patient safety problems during ward rounds.1 In correspondence, Tanisha Jowsey saw the study, which involved observational methods to catalogue safety issues and then applied descriptive statistics as a method of analysis, as misappropriating the ethnographic tradition.2 We share with Jowsey her concerns about the way ethnography can sometimes be used as a label to describe observational studies where, in broad terms, a researcher counts or categorizes clinical activities. It is our view that ethnography is much more than observing. Rather, it is about the type of questions we ask of the social world. We wonder whether such methodological slippage would be tolerated, for example, with trial methodologies where issues of randomisation, power calculations and analytical rigor are heavily scrutinized. The journal's editors, Dixon-Woods and Shojania, acknowledge Jowsey's critique, and suggest that the methods of ethnography are evolving to reflect the changing demands of health research; where expectations of ‘time in the …