Exploring the pathology of quality failings: measuring quality is not the problem - changing it is

The USA can boast a long history of investigation into quality failings in health care. From Ernest Codman and Abraham Flexner in the opening decades of this century through to the intense activity of the 1980s and 1990s, muc h careful study has exposed extraordinary and at times scandalous deficienc ies in the quality of care (Millenson 1997; Chassin & Galvin 1998; Schuster et al. 1998). Yet we are still far from developing 'industrial strength' q uality in health care: in all but a few isolated areas, such as general ana esthesia, 'six sigma quality' (i.e, a handful of errors per million) seems wishful thinking (Chassin 1998), Pockets of excellence and innovation notwi thstanding, the dominant experience of the past two decades has been an inc reasing ability to document quality failings and a seeming inability to mob ilize effective action (Coye & Detmer 1998). The rich literature on health- care quality that has sprung up over the past few decades has largely faile d to provide a clear direction for quality improvement activity. This paper analyses some of the reasons why this might be so. Contrasting the relativ e absence of progress on health-care quality with the relative success of d isease epidemiology provides some illuminating parallels. In essence, study of the quality of care has focused largely on providing a 'descriptive epi demiology'. Much more work is needed yet to unravel the underlying patholog y of quality failings, in order to empower development of an 'aetiological epidemiology' of quality in health care. Such understanding is essential as a precursor to targeted and effective preventative and remedial action.