A defined minimum data set. Will it work for direct patient care?

Abstract

As health care becomes geographically and institutionally distributed, sharing clinical information becomes necessary for efficiency but harder to achieve. The computerization of patient data promises to facilitate its sharing and reuse. The kind and specificity of data needed, however, vary with the intended use, so defining a data set that is limited in size but broad in application has been problematic. Collecting information that is relevant to direct patient care, and useful for ancillary purposes, requires an understanding of how clinical data are recorded and used. There must be agreement on the vocabulary and the definitions of words. To support clinical decision-making, it must be possible to represent information at various cognitive levels and to different degrees of specificity. Accommodation of ambiguity and uncertainty should be possible. The contextual, temporal, and relational properties of clinical facts must be capable of representation in the data that are shared.