Telling Stories with Data

Abstract

The fidelity provided by rich, nuanced ethics consult narratives does not proscribe efforts to advance the profession by using data to assess performance and demonstrate value. While these two approaches have been described as in conflict with one another, the former sets the bar to which the latter should aim; to achieve this, consult data should, minimally, do two things: (1) tell the story of the case, as best as possible, in language easily accessible to both ethicists and non-ethicists alike; and (2) do so in a way that elements of interest can be aggregated, disambiguated, and analyzed. Here we present a three-component system designed with that in mind. First, the Armstrong Clinical Ethics Coding System (ACECS) produces a set of six codes that tell a simplified story of the case. These six codes form a meta narrative, which, like meta data, can be aggregated, disambiguated, and analyzed. Second, we use a data model to uncover the stories those data tell. The results allow us to understand relationships, recognize patterns, and reveal insights that are otherwise not discoverable when using narratives alone. Third, due to tremendous heterogeneity among cases, we represent findings visually using techniques borrowed from data storytelling, a powerful means to demonstrate the value of a consult service to administrators and others. The ability to derive meta narratives from an aggregate of case narratives allows us to describe, in detail, the work of our consult service in a manner directly comparable to other services that use the system while we preserve unprecedented case-level nuance. If adopted broadly, this functionality suggests a means to undergird the legitimacy of the ethics consult practice itself in an evidence-based manner similar to that which undergirds other domains of healthcare.