AUTHOR=Lazaridis Christos TITLE=Deciding Under Uncertainty: The Case of Refractory Intracranial Hypertension JOURNAL=Frontiers in Neurology VOLUME=11 YEAR=2020 URL=https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2020.00908 DOI=10.3389/fneur.2020.00908 ISSN=1664-2295 ABSTRACT=

A challenging clinical conundrum arises in severe traumatic brain injury patients who develop intractable intracranial hypertension. For these patients, high morbidity interventions such as surgical decompression and barbiturate coma have to be considered against a backdrop of uncertain outcomes including prolonged states of disordered consciousness and severe disability. The clinical evidence available to guide shared decision-making is mainly limited to one randomized controlled trial, the RESCUEicp. However, since the publication of this trial significant controversy has been ongoing over the interpretation of the results. Is the mortality benefit from surgery merely a trade off for unacceptable long-term disability? How should treatment options, possible outcomes, and results from the trial be communicated to surrogates? How do we incorporate patient values into forming plans of care? The aim of this article is to sketch an approach based on insights from Decision Theory, and specifically deciding under uncertainty. The mainstream normative decision theory, Expected Utility (EU) theory, essentially says that, in situations of uncertainty, one should prefer the option with greatest expected desirability or value. The steps required to compute expected utilities include listing the possible outcomes of available interventions, assigning each outcome a utility ranking representing an individual patient's preferences, and a conditional probability given each intervention. This is a conceptual framework meant to supplement, and enhance shared decision making by assuring that patient values are elicited and incorporated, the possible range and nature of outcomes is discussed, and finally by attempting to connect best available means to patient-individualized ends.