Complex decision making in intensive care: The role of medical expertise

The study of expertise has led to insight into the nature of expert performance in a variety of problem solving domains. In medicine, considerable study has focused on diagnostic reasoning, however the role of expertise in coping with decision complexity has remained to be more fully explored. In this thesis three groups of subjects, consisting of medical students (novices), residents (intermediates) and intensive care experts were each presented with complex cases of intensive care problems containing varying levels of uncertain and conflicting evidence. The subjects were asked to think-aloud as they worked through the problems and provided a management and treatment plan for each case. The audiotaped protocols were coded for key process variables in decision making and problem solving.; The results indicate that the strategies used by novices, intermediates and experts for dealing with complex cases containing ambiguous information varied considerably. When faced with anomalous evidence, expert physicians were found to disregard the anomalous evidence and focus on the patient's overall clinical condition. In contrast, non-experts used the anomalous evidence to evaluate diagnostic hypotheses and drive the decision making process. Expert physicians were also found to focus on situational aspects of the patient's condition to a greater extent than non-experts, occasionally employing recognitional strategies. Strategies for processing anomalous evidence were shown to relate to level of confidence in decision making.; The research has a number of important theoretical implications for the study of decision making and expertise in general and medical decision making in particular. A theoretical account of the role of expertise in complex decision making is provided which emphasizes the interaction between situation assessment and evaluative reasoning processes in the decision making of experts. In addition, the thesis provides empirical support for aspects of several emerging naturalistic models of decision making, and extends them to include greater consideration of processes involved in evaluation of evidence. Implications for education and design of decision support are discussed.