Introduction: The purpose of this study is to improve care by supporting clinical decision-making by identifying design requirements for computerized cognitive aides and communication tools. Methods: This project is divided into three phases: foundation research, prototype development, and prototype assessment. In phase I, we conduct one-week data collection visits in a BICU followed by data analysis sessions. Each visit includes: 1) Direct observation of clinical teams providing patient care. Probe questions enable researchers to request background and clarifying information situated in context to better understand motivations, information use, and decision making; 2) Structured interviews elicit knowledge from clinicians about their background, perspectives, work activity, information sources, and challenges they face; 3) Collection of computer-based and hard copy artifacts that clinicians use in their work. These include sign out sheets, personal notes, status boards, and equipment displays, among others. Through data analysis, we develop descriptive models of decision-making and patient care that can be used to understand the inventory of information used by clinicians. These models suggest the content and flow of information the project's prototype cognitive aide and communication system will help to manage. Results: Preliminary results have identified the complex network of human relationships that clinicians maintain and negotiate to provide patient care. For example, the bedside nurse is the central figure of an ICU patient's clinician network. Daily, this individual maintains at least 35 distinct relationships with other hospital entities to effectively provide patient care. Furthermore, the nurse is informally responsible for reconciling numerous conflicts among information sources, protocols/guidelines, unit policies, physician orders, consultant recommendations, priorities of care, patient and family preferences and requests. Conclusions: Use of CSE can identify the complexities of patient care that face individual clinicians and the decisions they make. This understanding may be used to develop requirements for a computer-based decision support to assist communication between network entities, improving decision-making and collaboration.
(C) 2013 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins