ABSTRACT
Adoption of technology has been proposed as a means to reduce medical error (Kohn, Corrigan, & Donaldson, 1999). This reduction might be derived through automation of tasks, monitoring, and improved flexibility of data handling directed toward better situation assessment (Sarter, Woods, Sc Billings, 1997). Studies of accidents, however, consistently demonstrate that new computerized systems predictably affect human problem-solving ability in ways that often contribute to accidents. Experience with technology’s contribution to new failures is widespread; ranging from personal cata strophes by unintentional but unrecoverable key strokes that wipe out entire files (Norman, 1983), to software problems that crashed NASA’s Mars exploration mission (Young et al., 2000). As health care moves toward increasing dependence on com puterized tools-order entry, electronic medical records, and medication administration to name just a few-the design of these computerized tools to support human work (e.g., make it easier, faster, safer, and more accurate) also increases in impor tance. Recent reports including the failure of com puterized order entry at Cedars Sinai (Chin, 2003), unexpected consequences with implementation of patient care information systems reported by Ash, Berg, and Coiera (2004), and bar-coded medication administration (BCMA) by Patterson, Cook, and Render (2002), lends credence to this view. Cognitive work analysis (CWA) is a method that models how environmental, organizational, individ ual, and technical constraints contribute to work in
order to design tools that support work. This methodology, applied to health care systems, could improve safety, efficacy, efficiency, and the accep tance of computerized tools by health care providers.
