Systematic observation in healthcare: Utility and limitations of a threat and error management-based safety audit

Sammanfattning: Improving teamwork has become a major safety goal for healthcare organizations. Audit tools currently available are useful, but they remain inadequate because they are reactive and fail to provide context for “the interaction between people and the operational context (i.e., organizational, regulatory and environmental factors) within which people discharge their operational duties” (Maurino, 2005). Accurate and relevant information about real-world team behavior is theorized to confer the ability to address, through process design &/or training, significant issues which can then be re-assessed through repeat observations. In the mid-1990s, the Federal Aviation Administration (FAA) funded collaboration between the University of Texas and Continental Airlines to directly observe in-flight behaviors associated with safety and risk. This methodology, now known as the Line Operation Safety Audit (LOSA), was instrumental in developing the Threat and Error Management (TEM) model of cockpit work performance. In 2006, the FAA made TEM-based LOSA a “voluntary safety recommendation,” and all major US commercial air carriers engage in this on a regular basis as a component of their safety management systems (FAA, 2006). This thesis describes the adaptation of LOSA to a Threat and Error Management-based Clinical Operation Safety Audit (COSA), and reports a series of 30 observations of trauma team activations in the Emergency Department at an American College of Surgeons accredited level 1 trauma center in the United States of America. Results of these observations showed discrepancies between work as designed and as executed, as well as other behaviors, associated with increased risk to patients. Analysis of data revealed important areas for targeted improvement based on risk created by the healthcare system during normal clinical operations. Systematic observation following the COSA protocol can become a vital and essential new tool to assist in improving patient safety in healthcare. The bulk of this thesis considers the criticality of context in work analysis throughout the discussion section. Though concepts of threats and undesired states were easily adaptable to healthcare, error was found to be too narrow a concept. I therefore propose discarding error for a more open and inclusive interpretation of performance: Task Adaptation. We therefore propose to widen our scope and continue to develop Threat Management and Task Adaptation-based COSA throughout the hospital to enhance system performance and improve patient safety.