Volume 6, No. 2 • Summer 1991

Meeting Targets Human Error in Anesthesia

David Gaba, M.D.; Steven Howard, M.D.

Anesthesiologists interested in the role of human error in anesthesia accidents solicited the help of experts on error and human performance at a Spring symposium to explore ways of optimizing the performance of anesthesiologists to improve patient safety.

The three-day meeting, thought to be the fast ever held on the subject of human error in anesthesia, was sponsored by the APSF and the FDA. The small group of invited anesthesiologists, engineers, and psychologists met in Pacific Grove, CA, to investigate research opportunities for combining expertise to advance relevant knowledge in their respective fields.

80% of Accidents

Recognizing that human error has been blamed for approximately 80 percent of anesthesia-related operating room accidents, members of the APSF executive committee organized the meeting with the hope of learning from the extensive research that has been conducted on human error and human performance in fields such as aviation, nuclear power, spaceflight, and military operations. Researchers from these are-as discussed how their findings might be relevant to the world of anesthesiology, while anesthesiologists discussed problems unique to the operating room and encouraged human performance experts to consider the medical arena as a new domain of study. Because the field of anesthesiology is less centralized and regulated than the other involved domains, it presents some unexplored challenges.

The meeting opened with a history of the investigation of patient safety in anesthesia and the recent creation of the APSF, presented by Ellison C. Pierce, Jr. M.D., of Harvard University. Mr. Joseph Arcarese of the FDA’s Center for Devices and Radiological Health described the rationale for the FDA’s sponsorship of the conference through its role as the regulatory body for medical devices. During the meeting he applauded anesthesia for being the only medical specialty trying to make profession-wide changes. Arcarese explained that the FDA was eager to become involved because of the opportunities to be influential in terms of patient safety and to use anesthesiology as a “window on & rest of medical practice.

Also during the opening session, Jeffrey Cooper, Ph.D., of Harvard, outlined the epidemiology of anesthetic mishaps and emphasized the growing realization of the role of human performance in anesthetic outcome. The “fierce independence” of anesthesiologists and paucity of protocols (including lack of standardized equipment), “macho” attitudes, and production pressure are among the factors that will make it difficult to improve anesthetic outcome through changes in human performance, he explained.

Basis for Decisions

The second session of the conference focused on human errors during dynamic decision making. International experts Jens Rasmmsen, Ph. D., of the Riso National laboratory in Denmark and James Reason, Ph.D., of the University of Manchester in the U.K., presented overviews of the cognitive psychology of human error Rasmussen defined three abstract levels of cognitive operations as skill-based, rule-based, and knowledge-based activities. In this model of dynamic decision making, mental and physical resources cycle among these cognitive levels during a rapid repetitive loop of observation of, decision about, and action upon multiple problems.

Dr. Reason explained that even though accidents are complex events that cannot be controlled, often the general failures that lead to accidents can he avoided, thus helping to prevent accidents. His Generic Error Modelling System (GEMS) incorporates Rasmussen’s classification of cognitive operations and emphasizes the matching of environmental cues to a store of rule-based responses. In other words, a person tends to respond in a way that is natural given the context of a situation. Reason also discussed the importance of “Went errors.” whose impact is masked until uncovered by evolving mishap.

Dynamic Decisions

Also in the context of dynamic decision making, David Woods, Ph.D., of Ohio State University and Veronique De Keyser, Ph.D., of the University of Liege in Belgium, discussed methods for studying the anesthesiologists’ task environment. Woods has analyzed experts’ comments at “mortality and morbidity” case conferences and has studied human/ machine interactions after new patient monitors were introduced in the Ohio State cardiac surgery operating rooms. De Keyser has conducted traditional structured interviews and direct observation of anesthesiologists during a variety of cases in an effort to map the causes of anesthesia incidents and to analyze the human factors that make error more likely.

Another session of the conference was devoted to the effects of fatigue and mental workload on human error. William Derment, M.D., director of the Stanford University Sleep Research Center, explained that a “sleep debt” can accumulate in people who consistency don’t get enough sleep. He indicated that chronic sleep deprivation is extremely common in our society and that the associated impairment of performance might contribute to anesthetic accidents, just as it is now believed to be a major cause of accidents in the transportation industry. Colonel Gerald Krueger, Ph.D. of the US. Army presented data from Operation Desert Storm noting that the most sleep deprived members of a group involved in sustained and continuous operations are often those with the highest level of responsibility He emphasized the need for developing rational sleep plans for all levels of personnel engaged in around-the-clock duties.


In the same session Daniel Gopher, Ph.D., of Technion Institute of Technology in Israel, explained that there are well-known limits of the individual’s ability to perform multiple simultaneous tasks. However, his research has shown it is possible to teach allocation of attention to allow the operator to cope with high workload in complex situations.

In the fourth session, Messrs. John Chappelow of the Royal Air Force in the U.K. and Robert Lee of the Bureau of Air Safety Investigation, Canberra, Australia, described similarities and differences between anesthesia and aviation. They felt researchers exploring human error in anesthesia could learn a lot from the field of aviation in terms of accident investigation, standardization of equipment and operational protocols, promotion of team coordination, and the organization of training.

On this last issue, Dr. William Runcimm of Royal Adelaide Hospital in Australia and Dr. David Gaba of Stanford University, advocated formal training for anesthesiologist on the management of crises in the operating room and described their research in this area. Runciman had developed a generic protocol for the initial management of an anesthetic problem. Gaba described the Anesthesia Crisis Resource Management (ACRM) training program he designed (with funding from the APSF). Modeled on pilots’ “Cockpit Resource Management” training, the course teaches anesthesiologists the fundamental skills of crisis management.

During the final session, Caba and Abigail Sellen, Ph.D., of the University of Toronto, presented critical summaries of the conference and directed discussion of a research and policy agenda concerning human performance in anesthesia. The group developed a set of broad research goals to provide the basis for improvements in the performance of anesthesiologists:

Define the “task characteristics” of the anesthesia domain more carefully, including a taxonomy of error modes and an analysis of successful adaptive strategies. This wig involve the application of existing models of dynamic decision-making to anesthesiologist behavior during both mA and simulated cases and to retrospective reports of crisis situations.

Determine the prevalence of sleep deprivation and fatigue in anesthesiologists during and after residency training. Evaluate the effects of fatigue on the performance of anesthesiologists.

Evaluate the effects of workload on human error during anesthesia. This will require the analysis of strategies by which expert anesthesiologists cope with high workload states.

Continue the development of simulator-based training programs in anesthesia crisis management. This will include the development of widely accepted standardized protocols for the initial management of perioperative crisis situations.

Analyze the organization of the “anesthesia system” to define the factors that have an effect on patient safety. “Latent errors” will have to be examined and the sociology and psychology of the complex network of incentives and disincentives that promote or detract from patient safety will be examined.

As pathfinders in this field, many researchers attending the meeting hope that further study of current problem of human error in anesthesia will lead to the ability to more effectively solve problems of the future and lead to improvements in other dynamic medical specialties such as intensive care or cardiology.

Open Meeting Planned

The group planned to generate a special interest committee within the Anesthesia Patient Safety Foundation to promote research and greater awareness in the anesthesia community of human performance issues as a limit to patient safety. An open scientific meeting to exchange research findings among a broader range of investigators has been planned for 1993.

Dr. Gaba, Stanford University and the Veteran’s Administration Medical Center, Palo Alto, CA is -Secretary of the APSE He and Dr. Howard of the Department of Anesthesiology, Stanford, University organized this conference.