BACKGROUND/RATIONALE:
To improve patient safety, it is critical to understand how clinical systems actually work, what factors make them work well (or not so well), and why adverse events occur. Human factors techniques used in other domains permit extraction of detailed information about system performance and risks to safety from any deviation beyond expected or routine system function. We investigated a similar approach in medicine, defining a non-routine event (NRE) as any event that is perceived by clinicians or skilled observers to deviate from ideal care for that specific patient in that specific clinical situation.
OBJECTIVE(S):
The goal of this project was to demonstrate the value of studying NREs during actual patient care in the field of anesthesiology. We sought to determine if rigorous delineation of the factors contributing to the occurrence of, and recovery from, NRE facilitates understanding of what distinguishes safe from unsafe care. Using anesthesia as a model of a structured medical work environment, we collected prospective and retrospective data about NREs, developed a useful classification of these events, and began to relate them to clinical outcomes.
METHODS:
Direct observation and videotaping of NREs during actual patient care included behavioral task analysis and measurement of workload and situation awareness. A sophisticated data collection and analysis system was designed and constructed. NRE were also identified by systematic query of clinicians in the recovery room, by a traditional QA reporting system, and by chart screening. Structured interviews examined providers’ knowledge and decision processes with regard to NRE etiology and management. Each NRE was categorized and logged in a database. Custom software and hardware tools were developed to facilitate event analysis.
FINDINGS/RESULTS:
We collected video data from 404 elective surgical cases representing a cross-section of anesthetic techniques, surgical procedures, and patient complexity. 34% of these cases contained at least one NRE. More than one NRE occurred in 23% of the NRE-containing cases, with a total of 187 NREs across all cases. The patient was affected in 78% of NREs (36% of all cases) and some injury occurred in 19% of NREs. NREs that resulted in patient impact (events associated with detectable, undesirable physiological changes or injury) occurred in 78% of NREs (36% of all cases) and patient injury occurred in 19% of NREs (8.9% of all cases). Airway management NRE were the most frequent type of NRE. In an interim multivariate logistic regression using 332 cases, provider experience (OR of 1.62; 1.03-2.56) and difficulty sleeping the previous night (OR of 2.00; 1.03-3.90) were significant. To assess further the epidemiology of anesthesia NREs, we have studied 412 cases from anesthesia providers in the recovery room (RR) using a retrospective survey instrument. The incidence and distribution of NRE was similar to those collected on video: 38% of the RR cases contained at least one NRE. The patient was impacted in 56% of RR NREs, with 17% associated with patient injury. The overall yield of events using the NRE construct was significantly higher than with traditional QA methods.
IMPACT:
This research examined the potential value of new methods, tools, and approaches to evaluating the safety of clinical systems. Preliminary data suggest that prospective collection of NREs will not only yield knowledge of risks to patient safety but may also provide insight into fundamental issues of human cognition and the nature of expertise. Future work will apply these methods to other clinical domains, and will assess the impact of specific safety interventions.
PUBLICATIONS:
Journal Articles
- Howard SK, Gaba DM, Smith BE, Weinger MB, Herndon C, Keshavacharya S, Rosekind MR. Simulation study of rested versus sleep-deprived anesthesiologists. Anesthesiology. 2003; 98(6): 1345-55; discussion 5A.
- Weinger MB, Dunn EJ, Cohn F. Dying of gallstones. The Hastings Center Report. 2003; 33(1): 14; discussion 15-6.
