|
The empirical data can best be interpreted in accordance with an explicit model of the cognition involved in the anesthetist's performance. Several investigators[62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] have written about the cognitive elements in anesthesiology.[65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] The model developed here is explicit, comprehensive, and specific to anesthesia. It is described in detail as a framework for understanding the empirical data and provides a vocabulary for discussing the elements of both successful and unsuccessful performance by anesthetists. This model draws heavily from the work of a number of other investigators[63] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] studying human performance in a variety of complex, dynamic worlds.
The entire model, which is shown in Figure 83-1 , depicts the anesthetist as working at five different interacting cognitive levels to implement and control a core process (see Table 83-2 ) of observation, decision-making, action, and reevaluation. The core process must then be integrated with the behavior of other team members and with the constraints of the work environment.
The division of mental activities into levels follows the work of Rasmussen[35] [96] [97] and Reason.[82] [98] [99] Having multiple levels supports parallel processing (performing more than one task at a time) and multitasking (performing only one task at a time but switching very rapidly from one task to another). The anesthesia task analyses[104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] and the direct observations of anesthetists in Toronto[100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] and Tuebingen[111] [112] [113] have provided clear evidence for the occurrence of parallel processing and multitasking. Table 83-1 gives an overview of the mental activity levels.
At the sensorimotor level, activities involving sensory perception or motor actions take place with minimal conscious control; they are smooth, practiced, and highly integrated patterns of behavior. At the procedural level, the anesthetist performs regular routines in a familiar work situation. These routines have been derived and internalized from training and from previous work episodes. A level of abstract reasoning is used during preoperative planning, and intraoperatively it is used in unfamiliar situations for which no well-practiced expertise or routine is available from previous encounters.
Rasmussen's model[96] [97] was extended by the explicit addition of two additional levels of mental activity that provide for dynamic adaptation of the anesthetist's own thought processes. This ability to "think about thinking" in order to control one's own mental activities strategically is called metacognition by psychologists, [84] [93] and it is
Figure 83-1
Cognitive process model of the anesthetist's real-time
problem-solving behavior (see text for detailed description). Five levels of cognition
operate in parallel. The core process involves a main loop (arrows)
of observation, decision, action, and reevaluation. The core process is managed
by two levels of metacognition, involving supervisory control, allocation of attention,
and resource management (above the core process).
Each component of the model requires different cognitive skills, and each component
is vulnerable to a different set of performance failures or "errors." BP, blood
pressure; CO, cardiac output; MAP, mean arterial pressure; SVR, systemic vascular
resistance. (From Gaba DM, Fish KJ, Howard SK: Crisis Management in Anesthesiology.
New York, Churchill Livingstone, 1994.)
Level of Control | Explanation | Comments |
---|---|---|
Resource management level | Command and control of all resources, including teamwork and communication | Incident analysis shows a huge contribution of lacking resource management and communication skills to the development of incidents and accidents; the importance of these factors is reflected in the ACRM principles and simulation training courses (see Chapter 84 ) |
Supervisory control level | Metacognition: Thinking about thinking | Dynamic adaptation of the thought process, decision-making (e.g., avoiding fixation errors), scheduling and remembering actions (e.g., prospective memory tasks) |
Abstract reasoning level | Utilization of fundamental medical knowledge, search for high-level analogies, deductive reasoning | Often in parallel to other levels; in emergency situations often too slow and too sensitive to distractions in high workload situations |
Procedural level | Precompiled responses, following algorithms, heuristics, "reflexes" | Recognition-primed decision-making—experts are more often on this level; special errors may occur due to lack of checking for the appropriateness of the "procedure"; less experienced may misuse this level for not well-considered, unadapted "cookbook medicine" |
Sensorimotor level | Use of all senses and manual actions, "feeling, doing, hearing," sometimes subconscious control of actions | Experts perform smooth action sequences and control their actions by direct feedback of their senses (e.g., action sequences of placing an IV line or endotracheal intubation; skill-based errors like slips and lapses may occur |
1. Observation |
2. Verification |
3. Problem recognition |
4. Prediction of future states |
5. Decision-making |
a. Application of precompiled responses (recognition-primed decision-making) |
b. Decision-making using heuristics and probability |
c. Decision-making including abstract reasoning |
6. Action implementation |
7. Reevaluation (avoiding fixation errors) |
8. Start again with 1 (loop continues) |
|