OTHER STUDIES OF "VIGILANCE" AND DECISION-MAKING BY
ANESTHETISTS
Studies of "Vigilance"
The dominant metaphor for the mental activity of the anesthetist
has been that of a vigilance task. "Vigilance" is the motto on the seal of the ASA.
What is vigilance, and to what degree does it capture the complex nature of the
anesthetist's work? "Vigilance, or sustained attention refers to the ability of
observers to maintain their focus of attention and to remain alert to stimuli for
prolonged periods of time" (J. Warm, presentation at the Panel on Vigilance, ASA
annual meeting, 1992). There is an enormous literature concerning vigilance. Many
laboratory studies have demonstrated vigilance decrements during prolonged vigilance
tasks that are exacerbated or ameliorated by a variety of factors. These studies
have generated considerable controversy. Some psychologists believe that the results
of laboratory studies of vigilance have little application to complex real-world
task domains.[244]
[245]
Certainly the anesthetist's work involves vigilance as a necessary component, because
if new stimuli are not perceived, no meaningful work can be accomplished. However,
the anesthetist's task is much more complex than just vigilant alertness to stimuli.
Therefore, vigilance is a necessary but not sufficient condition for appropriate
performance.[64]
Several studies have attempted to quantitate the vigilance of
anesthetists to changes in clinically important variables using low-fidelity simulation.
Beatty and associates[246]
had anesthetists watch
for changes in displays of six vital signs on a video monitor. Denisco and colleagues
[247]
used videotapes containing abnormal changes
of anesthesia flow meter settings and physiologic monitor displays. The raw reaction
times were not reported—only the "vigilance scores." Such studies purported
to demonstrate a degradation of performance for sleep-deprived and fatigued residents,
but there were methodologic flaws. For example, in the study of Denisco and coworkers,
[247]
the subjects were never told the threshold
of change for them to report.
In addition, the work environment of the anesthetist is much more
complex than that presented in these low-fidelity simulations. Although it can be
argued that complexity will worsen vigilance performance, this is not necessarily
the case. Complexity can combat boredom, which is a distinct possibility in vigilance
experiments. In addition, the real work environment often provides redundant data
cues of changes, offering multiple possibilities for their detection.
A study of rested anesthetists[248]
using a realistic anesthesia simulator (also see Chapter
84
) measured detection times (first awareness of a problem) for a variety
of intraoperative events embedded within a realistic case situation. Of interest,
an event that caused alarms to sound immediately, such as ventricular tachycardia/fibrillation,
was detected in 10 seconds or less (confirming its veracity and acting on it took
longer in many instances, however). Another event, occlusion of the intravenous
line, was detectable only by visual observation in the direction opposite from the
anesthesia machine and monitors. This event took several minutes (on average) to
detect, but it was corrected quickly after detection. The redundancy of cues in
the task domain was confirmed in that six different observational modalities were
used by at least 1 of the 19 residents studied to detect an endobronchial intubation
initially.
Other vigilance studies have focused on detecting the loss of
an existing monitoring modality during actual patient care. In one study,[249]
the esophageal stethoscope was occluded with a clamp following a staged distraction
(conversation or a loud noise). Subjects most commonly detected the occlusion by
observing the clamp, rather than by noting a loss of sound. Although this study
purported
to show a difference in vigilance related to the use of automatic blood pressure
measurement devices, the artificial distraction could have biased the results; the
subjects typically failed to hear the loss of sounds, and there was a significant
result for only one group of residents, arbitrarily stratified by length of training.
Another study[250]
that evaluated reaction time
to loss of esophageal stethoscopy more carefully showed that 13% of occlusions were
detected after 1 minute. However, this study required a manual response, and the
investigators noted that in some instances of delay, the subject was involved in
a manual clinical activity such as administering blood or drugs.
Apparent reductions in vigilance could be due either to a reduction
of "spare capacity" to attend to the stimuli immediately or to a reduction in the
overall alertness of the anesthetist. Enhancements in the display and annunciation
of relevant information would be a useful ameliorative strategy in the first case,
whereas this strategy would not be useful if the anesthetist's overall alertness
is impaired, as by sleep deprivation or illness. The influence and impact of fatigue
and possible countermeasures are covered later in the section on performance-shaping
factors.
