An unexpected finding in our study was that the review of guidelines prior to the simulation test was associated with a statistically significant, almost eight-fold, increase in the delay in requesting the EMT to help in the patient's compartment.
This finding suggests a possible 'inhibitory' effect of reviewing guidelines on resource thinking. A possible explanation is that focusing the working memory on technical guidelines impairs the physicians' overall perspective, as working memory has a limited capacity. The limited capacity may again contribute to information overload and consequently compromise the overview.
It has been documented, that lack of overview jeopardize teamwork and consequently decrease the standard of the resuscitation [4, 12, 16, 25–30].
Although not statistically significant, a similar pattern of lowest delegation rate in the review group than in the other two groups was seen in delegation of chest compressions and defibrillation.
The incidence of successful resuscitation, in real life, increases significantly with the performance of on-going chest compressions [31–35]. Therefore, the hands-off ratio may be the most important parameter in the evaluation of resuscitation tests . Never calling for help from the driver was associated with a statistically significant increase in hands-off ratio, rising from 39% if help was requested to 54% if not (p < 0.01). This finding suggests that stopping the ambulance and unifying all available human resources in the patient's compartment may be better than keeping the ambulance moving, thereby leaving the EMT in the driver's cabin.
Strengths and limitations
This simulation study was performed in an authentic setting, as a real ambulance was used and the EMTs on duty on the day of the experiment comprised the team together with the junior physician studied. By choosing this model, it became possible to study the isolated performance of a junior physician working in a small resuscitation team as opposed to a larger, in-hospital team, with senior physicians taking the lead.
The small-intervention format was chosen because the interventions should be easy to administer, should cause no delay in resuscitation attempts, and, in case of success, be easy to implement. The interventions were tested on junior physicians because we had previously shown how this group of physicians showed large variations in their resuscitation skills .
While simulation offers many advantages, it also has potential problems. First, it is impossible to keep all variables completely stable, as different EMTs and ambulances were used according to time and place for the simulations. Second, the set-up increases the complexities of the tasks facing the junior physician because he or she has to deal with a number of unfamiliar challenges: should medicine be prepared in syringes in advance, where to put the medicine bag, and where to place oneself, just to mention a few choices that must be made by the junior physician. These drawbacks and potential sources of bias of our "mobile laboratory" are, however, worthwhile in our opinion, as the setting allows us to get a better understanding of the real performance of young physicians compared to a conventional laboratory setting.
Participants were all volunteers. This could introduce bias if only physicians proficient in ALS - or feeling so, at least - would participate, perhaps after refreshing guidelines. If this was the case, the results may be even more relevant as these would then reflect the better part of the group and thereby overestimate the skills in the underlying population. Regarding the number of participants, an unfortunate, but frequent, problem was, that participants who had already volunteered to participate were hindered in participation due to extreme workload, changes in work-schedule, exhaustion after night-duty, or alike. This caused the number of participants to be lower than we expected from our previous experiences .
Comparisons of the three groups (control, team briefing, and review) showed no differences when it came to technical resuscitation skills. This could be because the interventions were too small-scaled to trigger results. However, it is also possible that the lack of statistically significant effects is due to the possibility of a type 2 error (rejecting an effect that is there) due to small sample sizes (13, 16, and 17 participants in the groups, respectively). Another explanation could be the possibility of a ceiling effect, that is, if the participants had high levels of resuscitation skills before the experiment it would leave little room for improvement. Also, the intra-group variation was large and might have outweighed the inter-group variation and thereby masked possible differences. Finally, imprecise data collection could also be an explanation for the large variations, however, as the intra- and inter-observer variability coefficients were all very high (> 0.98) this is unlikely.
Fixation errors are known from clinical practise: information that should change the course of the resuscitation attempt is ignored (the team leader believes cardiac arrest is due to myocardial infarction and ignores intelligence about ingestion of toxic substances, e.g.) . The underlying mechanism may be inattentional blindness, a term from the psychological literature, describing how a preoccupied mind may fail to shift attention intentionally . Rehearsing guidelines may cause the physician to focus strictly on following guidelines and thereby miss being a team leader. A parallel to this can be the first phase of learning new skills: focus of the novice will be to understand the task and avoid mistakes .
A team briefing may help the junior physician to delegate tasks, as it ensures that relevant information is shared between the team members and helps the team to agree on goals, identify key priorities, and identify ways to reach these goals [4, 13–15, 40]. It may be especially important to help junior physicians to value teamwork, as they often feel time-pressured and therefore may be reluctant to invest time in team briefings .
Hunt et al. have advocated for team briefing prior to hospital transfer of critically ill patients . In a recent study, Westli et al. investigated teamwork skills and behaviour correlated to medical treatment quality of trauma teams in simulated settings. They suggested that trauma teams could be significantly more effective if communication and information exchange skills were strengthened by team briefing and establishing a shared mental model . This is further emphasized by Neily et al. who recently published how a medical team training program reduced surgical mortality during a three-year period with more than 180,000 procedures examined . Hunziker et al. compared the influence of a short leadership instruction versus a short technical instruction in a high-fidelity cardiac arrest simulation . Among their findings was, that leadership instruction resulted in a better overall performance regarding time to first compression and hands-off time, while technical instructions, although improving e.g. arm position during chest compressions, did not. Our findings seem to be coherent with Hunziker et al. as we found review of guidelines delayed requesting help
The use of checklists and timeouts has been shown both to improve understanding and to mitigate potential errors, especially in elective surgery [15–18]. Our study suggests that similar advantages can be made with the transportation of unstable patients.
Rall et al. has proposed the "10-seconds-for-10-minutes-principle", a metaphor for a formal "time out" during teamwork. They argue, that time "lost" by a quick team briefing (the "10 seconds") is "won" by a massive increase in team efficiency that saves time, improve treatment, and increases safety (the "10 minutes") . Translated to practise, they teach the team leader to take the time needed, take a deep breath instead of making a quick diagnosis and start treatment within a second, but make a formal time out instead .
Our structured team briefing intervention took place under relaxed circumstances, and the physicians were told to use the time they felt necessary resulting in a mean time consumption of 206 seconds (approximately 3 1/4 minutes). Our data do not allow us to extrapolate to time consumption in real clinical life. It does however equal findings from Lingard et al. in an evaluation of a preoperative checklist where 92% of team briefings took between 1-4 minutes . Thus it seems plausible that a structured team briefing could be applied at real inter-hospital transportations without consuming too much of precious time.
Our results suggest that reviewing guidelines might compromise the ability to focus on other aspects of resuscitation, such as teamwork, warrants the need for further studies focusing on how to avoid this cognitive impairment. However, our recommendation goes further than just a proposal for conducting additional studies: we believe it is necessary to perform this under very realistic settings, or, as Yeung and Perkins has stated, to do 'road testing of how treatment algorithms derived from evaluation of resuscitation science interact with each other before their application in real life.'