Study design
This was a prospective observational study, conducted at a tertiary hospital over a 3-month period.
Participants
All registered trainees in critical care specialties (anaesthetics, EM, and ICU) were eligible to participate and there were no exclusion criteria. Informed consent was sought and obtained. Participants were assigned a study number in order of recruitment to blind the supervising investigators.
Procedure
Each participant completed i) a demographic questionnaire and ii) two psychological measures, and iii) a treadmill exercise test according to a modified Bruce Protocol (15) to determine maximal heart rate. Heart rate data was recorded using a heart rate monitor worn during clinical duties for a 3-month period. Participants completed an airway intubation diary for each intubation performed. Finally, participants attended a simulation session in which they were required to manage an endotracheal intubation.
Measures
Demographic questionnaire
Prior to participation in the study, each participant completed a questionnaire including: demographic characteristics including age, clinical role in the hospital (anaesthetics, ICU or ED), experience in that role (months), level of training, subjective level of expertise in intubation, and number of endotracheal intubations.
Psychological assessment tools
The DASS-42 is a self-reported scale designed to assess stress, anxiety and depression. It is a 42-item scale on which participants’ rate to what extent they agree with each statement. Each statement is associated with one of the above psycho-emotional states, with agreement ranging from ‘does not apply to me at all’ (0) to ‘applies to me very much’ (3). A higher score indicates a higher level of that particular state (16).
The STAI-AD was used to measure baseline anxiety associated with a subjective stress response. It consists of a ‘state anxiety’ and a ‘trait anxiety’ scale used to measure the anxiety associated with the participant’s experience of stress. The STAI-AD consists of 20-items for each ‘state’ and ‘trait’ anxiety, with a total of 40 items (e.g. ‘I am tense’). Participants rate their level of agreement using a 4-point scale to each statement, regarding how they felt at that given moment and in general. Scores on the STAI-AD range from 20 to 80 with a higher score indicating greater anxiety. The STAI-AD was administered prior to participation in the study as a measure of participants’ baseline stress response, as anxiety is an emotional state strongly associated with the stress response (17).
Treadmill Exercise Test
Maximum functional heart rate (MFHR) was calculated for each participant using a treadmill exercise test according to a modified Bruce protocol (15) (see Appendix 1). Participants ran at each stage for 3 minutes before the speed and gradient was increased to progress to the next stage. MFHR was recorded and the test concluded at the point at which the participant was unable to continue, or when they had completed all levels of the modified Bruce protocol. For participants unable or unwilling to participate in the stress test, the MFHR was calculated at their age in years subtracted from 220 (18).
Heart rate measurement
Heart rate was continuously measured using a FitBit® Charge 2 (FitBit Inc 2007, California, United States) and recorded throughout the participant’s work day. Second-by-second heart rate data was extracted for analysis from 15 minutes prior to the recorded intubation time to 5 minutes post recorded intubation time. This 20 minute intubation time period was selected due to the nature of the study population and expected intubation events. It was anticipated that the majority of intubations would be elective procedures completed by anaesthetic trainees, with a lesser amount performed as emergency procedures in one of the three critical care environments. We estimated that preparation time for elective intubation would take approximately 15 minutes, therefore capturing any anticipatory heart rate changes. Any stress-inducing post-intubation events such as desaturation, misplacement and re-insertion of the tube, as well as recovery from the complex task induced stress response, were intended to be captured in the 5 minutes following intubation. Intubation time was based on the online diary entries made by participants.
Heart rate changes were observed in two ways: percentage rise (median) across the 20-minute intubation period, and percentage rise at point of intubation. The heart rate range was calculated by subtracting the baseline working heart rate (BWHR) from the maximum functional heart rate (MFHR). The BWHR was calculated by generating the median of the minimum heart rate from each intubation that the participant recorded from the clinical environment. The BWHR was selected as trainees were only expected to wear the FitBit® when they were in the workplace. The heart rate change was selected as our comparison data over absolute heart rate values as we anticipated variability in the heart rate ranges within individuals in the study population. Clinical heart rate changes were collected as a median for all clinical intubation data entries, the resultant number then compared with the heart rate changes from the simulation scenario.
Clinical airway intubation diary
Participants completed an online diary for each intubation. Data collected included: date and time of intubation, type of intubation, emergency case (yes/no), subjective case difficulty, perceived stress level, level of supervision, and the Samn-Perelli 7-point fatigue scale (1 = fully alert, wide awake, 7 = completely exhausted, unable to function effectively). As studies have shown that caffeine can affect heart rate (19), participants’ daily consumption of caffeine (tea, coffee or other caffeinated drinks) were recorded. Any regular medication was also noted.
Simulated Airway Scenario
Participants attended a simulated scenario requiring the trainee to perform an airway intubation. This took place in an artificial environment with a high-fidelity mannequin representing a deteriorating patient. The session was tailored to the participant’s current training environment (Emergency Department, Intensive Care Unit, Anaesthetic Department). This allowed the scenarios to maintain consistency in timing of phases of clinical deterioration, diagnostic difficulty, and management difficulty. We did not attempt to adjust scenario difficulty for each individual’s perceived level of expertise as the trainees ranged from junior registrars to senior fellows and this would have introduced additional variability into the scenario. Rather, the scenario was designed to be a replicable emergency situation expected to be to manageable for all participants but difficult enough to challenge them. The simulation concluded one-minute post intubation, or after fifteen minutes if the intubation was not successful. See Appendix 2 for a description of the simulation scenario.
Data management and statistical analysis
As there were no comparable studies at the time which addressed the heart rate response to stress during airway intubation when comparing simulation environment with the clinical environment, sample size was determined for this study by participant availability. All data were analysed with SPSS software V23.0 (SPSS Inc. Chicago, IL), and graphed with GraphPad Prism version 5.00 for Mac (GraphPad Software, La Jolla, CA).
Demographic data were described using mean and standard deviation and heart rate data were described using median and interquartile range when appropriate. Median heart rate changes were selected given the small sample size, to reduce the power of outliers in the analysis.
Prior to all analyses and given the sample size of less than 50, a univariate normal distribution was tested by the Shapiro-Wilk test for the physiological responses to airway intubation. The level of statistical significance was set at p < 0.05.
Grouped data were analysed using a Wilcoxon signed rank test for comparison of median changes in heart rate over the 20-minute peri-intubation recording period between the clinical and simulation environment. Mann-Whitney U test was applied for comparison of median changes in heart rate at point of intubation between the clinical and simulated environment.
Individual data were analysed using Wilcoxon signed rank test was used to compare the heart rate changes at the time of intubation for individual study participants between the clinical and simulation environments.