Human Factors in Emergency Department eCPR: A Prospective Observational Study of Teamwork, Stress and Environmental Demands

doi:10.21203/rs.3.rs-8216191/v1

Download PDF

Research Article

Human Factors in Emergency Department eCPR: A Prospective Observational Study of Teamwork, Stress and Environmental Demands

https://doi.org/10.21203/rs.3.rs-8216191/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

BACKGROUND

Extracorporeal cardiopulmonary resuscitation (eCPR) is a high-acuity, low-occurrence (HALO) intervention requiring advanced technical skills and coordinated team performance. This study examined teamwork, provider stress and environmental conditions during Emergency Department (ED) eCPR events, interpreted using Reid and co-authors’ conceptual framework of self, team, environmental and system factors in high-performance resuscitation teams.

METHODS

We conducted a prospective observational study at two EDs with ethical approval. A pilot phase involving simulation and trauma activations was used to refine research logistics and methods. Recruitment for the eCPR study arm used convenience sampling with eligibility criteria including ED attempts at cannulation for adult cardiac arrest. Patient variables included demographics, Utstein elements and key eCPR timings. During eligible cases up to four providers were pre-consented and enrolled (cannulators, team leaders and defibrillation nurses). Team members were monitored for heart rate, salivary cortisol and psychometric stress-response indices. Global teamwork was assessed using the Mayo High-Performance Team Score (MHPTS). Environmental factors were measured via ambient noise monitoring (dB), team headcounts and observation of workflows.

RESULTS

Thirteen pilot cases and six eCPR cases (including 18 providers) were observed. Patients had a mean age of 44.7 years (SD 10.0), and five of six arrests presented with a shockable rhythm. Providers demonstrated significant physiological stress, with cortisol levels elevated compared to baseline (p = 0.0023). Team performance scores were consistently high (MHPTS x̄=12; IQR 10–13). Ambient noise peaked at a mean of 93.3 dB (SD 6.4), with overall mean levels of 75.2 dB (SD 7.9). Resuscitation rooms were crowded (headcount x̄=20; IQR 18–29). Despite stress, noise and crowding, team performance remained robust. Observations suggested that predefined roles, crowd-management strategies and regular verbal summaries helped maintain performance.

CONCLUSION

ED eCPR teams displayed consistently high performance despite substantial provider stress and challenging environmental conditions. Role clarity, crowd control and structured team summaries may help preserve team reliability. Targeted training in these behaviours could further strengthen ED eCPR delivery.

Extracorporeal cardiopulmonary resuscitation (eCPR) is an emerging intervention in Emergency Departments (EDs) for selected cardiac arrest cases.^1–3 Recent trials have shown inconsistent results, but in out-of-hospital cardiac arrest (OHCA) eCPR has been shown to consistently to improve outcomes.^4–6 Extracorporeal membrane oxygenation (ECMO) is challenging for teams to deliver because of the requirement for advanced technical and non-technical skills. Therefore this study foregrounds the exploration of eCPR team responses from a human factors perspective including detailed observations of teamwork, environmental and individual factors.²⁷

eCPR is recognised as technically demanding but also relies on a time-dependent and coordinated team response. Such cases share some characteristics with the more common ‘trauma call’ responses in the ED setting. Both are characterised by an unscheduled and rapid assembly of a multi-specialty team. Pre-assignment of key roles is desirable, and frequently observable, but trauma and eCPR teams rarely consist of the same providers on more than on occasion (unlike in other high-performance industries such as sports and military aviation). Moreover, eCPR represents a high acuity low occurrence (HALO) event and while some design and non-technical aspects align closely with trauma there are notable additional challenges. ^{7 8} These elements may broadly include limited lead-time for eCPR⁷, lack of regular cases³, stress, team preparedness (i.e., training within roles), unfamiliar workflow and ED environment factors (i.e., noise, lighting and equipment).⁸ Therefore, in this study the overarching aim was to observe eCPR cases for non-technical skills (NTS), environmental factors and impact on specific team members.

Yule and co-authors defined NTS as “cognitive, social, and personal resource skills that complement clinicians’ technical skills and contribute to safe and efficient task performance.” We considered this definition as relevant to eCPR delivery in the ED setting. We also leveraged work by Yule on approaches to measuring healthcare team performance known as ‘surgical sabermetrics’. Additionally, Hicks and Petrosoniak work on human factors in emergency resuscitation bays was also viewed as highly relevant to eCPR.^9–11 Yule’s surgical sabermetrics is defined as the application of objective metrics to assess performance in healthcare based on the use of analytics (i.e., sabermetrics) in professional sports.¹⁰ This study aims to add to the existing literature on eCPR by exploring ECMO responses through the lens of ‘self, team, environmental and system’ factors.^9–14

Justification for the study is further highlighted by outlining gaps in the current literature.^13–15 Despite growing eCPR exposure and evidence-based discussions of ‘if it works’ and ‘technical approaches’, previous studies often explore patient outcomes and are small due to limited eligible cases.^13–15 The patient impact and technical difficulties are also highlighted, but eCPR studies infrequently focus on the impacts on teams, team members or environmental considerations.^4,16,17 As a result, little is known about non-technical (human factor) factors during eCPR which include team coordination, situational awareness and managing conflict. Further understanding of these human and ergonomic factors is important for designing robust eCPR systems and training. This study addresses this perceived literature gap by exploring non-techincal factors that are relevant to eCPR. We used a multi-methods approach to explore the ‘non-technical aspects of eCPR delivery in the Emergency Department’, drawing on recent work on translational simulation, medical education, teamwork, ergonomics and system design.

Study Design

A prospective observational study of eCPR cases complimented by a translational simulation pilot phase. Descriptions and analysis of team performance during eCPR are foregrounded. The methods were adapted from the literature on ‘surgical sabermetrics’ and ED ‘design thinking’.^9–12 Key aspects are reported using a STROBE diagram (Fig. 1). Patients did not receive additional intervention(s) outside of routine care as the result of enrolment.

Study Setting

Ethics committee [Internal Review Board] (Ref 2021/ETH11970) approval was granted pre-commencement. The need for informed patient consent was waived by the committee due to: (i) impracticality in cardiac arrest; (ii) no patient data being identifiable AND (iii) no change in management. The staff members enrolled provided written consent for participation in the study.

Conduct adhered to Australian guidelines on use of data and privacy. We obtained approval for two eCPR practicing hospitals. Both sites (Westmead Hospital and Royal Prince Alfred) are Level 1 Trauma and Cardiac Surgery centres. These sites both provide similar eCPR care, including a highly selective ED-based eCPR responses for refractory cardiac arrest.

eCPR Model of Care

The eCPR model of care is similar to the established major trauma response systems at the two sites. For eCPR“CODE ECMO” was the standard term used for activation.¹⁴ These activations operate on weekdays-only between 0700–1700. The aligned approach to eCPR at the two sites has been previously reported.¹⁴ CODE ECMO activations send a text message and/or pager message to pre-allocated eCPR staff. The various checklists used are presented on a free smart phone ‘app’ available for both Apple and Android platforms (i.e., https://apps.apple.com/au/app/westmead-ecmo-app).

Translational Simulation Pilot (February 2022 - March 2022)

During a 2-month pilot study we observed in-situ simulations and trauma cases to test feasibility of data collection. Individual, team, and environmental observations were recorded by two investigators in real time (Appendix 1; Appendix 2). Validated scores of teamwork were also recorded (Appendix 3). Logistical data collection challenges were identified and a standardised data sheet optimised for use in eCPR cases. Pilot data is presented in Appendix 1. In the a-priori protocol additional variables were approved (movement tracking, pedometers) but were deemed impracticable and excluded after the pilot.

A notable component of the pilot (Fig. 1) was the deliberate use of simulation and pilot trauma cases to refine our data-collection processes before enrolling eCPR cases. This preparatory phase allowed to testing of feasibility, anticipation of researcher workflows, and efficient processing of stress-response sampling. Further, the pilot allowed environmental measurements and teamwork assessments to reliably be integrated in real-time. The value of simulation for refining research protocols is increasingly recognised; similar benefits have been demonstrated by Fatovich and co-authors who showed that embedding research into resuscitation simulation improved recruitment into complex clinical trials.¹⁵

eCPR Cases (March 2022 - December 2024)

eCPR cases were included based on availability of investigators to observe (AC, HC, AS). We received assistance from an ED research assistant when available. The inclusion criteria were: (i) patient presenting with OUT-OF-HOSPITAL cardiac arrest (OHCA); AND (ii) Activation of eCPR team (criteria aligned with international standards)’; AND (iii) informed consent of all directly observed team members (eCPR providers) in advance of the OHCA arrival. The Exclusion criteria (Fig. 1) were: (i) eCPR case age ≤ 18; (ii) in-hospital cardiac arrests; (iii) cardiac arrest > 1-hour after presentation to the ED; (iv) cardiac arrest following trauma; (v) refusal of provider(s) to participate; (vi) no availability of researcher(s) to observe. Key health providers (i.e., cannulators) were consented in advance where possible to avoid interferences with workflow and eCPR preparation.

Case Sampling

For a convenience sample of “observation of processes, team and health provider factors” we aimed to enrol > 5 eCPR cases and > 15 team (discrete) team members. No power calculations were made. As noted, core eCPR team members were enrolled in advance. Team members such as eCPR cannulators routinely know they are part of the team (i.e., carrying a pager).

Eligible key eCPR providers included Team Leaders (Medical, Crowd Control and Nursing), CPR (defibrillation) nurses, eCPR cannulators and airway doctors. For each case a capped number of team members were enrolled (i.e., maximum of four due to limited monitoring equipment). The local eCPR team on-boarding process encourages team members to be familiar with their role and take a team lanyard (Fig. 4) to identify their role during the case.

Outcome Measures

The overall approach was to adopt a pragmatic data collection strategy informed by prior work in trauma and simulation settings.^8,11 The choice of outcomes aligned with testing a hypothesis that ‘effective eCPR performance is dependent on alignment between individual stress regulation, coordinated teamwork, and contextual affordances.’ Accordingly, we selected feasible, low-cost measures from 4 domains (i.e., self, team, environmental and system). Observation of actual eCPR performance under real world conditions provided a robust but limited number of cases to observe for these factors.^{8 11} We took the learnings from work by Yule (sabermetrics), Hicks (ED trauma) and LeBlanc (team stress) as well as experience from the pilot to select a final list of feasible outcome measures.^{9 14 18}

In more detail, the sampling strategy sought to capture complementary dimensions of performance across key interrelated domains relevant to NTS performance in eCPR. These domains included * Individual Factors (focussing on individual response outcomes); ** Team Factors (focussing on teamwork outcomes) and *** Environmental Factors (focussing on environmental outcomes). The outcome measures used were conceptually linked rather than discrete. Specifically, the measures were chosen to broadly represent the interdependent domains within Reid’s 2018 framework (e.g., self, team, environment and system factors that relate to performance in high stakes resuscitations).¹⁸ The pilot work (Appendix 1) informed the feasibility of data collection and assisted with refining the data collection sheets. After the pilot study we adopted an integrated data collection strategy to explain how Reid’s domains. The specific measures (listed below) were chosen because each domain can interact and influence the others. For example, environmental noise may influence stress and impede teamwork - whilst clear system design and team training could mitigate such effects. In the resulting discussion section, we provide interpretations of the resulting data set within the context of the literature on non-technical skills and human factors in healthcare. This approach allows us to move beyond summarising results and instead discuss how the domains collectively shape performance during eCPR. This informs the overarching goal of providing recommendations for eCPR training and systems. For each eCPR case enrolled two researchers collected the following outcomes for each case:

(i) Utstein and eCPR Patient Outcomes - Utstein outcomes were available for each case using a pre-existing OHCA registry.³ Enrolment did not alter the clinical management of patients. Cases are reported with cardiac arrest outcomes relevant to appropriateness of selection (i.e. rhythm, time of arrest pre-hospital), the eCPR process (i.e., cannulation start time, finish time) and mortality at 1-day/28-days.

(ii) Salivary Cortisol*- Hypothalamic Pituitary Adrenal system activation can be reliably measured using salivary testing by Enzyme-Linked Immunosorbent Assay (ELISA). Cortisol ELISA sampling shows correlation narrow limits of agreement with plasma cortisol.¹¹¹⁹ Samples were taken both at baseline (matched time to eCPR case) and 0.5 to 2 hours following the eCPR case (i.e., cortisol typically peaks at around 45 minutes post event). Available participants were asked to roll a specific salivary sampling tampon in their mouth until saturated (i.e., 1-minute). Samples were placed in collection tubes and frozen for analysis at RPA hospital.

(iii) Heart Rate Monitoring*- Heart rate (HR) is considered a useful dependent stress measure.²⁰ Heart rate variability (HRV) is superior to raw heart rate as a measure, but we did not have resources to cover devices to measure HRV. Polar H10 mobile bands (Polar Electro, Kempele, Finland) supplied by the local Simulation Centre ($580 for purchase of 3) for the duration of study. These bands were linked to an IOS device by Bluetooth (https://support.polar.com/au-en/polar-team-app) that recorded HR continuously. Baseline HR was not measured so data is limited to descriptive reporting.

(iv) State Trait Anxiety Index (STAI)*- The State-Trait Anxiety Inventory (STAI) is a widely used psychological scale for measuring anxiety in adults. It was developed by Spielberger and colleagues. The STAI is divided into two self-report scales (Y1 State and Y2 Trait). In a previous studies of stressful simulations use of the scales were sensitive to anxiety increases.¹¹ The scores on each item are collated into a anxiety score, which can range from 20 to 80 with a high reported level of internal consistency.²¹ The state anxiety scale of the STAI consists of 20 statements (e.g., ‘‘I am tense’’) to which respondents indicate agreement on a four-point scale as to how they feel at the given moment. Paper forms were used to collect both Y1 and Y2 scores within 6 hours of the eCPR event.

(v) Mayo High Performance Teamwork Score (MHPTS)** - We measured the validated items of the MHPTS score for all cases. MHPTS is a score of teamwork with a score of 0–16 (Appendix 3). For a secondary measure to assess teamwork factors more specific eCPR we used a second team checklist derived from work by Weller and co-authors (Appendix 2).²² A combination of both MHPTS (Appendix 3) and the Weller derived checklist (Appendix 2) is relevant to eCPR because it occurs in a critical care team context.¹⁴²³ In the manuscript we denote the Weller checklist as the “Modified Auckland Score”. Investigators (two for each case) also took notes on major teamwork omissions. Discrepancy in scoring/observations/notes were resolved by consensus or by use of a median score.

(vi) Ambient Noise Level (Decibels)*** - Room volume was continuously monitored via a National Institute for Occupational Safety and Health (NIOSH) Sound Level Meter system. A calibrated microphone was connected to a spare dedicated smartphone.²² Sound levels have been measured extensively in operating rooms with similar methods.^23–26 A-weighted sound levels (LAeq) and the C-weighted peak sound pressure (LCpeak) were displayed on the device and recorded onto the paper sheet (Appendix 2). As this system provides a diachronic trace of these values, decibel values were manually recorded each minute, and the peak volume captured noted at patient departure. Microphones were placed to the centre left of the resuscitation room which is part of an open area Resuscitation Bay measuring approximately 60m² (Fig. 3).

Analysis Plan and Bias

Extensive comparative analysis was limited due to a limited sample size (Fig. 1). As a result, we describe the majority of the data set descriptively. Post-hoc comparative analyses using t-tests was applied by a progressional statistician to a limited number of normally distributed variables with larger sample sizes (i.e., noise level (DB). The analysis was conducted using IBM SPSS Statistics (version 19). There was no control for confounders or use of any sensitivity analyses. The literature provides a number of comparable studies which assist in interpreting and make meaning the specific results.

Table 1 shows characteristics of the participants enrolled by convenience sampling over a 2-year period. This period commenced after a pilot (Appendix 1) that enrolled 13 cases. Data were collected from eight actual trauma cases, one high-fidelity ECMO simulation, and four major trauma simulations (Fig. 1). For the eCPR main study 6 cases provided 155 minutes of observable case time and enrolled 18 healthcare providers.

Table 2 outlines the Utstein outcome data related to eCPR cases as well as observed team performance and ambient noise data. The criteria for “CODE ECMO” activation was appropriately triggered in all cases.¹⁴ 5/6 patients had a shockable rhythm and brief return of spontaneous circulation (ROSC) was achieved in two cases. There were no cannulation failures. One case was palliated in ED. Four patients were alive at 24 hours of which two survived to neurologically intact hospital discharge.

Individuals and Stress

Among healthcare provider participants (Table 1) there were balanced numbers of males and female (n = 9). Most providers had a medical role (n = 11). Missing data related to these individuals is outlined in Fig. 1. Reports for heart rate, STAI and salivary cortisol are presented descriptively in Table 1. Mean HR rate was 97.0 (SD 10.8). In terms of comparison between baseline and post eCPR the following comparisons were made (student’s t-test): (i) STAI baseline score mean was 35.5 (SD 8.0) versus post eCPR mean 41.8 (SD 9.67) (p = 0.13); (ii) Salivary cortisol baseline 2.53 nmol/L (SD 2.1) versus post eCPR 8.41 (SD 5.0) (p = 0.0023)

Teamwork

Teamwork scores were consistently in a high range (MHPTS x̄=12, IQR 10–13). Some ‘errors’ were observed in teamwork. Suboptimal team behaviours included gaps in frequency of team summaries, attention not consistently drawn to ‘safety threats’, and a lack of consistency in closed loop communication. Headcount was observed in all cases and the crowded eCPR attendances were all higher than 18 providers (Headcount x̄=20; IQR 18–29). Teamwork observations suggested team (leader) summaries were done more than once during each case. These were typically used to bring teams to the ‘same page’ rather than for team reflexivity or decision making. Despite the high scores for MHPTS significant periods of loud (inaudible) cross talk were noted at least 3 times in 4/6 cases.

Environmental Factors and Noise

Ambient noise data was collected, and all cases took place in the same resuscitation area with microphone in the same location. Baseline noise (without a patient) in this area was measured over 10 iterations at 64.8dB (SD 9.4) during the pilot study. Noise volumes during eCPR cases were in a high range with mean peak 93.3dB (SD 6.4) and overall mean of 75.2dB (SD 7.9).^23–29 Fig. 2 further illustrates the ambient noise level over time with the cannulation attempt timings superimposed. A cordon separating clinical and corridor space (Fig. 3) was used in 3 of 6 cases. Typically, ⅔ of attendees remained outside of the clinical space regardless of the use of a cordon. For example, in Case 1 many providers (n = 32) attended but only 11 team members (Fig. 3) were observed inside the designated resuscitation area.

Table 1

Description of eCPR Team Healthcare Providers
Team Members and Case Number	eCPR Role	Heart Rate mean (SD)	Age (years)	Gender (M/F/O)	Experience (PGY)	STAI (Y2 TRAIT) (Usual)	STAI (Y1 STATE) (Post eCPR)	Baseline Salivary Cortisol (nmol/L)	Salivary Cortisol 1-3h e-CPR (nmol/L)
Provider 1 (Case 1)	Medical Team Leader	109.8 (6.8)	31	M	7	Not Recorded	Not Recorded	Not Recorded	Unable to obtain within 3h
Provider 2 (Case 1)	Defibrillation Nurse	73.0 (8.2)	34	M	13	Not Recorded	Not Recorded	Not Recorded	Unable to obtain within 3h
Provider 3 (Case 1)	Anaesthetist Cannulator	86.6 (17.3)	43	M	15	46	46	Not Recorded	Unable to obtain within 3h
Provider 4 (Case 2)	Medical Team Leader	101.2 (7.4)	43	M	19	34	43	3.6	9.5
Provider 5 (Case 2)	Crowd Control Leader	Not Recorded	40	M	17	32	45	2.1	12.9
Provider 6 (Case 2)	Anaesthetist Cannulator	104.8 (14.8)	52	M	26	24	21	2.0	6.1
Provider 7 (Case 2)	Defibrillation Nurse	Not Recorded	29	F	7	35	35	0.6	2.7
Provider 8 (Case 2)	Defibrillation Nurse	Not Recorded	28	F	8	30	56	6.7	16.7
Provider (Case 3)	Medical Team Leader	91.5 (10.2)	50	F	27	Not Recorded	Not Recorded	1.9	8.1
Provider 10 (Case 3)	Defibrillation Nurse	103.8 (11.1)	28	F	7	52	50	0.6	2.7
Provider 11 (Case 3)	Crowd Control Leader	Not Recorded	39	F	13	37	46	6.6	14.1
Provider 12 (Case 4)	Anaesthetist Cannulator	Not Recorded	45	M	19	Not Recorded	Not Recorded	Not Recorded	Unable to obtain within 3h
Provider 13 (Case 4)	Defibrillation Nurse	Not Recorded	53	F	30	Not Recorded	Not Recorded	0.6	2.9
Provider 14 (Case 5)	Medical Team Leader	98.7 (10.9)	44	M	21	Not Recorded	Not Recorded	2.1	Unable to obtain within 3h
Provider 15 (Case 5)	Defibrillation Nurse	93.3 (3.5)	50	F	28	Not Recorded	Not Recorded	Not Recorded	Unable to obtain within 3h
Provider 16 (Case 6)	Medical Team Leader	Not Recorded	43	F	20	32	41	1.1	Unable to obtain within 3h
Provider 17 (Case 6)	Anaesthetist Cannulator	Not Recorded	45	M	15	Not Recorded	Not Recorded	Not Recorded	Unable to obtain within 3h
Provider 18 (Case 6)	Defibrillation Nurse	107 (9.3)	28	F	6	33	35	Not Recorded	Unable to obtain within 3h

Table 2

Description of eCPR Cases and Teamwork Factors
Variable		Case 1	Case 2	Case 3	Case 4	Case 5	Case 6
Utstein	Age and Gender (years, M/F/O)	44 - Male	53 - Male	52 - Male	30 - Male	33 - Female	56 - Male
	Presenting Rhythm	VF	Torsades	VF	VF	PEA	VF
	Pre-hospital Time (mins)	50	32	50	66	38	Pre-hospital eCPR sheaths
	Total Time in ED (mins)	31	19	27	38	28	12
	ROSC in ED (pulse returned during time in ED)	No	Yes	No	No	No	Yes
	24-hour mortality	Died	Alive	Alive	Died	Alive	Alive
	28-day mortality	Died	Died	Alive	Died	Alive	Died
Team	Highest team headcount during the eCPR resuscitation	32	18	29	20	18	22
	Noise (Db; median [range])	71 [60–90] (Peak 96Db)	75 [61–92] Peak (98Db)	73 [68–87] (Peak 96Db)	80 [60–92] (Peak 81Db)	74 [60–97] (Peak 97Db)	72 [62–81] (Peak 92Db)
	Mayo High Performance Teamwork Scale (Core Items) (Range of Scores Possible 0–16) --------------------------- Additional teamwork field notes from ‘Modified’ Auckland Team Score (Appendix 2)	12	10	13	12	10	14
		Team leader summaries: “sometimes” Instruction/ verbal communication explicit: “sometimes” 4 periods of inaudible cross talk	Team leader summaries: “No” // Attention to potentially hazardous actions/ omissions: “No”. 4 periods of inaudible cross talk	All Auckland Score Variables Checked No Periods of Cross Talk	All Auckland Score Variables Checked 3 periods of inaudible cross talk	All Auckland Score Variables Checked “Noisy at handover” 3 periods of inaudible cross talk	All Auckland Score Variables Checked “Not noisy”; “Many called timeouts by leader” “Quiet sign held up by leader”
	**At start attempt of eCPR cannulation:
	**Team Headcount (inside and outside the resuscitation room)	Inside 11 Outside 21	Inside 9 Outside 9	Inside 10 Outside 16	Inside 12 Outside 8	Inside 10 Outside 8	Inside 14 Outside 10
	***At ECMO flow start:
	***Team Headcount (inside/outside)	Inside 18 Outside 10	Inside 11 Outside 6	Inside 9 Outside 20	Inside 6 Outside 12	Inside 4 Outside 9	Inside 12 Outside 8

The application of human factors principles is important when planning the implementation of an eCPR system. ECPR events are inevitably time-pressured, stressful, and require rapidly assembled teams to perform difficult technical skills.³⁰ The study findings were not unexpected (elevated stress responses, large numbers of personnel (often 20–30) and high noise levels, yet overall strong teamwork). However, rather than viewing these results as standalone endpoints, the observations of eCPR serve as starting points for presenting evidence-based reflections on how individual, team, environmental and system factors interact in this context. Given that eCPR is a high-acuity, low-occurrence (HALO) event and that eCPR team composition is disrupted by a reshuffle for each case, even small improvements in non-technical skills may influence performance. ³¹ We structure the following discussion according to Reid’s four domains: (i) individual factors with a focus on stress; (ii) teamwork measured using the MHPTS; (iii) environmental influences such as noise and crowding; and (iv) system design considerations.^{8 31}

Individuals and Stress

The overall observation that eCPR teams experience stress aligns with the wider literature.¹¹ Specifically, elevations of cortisol levels were observed when compared between baseline and post-eCPR (Table 2). Observed raw heart rates were above expected baselines for typical adults and STAI questionnaire scores returned increased, albeit not statistically significantly. ²¹ The observed rise in cortisol (µ > 6nmol/L) can be considered significant in the context of the literature on technical skills performance.³² Further the cortisol increases also align with various studies showing spikes in markers of stress following team-based resuscitation.^{11 33–35} The overall picture appears to be that eCPR elicits a stress response which risks attenuating individual focus and overall performance.³⁷

An optimal level of stress (or at least mitigating the negative effects) is important during such events.^{35 36} If one considers stress as a continuum from low to high, both under arousal and unattenuated hyperarousal could both lead to underperformance.⁸ Limited data available on the effects of stress during HALO cases suggests unmanaged stress could be a negative influence. These observations suggest a need to provide specific training related to stress.^{8 35} For example, some researchers have concluded that stress leads to attenuated perceptions, limited ability to perform tasks and poor allocation of attention.³⁰ However, the degree of the impact(s) is uncertain. In our limited sample stress did not appear to impair teamwork (Table 2). The explanation could be that impact may be dependent on various factors such as training, experience and coping strategies.

‘Stress Inoculation Training’ has been cited as one training strategy that aids future performance under stress, and this technique is used in our geographical area. Proponents suggest that stress is ‘inevitable’ in healthcare settings.^37–39 The effectiveness of this training is hard to measure in empirical studies especially at a patient outcome level. ⁴⁰ As noted above, the observation of stress in providers was not unexpected. For example, the findings align with studies of providers performing cardiac surgery and healthcare learners experienced immersive scenarios.⁴¹

A notable randomised study by Lizotte and co-authors examined neonatal emergencies using simulation. The team tasks observed included cases of infant cardiac arrest which elicited significant stress in learners.³³ Despite a measurable increase in stress (at similar levels to this study) the team performance remained consistently high across all scenarios. Furthermore, high performance of teams was observed regardless of whether the outcome was survival or death. Providers that were trainees performed better on advanced tasks than on basic skills.

Observations of stress during eCPR events may also tie in with Folkman and Lazarus’ theories.⁴² Their work, which underpins much of the current theory on stress, suggests that performance may be determined by individual assessment of the balance of ‘demand’ and ‘resources available’ to cope with a situation.⁴² Based on this model if healthcare providers perceive that overall resources are not sufficient to match demand(s) then this risks overwhelmed individuals and declining performance. Conversely, balance in the levels of demands and resources can be processed as low threat. Such balanced perceptions lead to a “reframing” that the emergency is a challenge rather than a threat.⁸ The observations in this study of both high performance and high stress may reflect a similar phenomenon occurring at an individual level in eCPR cases.

The finding of eCPR causing stress reminds educators of the need to prepare teams.⁴³ As noted above some academics are proponents of stress inoculation training (SIT). This type of training deliberately stresses teams to relearn their stress responses to increase preparedness and therefore performance in future cases.^{35 36} Simulation experts have argued that the effectiveness of SIT is dependent on skilled pre-briefing and debriefing as well as careful calibration of the level of induced stress and active management of performance anxiety among learners.^{38 40 44} SIT may enhance performance but needs to be carefully designed with skilled facilitation to maximise benefits and minimise harms.⁴⁰ Conducted in an ED setting, one exploratory study examined SIT and involved 23 participants (12 doctors and 11 nurses) attending deliberately stressful training. One of the key findings from these authors' data resonates with our experience of observing eCPR. That is teams should focus on strategies that promote team ‘cohesion’ and ‘collective stress management’.⁴⁰ In this study we did not train eCPR teams with SIT, but we did orientate (on-board) team members to the eCPR process. Further, many of the staff involved in cases have received some form of team training in their departments or as part of ‘trauma team training’ exercises.

Teamwork

In terms of team performance, MHPTS scores were generally in a moderate to high range (Table 2). A notable limitation to this observation is the potential for bias in scoring. Further, the recorded number does not always reflect the actual lived experience of team members. Within this limitation, it is our view that teamwork was of a high level despite stress, crowding and noise.

In comparison to other studies that examined effects of pre-briefing, self-appraisal and mental rehearsal our observed scores for MHPTS were in keeping with higher-end team performance.^45–47 Furthermore, work by Sorensen and co-authors highlighted ‘high-quality communication, non-technical skills training, and shared mental models are performance safeguards for teams experiencing team stress’. We also observed these characteristics. For example, regular summaries by team leaders and effective communication by team members.^{30 48} Matching these findings with Edmondson’s work on teamwork principles may add value. Her emphasis on reflexive, adaptive teamwork could explain the team performance.

In the wider literature high performing teams commonly demonstrate ‘adaptive teamwork’, ‘adaptive expertise’ and are noted to have an ability to manage uncertainty.^{49 50} An area for future improvement would be ensuring providers are able to speak-up and thereby raising awareness of errors or risks, increasing frequency of summaries and reducing the periods of ‘cross talk’. All these problems were observed in the cases (Table 2).⁴⁸ The findings remind us that teaming is a learnable skill and that organisations seeking to improve outcomes should include teamwork in their training programmes.⁴³

Environmental Factors and Noise

Noise during eCPR events could be related to both communication and device operation.^25–28 We found a wide range of peaks and troughs of noise within each case and the critical events (cannulation and ECMO flow). On average the mean peak noise level of 93.3dB (SD 6.4) and an overall mean volume 75.2dB (SD 7.9). These levels were similar in peak to work by Alison and Co-authors but higher overall than the baselines reported across ED in their work.⁵¹ There is evidence in operating theatre settings that such metrics correlate with teamwork scores and may predict critical periods during a procedure.⁵² For example, noise peaks may indicate positive teamwork such as increased communication (i.e., ‘relational coordination’) but also indicate task-irrelevant noise that could add to cognitive load.^{27–29 53}

We acknowledge limitations in using ambient noise as a single measure of environmental performance. However, it is also true that this is a common method in the literature. For example, in operating rooms mean noise levels ranged between 51dB and 79dB in a recent review of 16 studies.²⁸ The World Health Organization (WHO) advises noise levels to remain well below this level in clinical environments. This study showed that eCPR events were relatively noisy compared to most hospital environments.^{28 51} Australian safe work standards (https://www.safeworkaustralia.gov.au) state that exposure to noise levels above 85dB should lead employers to consider mitigation strategies.

Studies from North American contexts by Grantcharov and co-authors examining operating room ‘black box’ give credence use of various ‘sabermetrics’ including noise.⁵⁴ We believe applying this work to eCPR teams is relevant and further work could explore the use of these sabermetric variables for research and team feedback. Noise has been used as an ‘early warning’ sign for team strain. For example, monitoring the noise levels as a black box ‘warning’ allows allocation of team attention to the alleviating strain, addressing the patient or applying ‘crowd control’ measures. Such ‘noise alarms’ are often used outside healthcare settings and have been used in critical care environments.⁵⁵

Systems, Workflow and Crowd Control

Improving workflow for eCPR events requires an incremental process of design, testing, and refinement. This work on ‘systems’ should occur in the clinical environment and can appraise many factors including equipment and operational processes.⁸ We report images of the eCPR workflow in Fig. 3. We have previously described our work on this topic in detail.¹⁴ Approaches described in the wider literature include strategic arrangement of space around the patient. This would ensure an efficient location, labelling of equipment trolleys and operationalizing the most important eCPR processes as a team checklist. In planning the eCPR system we frequently used in situ simulation (low fidelity) and tabletop exercises to identify potential problems with workflow. Use of tabletop simulation aligns with work described above describing ‘mental rehearsal’ as a strategy to reduce stress as well as the emerging literature on Visually enhanced Mental Simulation (VEMS).^{46 56 57} Improvements following simulation exercises included ergonomic positioning of the ultrasound(s) required for eCPR allowing execution to be the focus rather than repositioning and lanyard ‘ticket’ system with cordon tape (Fig. 3; Fig. 4) to prevent observers or superfluous team members crowding round the bed.¹⁴

A further example of innovation was the use of a crowd control leader (CCL). This role (taken by a senior doctor or nurse) is distinct from the medical team leader with the role outlined in Fig. 4. This individual considers logistics, safety and communication. Further, the CCL promotes teamwork and ‘relational coordination’ during eCPR.⁴⁹ Introducing a CCL created a buffer that redistributed cognitive load and protected the team leader’s bandwidth.

The presence of this role in all the cases may explain the performance by teams despite stress and noise. A CCL, who is often experienced and not focussed on the clinical demands, can also increase the likelihood of the ‘small details’ being remembered such as patient positioning, use of checklists and clinical handovers to team members who arrive late. Late team member arrivals were common (Table 2). In stressful cases a CCL could have a role in attending to more than just crowd control. ⁸ Their actions could include re-orientating team members to key goals, actively mitigating stress at a team level, supporting the other team leaders and consideration of organising a clinical debriefing.^{44 58}

Limitations

An important strength of this study was the structured use of simulation and pilot trauma cases to iteratively refine our data-collection strategy before progressing to real-world eCPR events. The use of a recognised conceptual framework as a lens to interpret the data is also a strength. The study is limited by a small sample size, with only six eCPR cases and 18 participating healthcare providers, which may restrict the generalisability of findings. Most observations were conducted at a single centre due to logistical constraints, and 11 cases were missed because researchers were unavailable. The distribution of missed cases across shifts or weekdays was not recorded, introducing the potential for systematic bias.

Observational bias, including the Hawthorne effect, is also relevant, as team members were aware they were being observed. We attempted to reduce this effect by prospective planning and positioning observers in an off-centre (less visible) position. Missing data were substantial, with complete physiological stress measures (heart rate and salivary cortisol) available for only 7 of 18 staff, further limiting conclusions. Finally, the absence of blinding, control groups, and limited follow-up precludes stronger causal inferences regarding the impact of stress or teamwork during eCPR.

Future Directions

We plan to implement a continuous quality improvement process across the two eCPR sites with video analysis and a series of educational simulations. Future work could focus on eCPR teamwork, workflows and approaches across multiple centres to improve generalisability and potentially compare approaches to teamwork. From an educational point of view, low-cost scalable programmes such as VEMS should be further evaluated for their value in training eCPR teams and for HALO events.

This study demonstrates that eCPR teams can achieve consistently high performance even under considerable physiological stress and in crowded environments. Using Reid’s domains, we observed that individual stress, teamwork behaviours, environmental load and system design interacted dynamically during eCPR delivery. Clear role allocation, proactive crowd control and regular team summaries emerged as practical system features that were likely to mitigate the effects of noise and stress. These findings suggest that targeted education and training in human factors and non-technical skills may be a key driver for improving eCPR performance

Competing Interests:

No authors have a conflict of interest to declare

Author Contribution

ARC and MO conceived and designed the study. Data were acquired by ARC, AS, and staff from the Westmead Hospital Simulation Centre. Data analysis and interpretation were conducted by ARC, JC, and HC. All authors contributed to drafting the manuscript, endorsed the final draft and provided feedback on revisions. Charitable funding for the project was secured by ARC.

Data Availability

All data relevant to the study are included in the article or uploaded as supplementary information. No large language models were used to develop the manuscript.

Williams ST, Sykes MC, Boon Lim P, Salciccioli JD. The 2015 advanced life support guidelines: a summary and evidence for the updates. Emerg Med J. 2016;33(5):357–60.
Perman SM, Elmer J, Maciel CB, et al. 2023 American Heart Association focused update on adult advanced cardiovascular life support: an update to the American Heart Association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation. 2024;149(5):e254–73.
Chandru P, Barua S, Dennis M, Eslick A, Kruit N, Coggins A. A prospective 3-year review of out of hospital cardiac arrest presentations to a university-affiliated tertiary centre. Heart Lung Circ. 2019;28(Suppl 1):S198.
Zeymer U, Freund A, Hochadel M, et al. Venoarterial extracorporeal membrane oxygenation in patients with infarct-related cardiogenic shock: an individual patient data meta-analysis of randomised trials. Lancet. 2023;402(10410):1338–46.
Elliott AM, van Diepen S, Hollenberg SM, Bernard S. Extracorporeal cardiopulmonary resuscitation: life-saving or resource wasting? US Cardiol. 2024;18:e12.
Ostadal P, Rokyta R, Karasek J, et al. Extracorporeal membrane oxygenation in the therapy of cardiogenic shock: results of the ECMO-CS randomized clinical trial. Circulation. 2023;147(6):454–64.
Yannopoulos D, Bartos J, Raveendran G, et al. Advanced reperfusion strategies for patients with out-of-hospital cardiac arrest and refractory ventricular fibrillation (ARREST): a phase 2, single centre, open-label, randomised controlled trial. Lancet. 2020;396(10265):1807–16.
Hicks C, Petrosoniak A. The human factor: optimizing trauma team performance in dynamic clinical environments. Emerg Med Clin North Am. 2018;36(1):1–17.
Yule S, Flin R, Paterson-Brown S, Maran N. Non-technical skills for surgeons in the operating room: a review of the literature. Surgery. 2006;139(2):140–9.
Howie EE, Ambler O, Gunn EGM, et al. Surgical sabermetrics: a scoping review of technology-enhanced assessment of nontechnical skills in the operating room. Ann Surg. 2024;279(6):973–84.
Leblanc VR, Regehr C, Tavares W, Scott AK, Macdonald R, King K. The impact of stress on paramedic performance during simulated critical events. Prehosp Disaster Med. 2012;27(4):369–74.
Yule S, Janda A, Likosky DS. Surgical sabermetrics: applying athletics data science to enhance operative performance. Ann Surg Open. 2021;2(2):e054.
Lingard L. Joining a conversation: the problem/gap/hook heuristic. Perspect Med Educ. 2015;4(5):252–3.
Oliver M, Coggins A, Kruit N, et al. Implementing enhanced extracorporeal membrane oxygenation for CPR (ECPR) in the emergency department. Int J Emerg Med. 2024;17(1):71.
Fatovich DM, Carey S, Iliff J, Jowitt T, Weber DG, Vance JS. Integrating Research Practice Into Resuscitation Simulation Training Improves Recruitment Into Complex Clinical Trials. Emerg Med Australas. 2025;37(3):e70064.
Knudson KA, Funk M, Redeker NS, et al. An unbelievable ordeal: the experiences of adult survivors treated with extracorporeal membrane oxygenation. Aust Crit Care. 2022;35(4):391–401.
Rupprecht L, Lunz D, Philipp A, Lubnow M, Schmid C. Pitfalls in percutaneous ECMO cannulation. Heart Lung Vessel. 2015;7(4):320–6.
Reid C, Brindley P, Hicks C, Carley S, Richmond C, Lauria M, Weingart S. Zero point survey: a multidisciplinary idea to STEP UP resuscitation effectiveness. Clin Exp Emerg Med. 2018;5(3):139–43.
Valentin B, Grottke O, Skorning M, et al. Cortisol and alpha-amylase as stress response indicators during pre-hospital emergency medicine training with repetitive high-fidelity simulation and scenarios with standardized patients. Scand J Trauma Resusc Emerg Med. 2015;23:31.
Kim HG, Cheon EJ, Bai DS, Lee YH, Koo BH. Stress and heart rate variability: a meta-analysis and review of the literature. Psychiatry Investig. 2018;15(3):235–45.
Spielberger C, Gorsuch RL, Lushene R, Vagg PR, Jacobs GA. Manual for the State-Trait Anxiety Inventory. Palo Alto: Consulting Psychologists; 1983. Accessed 2025 Feb 5.
Malec JF, Torsher LC, Dunn WF, et al. The Mayo high performance teamwork scale: reliability and validity for evaluating key crew resource management skills. Simul Healthc. 2007 Spring;2(1):4–10.
Weller J, Frengley R, Torrie J, et al. Evaluation of an instrument to measure teamwork in multidisciplinary critical care teams. BMJ Qual Saf. 2011;20(3):216–22.
Huyan J, Ramkissoon C, Laka M, Gaskin S. Assessing the usefulness of mobile apps for noise management in occupational health and safety: quantitative measurement and expert elicitation study. JMIR Mhealth Uhealth. 2023;11:e46846.
Hall A, Kawai K, Graber K, et al. Acoustic analysis of surgeons' voices to assess change in the stress response during surgical in situ simulation. BMJ Simul Technol Enhanc Learn. 2021;7(6):471–7.
Peisl S, Sánchez-Taltavull D, Guillen-Ramirez H et al. m. Noise in the operating room coincides with surgical difficulty. BJS Open. 2024;8(5):zrae098.
Keller S, Tschan F, Beldi G, Kurmann A, Candinas D, Semmer NK. Noise peaks influence communication in the operating room: an observational study. Ergonomics. 2016;59(12):1541–52.
Mcleod R, Myint-Wilks L, Davies SE, Elhassan HA. The impact of noise in the operating theatre: a review of the evidence. Ann R Coll Surg Engl. 2021;103(2):83–7.
Au J, Hamilton S, Webb A. Decibels in the operating theatre: a study of noise levels during surgical procedures. ANZ J Surg. 2024;94(10):1841–5.
Sorensen D, Cristancho S, Soh M, Varpio L. Team stress and its impact on interprofessional teams: a narrative review. Teach Learn Med. 2024 Apr-May;36(2):163–73.
Reid C, Brindley P, Hicks C, et al. Zero point survey: a multidisciplinary idea to STEP UP resuscitation effectiveness. Clin Exp Emerg Med. 2018;5(3):139–43.
Tjønnås MS, Muller S, Våpenstad C, Tjønnås J, Ose SO, Das A, Sandsund M. Stress responses in surgical trainees during simulation-based training courses in laparoscopy. BMC Med Educ. 2024;24(1):407.
Lizotte MH, Janvier A, Latraverse V, et al. The impact of neonatal simulations on trainees' stress and performance: a parallel-group randomized trial. Pediatr Crit Care Med. 2017;18(5):434–44.
Harvey A, Bandiera G, Nathens AB, LeBlanc VR. Impact of stress on resident performance in simulated trauma scenarios. J Trauma Acute Care Surg. 2012;72(2):497–503.
LeBlanc VR. The effects of acute stress on performance: implications for health professions education. Acad Med. 2009;84(10 Suppl):S25–33.
Szulewski A. A doctor in the headlights. AEM Educ Train. 2019;3(2):200–1.
Driskell JE, Salas E, Johnston J. Does stress lead to a loss of team perspective? Group Dyn. 1999;3(4):291–302.
Meichenbaum D. Stress inoculation training: a preventative and treatment approach. In: Lehrer PM, Woolfolk RL, Sime WS, editors. Principles and Practice of Stress Management. 3rd ed. New York: Guilford Press; 2007. pp. 497–518.
Saunders T, Driskell JE, Johnston JH, et al. The effect of stress inoculation training on anxiety and performance. J Occup Health Psychol. 1996;1(2):170–86.
Brazil V, Orr R, Canetti EFD, Isaacson W, Stevenson N, Purdy E. Exploring participant experience to optimize the design and delivery of stress exposure simulations in emergency medicine. AEM Educ Train. 2023;7(2):e10852.
Carnevali L, Bignami E, Gambetta S, Barbetti M, Procopio M, Freyrie A, Carbognani P, Ampollini L, Sgoifo A. Cardiac autonomic and cortisol stress responses to real operations in surgeons: relationship with individual psychobiological characteristics and experience. Biopsychosoc Med. 2023;17(1):5.
Folkman S, Lazarus RS. The relationship between coping and emotion: implications for theory and research. Soc Sci Med. 1988;26(3):309–17.
Murphy M, McCloughen A, Curtis K. The impact of simulated multidisciplinary trauma team training on team performance: a qualitative study. Australas Emerg Care. 2019;22(1):1–7.
Sawyer T, Eppich W, Brett-Fleegler M, Grant V, Cheng A. More than one way to debrief: a critical review of healthcare simulation debriefing methods. Simul Healthc. 2016;11(3):209–17.
Iattoni M, Ormazabal M, Luvini G, Uccella L. Effect of structured briefing prior to patient arrival on interprofessional communication and collaboration in the trauma team. Open Access Emerg Med. 2022;14:385–93.
Lorello GR, Hicks CM, Ahmed SA, Unger Z, Chandra D, Hayter MA. Mental practice: a simple tool to enhance team-based trauma resuscitation. CJEM. 2016;18(2):136–42.
Stocker M, Menadue L, Kakat S, et al. Reliability of team-based self-monitoring in critical events: a pilot study. BMC Emerg Med. 2013;13:22.
Lee JY, Szulewski A, Young JQ, Donkers J, Jarodzka H, van Merriënboer JJG. The medical pause: importance, processes and training. Med Educ. 2021;55(10):1152–60.
Bogdanovic J, Perry J, Guggenheim M, et al. Adaptive coordination in surgical teams: an interview study. BMC Health Serv Res. 2015;15:128.
Gamborg ML, Mylopoulos M, Mehlsen M, Paltved C, Musaeus P. Exploring adaptive expertise in residency: the (missed) opportunity of uncertainty. Adv Health Sci Educ Theory Pract. 2024;29(2):389–424.
Short AE, Short KT, Holdgate A, Ahern N, Morris J. Noise levels in an Australian emergency department. Australas Emerg Nurs J. 2011;14(1):26–31.
Kennedy-Metz LR, Wolfe HL, Dias R, Yule SJ, Zenati MA. Surgery task load index in cardiac surgery: measuring cognitive load among teams. Surg Innov. 2020;27(6):673–80.
Purdy EI, McLean D, Alexander C, Scott M, Donohue A, Campbell D, Wullschleger M, Berkowitz G, Winearls J, Henry D, Brazil V. Doing our work better, together: a relationship-based approach to defining the quality improvement agenda in trauma care. BMJ Open Qual. 2020;9(1):e000749.
Incze T, Pinkney SJ, Li C, et al. Using the operating room black box to assess surgical team member adaptation under uncertainty: an observational study. Ann Surg. 2024;280(1):75–81.
Guisasola-Rabes M, Solà-Enriquez B, Vélez-Pereira AM, de Nadal M. Effectiveness of a visual noise warning system on noise levels in a surgical ICU: a quality improvement programme. Eur J Anaesthesiol. 2019;36(11):857–62.
Brazil VA, Speirs C, Scott C, Schweitzer J, Purdy E. Recommendations for the design and delivery of Visually Enhanced Mental Simulation: insights from participants and facilitators. J Healthc Simul. 2025;1–11.
Demir S, Tunçbilek Z, Alinier G. The effectiveness of online visually enhanced mental simulation in developing casualty triage and management skills of paramedic program students: a quasi-experimental research study. Int Emerg Nurs. 2023;67:101262.
Butchibabu A, Sparano-Huiban C, Sonenberg L, Shah J. Implicit coordination strategies for effective team communication. Hum Factors. 2016;58(4):595–610.

No competing interests reported.

Download PDF

Editorial decision: Revision requested
29 Dec, 2025
Reviews received at journal
28 Dec, 2025
Reviewers agreed at journal
23 Dec, 2025
Reviews received at journal
22 Dec, 2025
Reviewers agreed at journal
22 Dec, 2025
Reviewers agreed at journal
21 Dec, 2025
Reviewers agreed at journal
21 Dec, 2025
Reviewers agreed at journal
18 Dec, 2025
Reviewers agreed at journal
16 Dec, 2025
Reviewers invited by journal
16 Dec, 2025
Editor invited by journal
01 Dec, 2025
Editor assigned by journal
27 Nov, 2025
Submission checks completed at journal
27 Nov, 2025
First submitted to journal
26 Nov, 2025

You are reading this latest preprint version

Human Factors in Emergency Department eCPR: A Prospective Observational Study of Teamwork, Stress and Environmental Demands

Status:

Version 1

Abstract

BACKGROUND

METHODS

RESULTS

CONCLUSION

Figures

BACKGROUND

METHODS

Study Design

Study Setting

eCPR Model of Care

Translational Simulation Pilot (February 2022 - March 2022)

eCPR Cases (March 2022 - December 2024)

Case Sampling

Outcome Measures

Analysis Plan and Bias

RESULTS

Individuals and Stress

Teamwork

Environmental Factors and Noise

DISCUSSION

Individuals and Stress

Teamwork

Environmental Factors and Noise

Systems, Workflow and Crowd Control

Limitations

Future Directions

CONCLUSION

Declarations

Competing Interests:

Author Contribution

Data Availability

References

Additional Declarations

Supplementary Files

Status:

Version 1