Factors influencing workload and stress during resuscitation – A scoping review

Abstract

Aim

This scoping review aimed to identify potential variables influencing healthcare provider’s perceived workload or stress when performing resuscitation on patients in cardiac arrest.

Methods

We searched Medline, EMBASE, PsycINFO, Cochrane, and Allied Health Literature (CINAHL) to identify studies published prior to February 1, 2024. We used a PECO format for this review: the population were healthcare providers performing resuscitation during simulated or real cardiac arrest; the exposure was the presence of any factor that could impact perceived workload or stress; and the comparator was the absence of any specific factor. Outcome variables, including self-reported questionnaires, objective and subjective measures, and any variables identified to have impact on workload and/or stress were extracted.

Results

Of the initially identified 10,165 studies, 24 studies (20 RCTs, 2 quasi-experimental studies and 2 observational studies) were ultimately included. Among them, a wide variety of factors influencing perceived stress or workload were identified. High heterogeneity among studies was observed. We categorized factors into the following entities: (1) team composition and roles; (2) telemedicine; (3) workflow; (4) tools; (5) cognitive aids; (6) presence of friends and family, and (7) provider experience and exposure, representing the modifiable factors for future interventions.

Conclusion

This scoping review provides an overview of factors influencing workload and stress during real and simulated cardiac arrest resuscitation. These findings highlight the need for targeted strategies to effectively manage workload and stress during resuscitation.

Introduction

Cardiopulmonary resuscitation (CPR) is a challenging task and requires dedicated and focused attention of a healthcare professional team. The effective delivery of high-quality CPR for patients in cardiac arrest is crucial for their survival and yet a demanding responsibility for resuscitation team members. It requires precise and accurate delivery of guideline-driven protocols while simultaneously working cohesively as a team.¹ A few studies have shown that provider workload, which includes mental and physical workload, can impact human performance.2, 3 Similarly, stress experienced by healthcare providers during resuscitation potentially affect individual or team performance adversely.4, 5 Understanding factors influencing workload or stress experienced by resuscitation team members during CPR might inform future educational and implementation strategies aiming to improve CPR-performance and team-based clinical performance. Despite emerging literature investigating workload and stress during resuscitation,⁶ it is unclear what factors influence workload or stress experienced by individual resuscitation team members. Once influencing factors are identified, possible strategies alleviating stress or workload might be developed to improve performance and patient outcomes. Thus, our review aimed to identify and describe potential factors influencing healthcare providers’ perceived workload or stress when performing resuscitation on patients in cardiac arrest in real or in simulated settings.

Methods

This scoping review was conducted by the Education, Implementation and Teams (EIT) Task Force of the International Liaison Committee on Resuscitation (ILCOR), as part of informing International Consensus on Cardiopulmonary Resuscitation and Emergency Cardiovascular Care Science Treatment Recommendations (CoSTR). The study protocol was available on the CoSTR website (EIT 6401 TF ScR).

Eligibility criteria

Our research question: “Amongst healthcare providers, what variables influence (i.e., increase or decrease) providers’ workload and/or stress during cardiac arrest, in both real-world and simulated scenarios” triggered the PECOST framework for this scoping review:

• •Population: healthcare providers performing resuscitation on patients in cardiac arrest or on manikins in a simulated setting.
• •Exposure: presence of any factor that would possibly impact the healthcare provider’s perceived workload or stress.
• •Comparator: absence of any specific factor.
• •Outcome variables: healthcare provider’s perceived workload or stress, including self-reported questionnaires, objective and subjective measures.
• •Study designs: Randomized controlled trials (RCTs) and non-randomized studies (non-RCTs, interrupted time series, controlled before-and-after studies, cohort studies), unpublished studies (e.g., conference abstracts, trial protocols), letters, editorials, comments, case reports. All publications in any language were included if there was an English abstract.
• •Timeframe: All years from inception to February 1, 2024.

Studies that did not report on workload or stress during cardiac arrest, only reported on workload or stress during non-cardiac arrest resuscitation scenarios (e.g., trauma resuscitation), described educational or other interventions that aimed at reducing workload or stress during cardiac arrest, and studies that examined the effect of patient status (survival or mortality) on the stress and/or workload experienced by resuscitation teams were excluded.

The rationale for the inclusion and exclusion criteria were to isolate factors associated with workload and/or stress. We excluded experimental studies that evaluated educational interventions aiming at reducing or modifying provider workload and/or stress, as we focused on factors inherent to typical clinical environments and their correlation with provider workload or stress.

Definitions

Workload and stress

Workload refers to cognitive, physical, and emotional demands placed on individuals while performing tasks.⁷ The National Aeronautics and Space Administration Task Load Index (NASA-TLX) is a validated tool which measures workload based on ratings on six subscales: mental demand, physical demand, temporal demand, performance, effort, and frustration level.⁸ Stress refers to the psychological and physiological response of individuals facing demanding situations or challenges that exceed their perceived ability to cope.⁹ Stress-related indicators included physiologic markers (e.g. salivary cortisol and amylase, heart rate, heart rate variability, blood pressure). Subjective scales include the State-Trait Anxiety Inventory (STAI).¹⁰

Information sources and search strategy

The search strategy included keywords, MeSH terms and Embase exploded terms related to our study question and the PECO framework, developed by a professional information specialist. The detailed description of the search strategy is provided in the Appendix 1. We searched Medline, EMBASE, PsycINFO, Cochrane, and Allied Health Literature (CINAHL) initially in Sep 2021. The search was updated in Feb 2023.

Study selection, data extraction and synthesis of results

Titles and abstracts were independently screened by two reviewers (CHL, CWY). Disagreements were resolved via discussion to reach consensus with full text analysis.

Data was extracted from the final list of studies by two authors (CHL, CWY), differences were resolved by discussion. Data extracted included author, publication year, country, study design, population, sample size, intervention and/or comparison (if applicable), outcome measures, and results.

We reported:

• (1)self-reported questionnaires such as NASA Task Load Index (NASA-TLX), Subjective Workload Assessment Technique (SWAT).¹¹
• (2)objective and subjective measures of stress, including physiologic measures (heart rate, cortisol, salivary amylase, blood pressure), validated scales including State Trait Anxiety Inventory (STAI), ¹² visual analogue scale for anxiety (VAS-A).¹³
• (3)other variables identified having possible impact on workload and/or stress, like family presence during resuscitation, use of mechanical chest compression devices, resuscitation team size, change of workflow, use of cognitive aids.

After data extraction, we employed a narrative synthesis approach, which involved the systematic collation and summary of the data from the included studies. Consistent with the methodological framework of scoping reviews, no meta-analysis was conducted. The review conforms to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines specifically extended for Scoping Reviews.¹⁴

Results

We identified 10,165 studies; 7555 titles and abstracts were screened after removal of duplicates. We excluded 7464 studies (reasons in Fig. 1), leaving 91 articles for full-text review. After full-text review, 24 articles were included for the final data extraction.6, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37

Fig. 1.

PRISMA flowchart of the included studies.

Study characteristics

The included 24 studies are heterogeneous in study design. Twenty are RCTs, 15, 16, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 32, 33, 34, 35, 36, 37 two are experimental designs without randomization,17, 31 and two are observational studies.6, 30 All but two studies6, 30 were done in simulation-based settings, with seven using pediatric scenarios,15, 19, 25, 26, 31, 32, 38 all other used adult clinical scenarios.

Besides extracting variables influencing individual healthcare providers' workload and stress during resuscitation, we also included measurement tools for workload and stress. The NASA Task Load Index 15, 16, 18, 19, 21, 22, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37 was the primary method for measuring subjective workload. Furthermore, the State-Trait Anxiety Inventory,17, 24 Visual Analogue Scale (VAS),17, 24, 39 and structured survey questions 6, 20, 29 were utilized to assess stress. Physiologic stress markers included salivary cortisol and alpha-amylase levels,³⁰ heart rate and blood pressure.¹⁷

Variables that may have had an influence on perceived stress or workload were extracted and categorized into the following separate entities:

• (1)team composition and roles (like designated nursing team leader, CPR coaches or comparison of workloads between resuscitation team leaders and members);
• (2)telemedicine involving teams supervised or led remotely;
• (3)workflows such as prioritization of CPR automation or task-focusing techniques;
• (4)tools including CPR feedback devices;
• (5)cognitive aids (eg. Manuals, smart apps);
• (6)presence of friends and family during resuscitation, and.
• (7)provider experience and exposure.

Categorization was guided by the intrinsic nature and underlying similarities of each factor during iterative process of discussions, ensuring a coherent grouping that reflects the multifaceted dynamics modifiable factors in resuscitation scenarios.

• (1)Team composition and roles

One simulation RCT suggested that the presence of a nursing team leader can significantly alleviate the medical team leader’s workload during emergency resuscitation.³⁴ The nursing team leader actively took over tasks like timing of 2-min cycles, rhythm checks, and defibrillation, monitoring CPR quality, and prompting drug administration. Charles et al. measured effects of team size on healthcare providers’ workload. Teams of two participants had significantly higher NASA-TLX scores than teams of three participants.²⁶ The presence of a CPR-coach, which is a designated person responsible for monitoring chest compression rate, depth, and compression fractions, and providing feedback to optimize CPR quality, decreases mental workload but increases physical workload among CPR providers,²⁸ but did not impact workload for resuscitation team leaders.22, 28 Additional designated hands-off recorder/time coach, on the other hand, increased team leader’s workload.³⁷ In another prospective cohort study of real pediatric resuscitations, team leaders reported higher mental load, whereas chest compressors reported higher physical workload.⁶ (Table 1a).

• (2)Telemedicine

Table 1.

Study details with influencing variables that might influence perceived stress or workload: a. team composition and roles; b. telemedicine; c. workflows; d. tools; e. cognitive aids; f. presence of friends and family during resuscitation; g. provider experience and exposure. ALS, advanced life support; CA: Cognitive aids; CI, confidence interval; CPR, cardiopulmonary resuscitation; EMT, emergency medical technician; IQR, interquartile range; PICU, pediatric intensive care unit; SD, standard deviation; CPR coaches: designated roles for monitoring and providing feedback on the quality of CPR, including compression rate, depth, and interruptions.

Article (Country)	Design and setting	Intervention/Exposure	Control	Team Composition	Participants (n)	Workload/Stress Measurement	Key Findings
a. Studies with team composition and roles as factors influencing healthcare provider’s workload or stress
2020 Tofil (Canada)29	Nested Randomized Controlled Trial; Pediatric Simulation	Addition of a CPR coach	No CPR coach	Two CPR providers, a team leader, an airway provider and a CPR coach or bedside provider	Pediatric healthcare providers	NASA- TLX scores	1. Adding a CPR coach increases physical workload and decreases mental workload of cardiopulmonary resuscitation providers. 2. There was no change in team leader’s workload between the groups.
2020 Badke (US)22	Randomized Controlled Trial; Adult Simulation	First assistant assuming the role of CPR Coach	No designation of first assistant	Resuscitation team size was not standardized, as it was dependent on the availability of PICU providers at the time of the simulation.	Critical care, pediatric, cardiac, or emergency medicine physicians, nurse practitioners, nurses, pharmacists, respiratory therapists, and medical students	NASA- TLX scores	There were no significant differences in team leader workload for scenarios that included a CPR Coach versus those without a CPR Coach.
2020 Roitsch (US)27	Randomized 2 × 2 factorial design study; Pediatric Simulation	Intervention 1: Team size Intervention 2: use of a tablet-based decision support tool		Team of 2 or 3, with one advanced provider per team	Healthcare providers (neonatologist, neonatology fellows, neonatal nurse practitioners, registered nurses, and respiratory therapists)	NASA-TLX scores	1. The NASA TLX scores were significantly increased within teams of 2 compared with 3 (mean difference = 6.2, 95 % CI = 0.4–12.0, P = 0.036, effect size = 0.42)
2021 Pallas (Australia)35	Randomized Controlled Trial; Adult Simulation	Designation of a nursing team leader	Usual care	Three doctors, two advanced life support (ALS) trained nurses, one additional non-ALS trained staff member, one investigator and one simulation confederate	Available ward staff (20 simulations, n = 120)	NASA-TLX scores	1. Similar total NASA-TLX in the Intervention and Control groups (p = 0.28) 2. Medical team leaders in the Intervention group had a significantly lower NASA-TLX (238.4, 95 % CI 192.0 to 284.7) compared to control group (306.3, 95 % CI 254.9 to 357.6; p = 0.02). 3. No statistically significant difference of NASA-TLX was observed between the nursing team leader of the intervention group (mean 223, 95 % CI: 189.3 to 256.7) and the senior control nurse (mean 255.5, 95 % CI 195.5 to 315.5; p = 0.15)
2023 Roman (US)6	Prospective cohort study; Pediatric resuscitation	team leaders	team members	actual resuscitation team in pediatric resuscitation	61 participants at 15 different resuscitation events including team leader data from 8 events	NASA-TLX, the stress numerical rating scale-11 (SNRS-11)	There was no difference in overall workload between team leaders and other team members (p = 0.601). There was higher mental demand in team leaders compared to other team members (p = 0.025).
Neveln 202338 (US)	Randomized 2 × 2 factorial design study; Pediatric Simulation	Intervention 1: Additional hands-on team member 2: Additional hands-off recorder/time coach	Baseline team of four	Team leader(neonatologist), pediatric resident, respiratory therapist, and a nurse	Healthcare providers from study site, 64 teams	NASA TLX scores	1. Significant higher team leader workload in teams with a recorder (p = 0.047). 2. No difference in team leader NASA TLX scores between teams with and without an additional team member
b. Studies with telemedicine as factors influencing healthcare provider’s workload
2019 Butler (US)19	Randomized Controlled Trial; Pediatric Simulation	Senior doctor act as remote team leaders in a separate control room (telemedicine)	Senior doctor present on site	Two doctors (a senior and a junior resident) and two standardized confederate nurses	Emergency Medicine residents (n = 20 teams)	NASA TLX scores	1. The telemedicine group had a higher workload compared to the usual care group (56 vs. 48, p = 0.020) 2. Across the seven sub-domains of the NASA TLX tool, there was a significantly higher mental demand in the telemedicine group.
2020 Couturier (US)37	Randomized Controlled Trial; Pediatric Simulation	Pediatric trained provider as remote assistant	No remote assistant	Participant and a confederate nurse	Emergency Medicine resident (n = 12)	NASA TLX scores	1. No significant difference was in overall workload 2. Frustration subscore components was statistically significantly greater in the control group 14 vs. 8 (p < 0.001).
2020 Gross (US)33	Randomized Controlled Trial; Pediatric Simulation	Proactive leader at a remote site (telemed leader)	Remote consultant, provide guidance on request (telemed consultant)	Participant and a confederate nurse	Emergency Medicine resident, Physician Assistants	NASA TLX scores	1. No significant difference between the two groups in the overall workload (p = 0.222). 2. When compared to the telemed leader group, the teleconsultant group experienced a higher level of mental demand (mean mental demand: telemed leader 14.1 vs. teleconsultant 17.0 out of 21, p < 0.05) and a higher level of frustration (telemed leader 7.9 vs. teleconsultant 14.7 out of 21, p < 0.05).
c. Studies with different workflows as factors influencing healthcare provider’s workload or stress
2018 Asselin (US)34	Randomized Controlled Trial; Adult Simulation	Goal-directed, and automation-assisted approach resuscitation	Standard state protocols and equipment	Two-provider team (1 EMT-B and 1 EMT-I/C/P).	Emergency medical technicians (EMTs) with regional and/or national licenses at the Basic (B), Intermediate (I), Cardiac (C), or Paramedic (P) levels	Heart rate, salivary amylase, Borg Rating of Perceived Exertion scale, NASA TLX scores	1. Reduced physical exertion and lower perceived workloads in automation-assisted teams
2013 Hunziker (Switzerland)23	Randomized Controlled Trial; Adult Simulation	10-minute instruction with two task-focusing questions (“what’s the patient’s condition?”; “what immediate action is needed?”) when feeling overwhelmed during simulated resuscitation	Usual care	Participant and confederate nurse	4th year medical students	perceived levels of stress measured on a Likert scale ranging from 1–20	1. Significantly smaller amounts of perceived stress and overload compared to the control group (difference of mean perceived stress: −0.6 (95 % CI-1.3, −0.1), p = 0.04)
d. Studies with automation tools or equipment failures as factors influencing healthcare provider’s workload or stress
Workload
2022 Wagner (Austria)15	Randomized Controlled Trial; Pediatric Simulation	Feedback device for chest compression and ventilation	No feedback device	participant did chest compression and ventilations on their own	Medical students, fellows, nurses, and consultants from Neonatal Intensive Care Unit (n = 40)	NASA-TLX scores	1.The average workload for chest compression task was 37 % in the feedback group, which was 3.5 % higher than no-feedback group (P = 0.02) 2.The average workload for ventilation task in the feedback group was 36 %, which was 8 % higher than no-feedback group (P < 0.001)
2018 Brown (Canada)25	Randomized Controlled Trial; Pediatric Simulation	Real-time visual CPR feedback device	Usual care	1 leader and 2 CPR providers	residents, fellows, physicians, nurses and nurse practitioner (n = 108 teams)	NASA-TLX scores	1. CPR providers reported comparatively higher physical workload. CPR providers reported significantly higher average workload (control 58.5 vs. feedback 62.3; p = 0.035) with real-time feedback provided compared to the group without feedback. 2. For teams with real-time feedback, there was a significant difference in average workload between team leader (TL) and CPR providers (CPR-P) [TL 56.1vs. CPR-P 62.3, p = 0.001]. 3. Team leaders had significantly higher mental demand (p < 0.001), but significantly lower physical demand (p < 0.001) and effort (p = 0.032) workloads compared with CPR providers
Stress
2020 Ontrup (Germany)30	Randomized Controlled Trial; Adult Simulation	Equipment failure (defective defibrillator)	No equipment failure	Participant and two confederate nurses	Medical students (human medicine) in their 7th to 9th semester	Salivary cortisol and amylase; Five-item questionnaires	1. Participants of both groups showed increased biological stress-levels, independent of group allocation. 2. Paradoxically, participants who encountered the equipment failure subjectively reported less stress.
e. Studies with cognitive aids and smart apps as factors influencing healthcare provider’s workload or stress
Workload
2020 Roitsch (US)27	Randomized 2 × 2 factorial design study; Pediatric Simulation	Intervention 1: Team size Intervention 2: use of a tablet-based decision support tools (DST) *Scenario A: hypoxemic, and bradycardic full-term newborn requiring intubation. *Scenario B: Similar to A, but requires CPR		Team of 2 or 3, with one advanced provider per team	Healthcare providers (neonatologist, neonatology fellows, neonatal nurse practitioners, registered nurses, and respiratory therapists)	NASA-TLX scores	1. Teams that used the DST scored workload significantly higher during scenario A (−DST − +DST mean difference =  − 7.5, 95 % CI =  − 14.2 to − 0.9, P = 0.027, effect size = 0.40) but in scenario B workload in teams using the DST was not significantly different (−DST − +DST mean difference = 5.1, 95 % CI =  − 1.6 to 11.9, P = 0.135). 2. Individual averages of NASA TLX scores of scenarios A and B was not significantly associated with a change in NASA TLX scores for teams using the DST compared with memory alone (−DST − +DST mean difference =  − 1.0, 95 % CI =  − 6.7 to 4.7, P = 0.721)
2020 Corazza (Italy)32	Non-randomized controlled study; Pediatric Simulation	PediAppRREST (audiovisual interactive app as cognitive aid)	Without the app	Three pediatric residents	48 pediatric residents divided into teams of 3; Five teams managed the case following usual care (control group), whereas 11 teams (intervention group) conducted the scenario using the support of the PediAppRREST app	NASA TLX scores	1. Using the app is not associated with increased team leaders’ workload; Team leaders’ perceived workload was comparable between the 2 groups; median NASA TLX score was 67.5 (IQR 65.0–81.7) for the control group and 66.7 (IQR 54.2–76.7) for the intervention group (P = 0.57).
2023 Corazza36 (Italy)	3-Group parallel randomized trial, Pediatric Simulation	PediAppRREST app; AHA PALS pocket reference card(card)	No cognitive aid(null)	Three members, with at least 1 PALS-certified resident for the team leader role	Residents in pediatrics, emergency medicine, and anesthesiology	NASA-TLX scores	1. No significant difference in team leader’s total workload 2. Mental demand subscores were significantly lower in the app group: mean 70.0, SD 20.9(app), mean 81.3, SD 14.7(card), mean 80.5, SD 13.4(null)
2021 Grundgeiger (Germany)16	Randomized Control Trial; Adult Simulation	Cognitive aid application (CA App) group	No application (No App) group	One qualified emergency physician as team leader and one qualified nurse	Prehospital emergency medicine physicians and acute care nurses (n = 67 teams)	NASA-TLX scores	1. For the physicians, the analysis of the NASA TLX scores indicated significantly lower mental demand, physical demand, and effort for the CA App group than the No App group 2. For the nurses, the analysis of the NASA TLX scores indicated significantly lower mental demand for the CA App group compared to the No App group.
Stress
2021 Lacour (Switzerland)24	Nested Randomized Controlled Trial; Pediatric Simulation	PedAMINES Smart App for dosing calculation	Usual care	Participant alone	Registered paramedics (n = 150)	State-Trait Anxiety Inventory, Self-assessment with 10-point Likert visual analogue scale (VAS), Heart Rate	1. Higher State-Trait Anxiety Inventory–perceived stress increase was observed during the scenario using the conventional methods (mean 35.4, SD 8.2 to mean 49.8, SD 13.2; a 41.3 %, 35.0 increase) than when using the app (mean 36.1, SD 8.1 to mean 39.0, SD 8.4; a 12.3 %, 29.0 increase). 2. On the Visual Analog Scale questionnaire, participants in the control group reported a higher increase in stress at the peak of the scenario (mean 7.1, SD 1.8 vs mean 6.4, SD 1.9; difference: −0.8, 95 % CI − 1.3 to − 0.2; P = 0.005). 3. Increase in heart rate during the scenario and over the 4 drugs was not different between the 2 groups.
2022 Sellmann (Germany)20	Randomized Controlled Trial; Adult Simulation	Medical emergency cognitive aid with text-based algorithm.	Usual care	Three to six physicians	Intensive care physicians (n = 520 physicians into 80 teams)	Structured questionnaire with 5-point Likert scales	1. In a high percentage, stress level of the participants was diminished. 2. Stress reduction using CA was more likely in “medical” than in “perioperative” subspecialties (3.7 ± 1.2 vs. 2.9 ± 1.2, p < 0.05)
f. Studies with family or friend presence during resuscitation as factors influencing workload or stress
2020 Zehnder (Canada)31	Observational Study; Real patient, neonatal resuscitation	Parental presence during resuscitation	None	Usual care	HCPs participated in neonatal resuscitation in the delivery room	NASA TLX scores	1. TLX score was lower when at least one parent was present (33; 16–47) compared with when no parents were present (46; 29–57) during the resuscitation (p = 0.0004)
2022 Willmes (Germany)21	Randomized Controlled Trial; Adult Simulation	Teams were randomized to a family presence	No family presence		Surgery, internal medicine and anesthesia residents (n = 1085 physicians into 325 teams)	NASA TLX scores	1. Family presence was associated with significantly higher ratings for the domains frustration (45 (30–70) vs 60 (30–75) difference 10, 95 % CI 5 to 15; p < 0.001), temporal demand (70 (50–80) vs 75 (55–85) difference 5, 95 % CI 5–10; p = 0.001) and mental demand (70 (55–80) vs 75 (60–85) difference 5, 95 % CI 0–5; p = 0.009), but no significant differences for the domains physical demand (60 (40–80) vs 65 (40–80) difference 0, 95 % CI 0–5; p = 0.20), effort (65 (50–75) vs 70 (45–80) difference 0, 95 % CI 0–5; p = 0.09) and performance (70 (50–80) vs 70 (45–80) difference 0, 95 % CI 0–5; p = 0.55).
2011 Bjørshol (Norway)28	Randomized Controlled Trial; Adult Simulation	Exposure to socio-emotional stress (an upset friend being emotional and obstructive)	without stress	Two paramedics	paramedics employed at Stavanger University Hospital, Stavanger, Norwa (n = 20 teams)	NASA TLX scores	1. ALS with socioemotional stress resulted in a significantly higher rating for mental demands, temporal demand, effort, and frustration compared with the control condition. 2. There was no difference in physical demands
2022 Sellmann (Germany)18	Randomized Controlled Trial; Adult Simulation	agitated relative with loud crying or mourning as well as walking around the room	withdrawn relative quiet crying, mourning, and quiet observation		residents in the 2nd to 3rd year in internal medicine (n = 355 teams, 113 control, 117 “agitated”), 105″withdrawn“	NASA TLX scores	1. The presence of a relative increased frustration, effort, and perceived temporal demands (all <0.05 compared to control); in addition, an “agitated” relative increased mental demands and total task load (both p < 0.05 compared to “withdrawn” and control group).
g. Studies with healthcare providers previous experiences as factors influencing workload or stress
2018 Fernández-Ayuso (Spain)17	Quasi-experimental study; Adult Simulation	Previous experience in health contexts	None	Second year nursing students		Heart Rate and Blood Pressure, state-trait anxiety questionnaire and perceived stress using Visual Analogical Scale	Both groups demonstrated a decrease in the vital signs and levels of stress/anxiety in subsequent simulation sessions, which suggests a positive adaptive process.

Three RCTs explored telemedicine.19, 32, 36 Butler et al. compared remotely-led resuscitation teams with on-site leaders. They found group using telemedicine experienced significantly higher workload and mental demand compared to groups receiving usual teaching.¹⁹ Another study randomized resuscitation groups to actively being led by remote team leaders or to remote consultants providing guidance on request of team members. Workload was significantly increased for resuscitation team members with a teleconsultant only.³² A study done by the same group revealed less frustration subscores when remote assistantance was provided.³⁶ (Table 1b).

• (3)Workflow

Adjustment of workflow, such as prioritizing chest compression automation with mechanical CPR device,³³ or deliberate re-orientation with task-focusing questions,²³ can reduce perceived workload and stress in simulation-based cardiac arrest scenarios (Table 1c).

• (4)Tools and devices

One simulation RCT used ventilation feedback device and chest compression feedback device, both increased workload for CPR providers.¹⁵ Another RCT investigating real-time feedback device found no significant effect on team leaders, while CPR providers doing chest compressions reported significantly higher workload.²⁵ Interestingly, equipment failure (ie. defective defibrillator) in a simulated cardiac arrest scenario did not result in increased stress for the resuscitation team.²⁹ (Table 1d).

• (5)Cognitive aids and smart apps

The influence of cognitive aids (eg. mobile applications, text-based algorithms) featuring resuscitation protocols or decision support functions, have been investigated in several studies. 16, 20, 24, 26, 31 A smart app designed to help pediatric drug preparation was effective in reducing acute stress in paramedics during simulated pediatric cardiac arrest scenarios,²⁴ whereas another smart app with a built-in resuscitation algorithm did not result in significant difference in team leaders’ workload between teams with and without the use of the app.³¹ However, the mental demand subscore was significantly reduced in a follow-up study.³⁵ The effect of a tablet-based decision support tool on workload results were inconclusive.²⁶ (Table 1e).

• (6)Family presence and socioemotional stress

The effect of disruptive family members,²¹ agitated relatives ¹⁸ or upset friends²⁷ on healthcare providers’ workload during simulated adult resuscitation was investigated in RCTs. Presence of next of kin significantly increased mental demands but did not change physical demands. Nonetheless, the studies utilized actors being obstructive and noxious as standardized family/friends, which may be different in real-world situations. An observational study of real pediatric resuscitations, however, showed that total NASA-TLX score was lower when at least one parent was present³⁰ This is compatible with ILCOR’s systematic review on family presence during resuscitation in pediatric and neonatal cardiac arrest stating mothers/fathers/partners being present during neonatal resuscitation was reasonable.⁴⁰ (Table 1f).

• (7)Provider experience

A quasi-experimental study found no association between level of clinical experience and subjective stress and physiologic parameters among nursing students during simulated resuscitations.¹⁷ (Table 1g).

Discussion

This scoping review has identified factors influencing workload and stress among healthcare providers during CPR, in both real-world and simulated scenarios, including such that can be modified. These are team dynamics such as the roles of nursing leaders and CPR coaches, the integration of telemedicine, the efficiency of workflow with an emphasis on CPR automation, and technological support, like feedback devices or cognitive aids, including manuals or smart applications.

Changes in team composition and roles, telemedicine and workflow, perceived workload and stress differed between team members. A designated medical team leader tends to experience increased workload, but this can be attenuated if a senior nurse is assigned as the nurse leader reducing the team leader’s tasks.³⁴ However, the addition of a CPR coach does not significantly affect the team leader’s overall workload. This may be due to the varied task performed by each team member. Adjustments of these modifiable composition and roles differently change workload. Considering family as potential participants during resuscitation, an ILCOR systematic review about family presence during resuscitation of adult in cardiac arrest also supports the notion that family presence may increase providers’ perceived stress.⁴¹

Remote team leader presence can also increase workload for teams. Conversely, adopting a goal-directed approach or utilizing task-focusing questions during resuscitations can reduce the perceived resuscitation team workload or stress. Initial alterations in workflows or team structures may inherently impose an external cognitive load, as these changes may not be sufficiently familiar to providers. Therefore, including such interventions into training programs might familiarize resuscitation teams with such interventions and reduces workload and stress during real resuscitation.

External support by cognitive aids have shown neutral or positive effects toward reducing stress or workload. Teams exposed to decision support tools experienced higher workload,²⁶ but this finding suggests that introducing new equipment without adequate familiarization could impose additional cognitive burden. The integration of team role diversification, telemedicine, and novel decision support tools in resuscitation settings highlights the critical need for comprehensive education and effective implementation strategies to prevent increased workload and stress among resuscitation teams.

The factors isolated in this review (team composition, roles, workflows, tools, telemedicine, cognitive aids, smart apps, socioemotional stress) are potentially modifiable, which could alleviate or increase their impact on workload or stress, and possibly on resuscitation performance. However, there may be additional factors influencing workload of resuscitation team members that were not covered in our review, because of absence of measurements.⁴²

Some studies excluded from our review provided valuable insights into potential interventions that can modify factors influencing workload and stress. Standardized communication during cardiopulmonary resuscitation in an RCT was found with reduced frustration scores compared to closed-loop communication.⁴³ Five weeks of small-group sessions on stress-mitigation strategies (eg. controlled breathing with relaxing techniques, mental rehearsal and revitalizing breathing techniques) reported lower stress visual analogue scale during simulation at the end of the study.⁴⁴

A discussion point is the NASA task load index that included originally six domains: mental, physical and temporal demands, performance, effort, and frustration, which each contributes to the overall workload. The sum of subscores derives the raw score as surrogate for workload, and variations in the NASA-TLX raw score may predominantly stem from specific components of the score. For instance, a team leader may have lower NASA-TLX raw score from using cognitive aids and appointing nurse leaders, predominantly due to ease of mental demand.²⁷ In contrast, CPR providers reported significantly increased physical demand but less mental demand with additional bedside CPR-coaches during simulated resuscitation.²⁸

In addition to workload and stress, resuscitation performance was also reported as an outcome variable in the included studies. Considering the limited number of studies specifically designed to manipulate workload assessing the impact on resuscitation performance, and if stress and/or workload affect individuals differently, resuscitation performance was intentionally excluded from this review to avoid speculation and preserve the integrity of the results. The complexity of the relationship between resuscitation performance and experienced workload or stress is multifaceted, influenced by individual resilience,⁴⁵ the variability of stress responses among healthcare professionals,⁴⁶ and the unpredictable nature of performance in emergency situations.²⁰ Reporting workload and stress alongside performance indices without careful analysis may inadvertently suggest a misleading correlation. Although our focus was on factors influencing providers' workload, some studies also reported on interventions and resuscitation performances, such as chest compression fractions18, 27, 34 and critical step compliance.31, 47 While these may suggest an association between perceived workload/stress and performance, they do not necessarily imply a causal relationship between the magnitude of workload and resuscitation performance. Nonetheless, the link between workload and/or stress and resuscitation performance remain a significant area of interest calling for well-designed studies investigating this complex interaction between factors, workload, and resuscitation team performance.

Limitations

This scoping review found only a limited number of studies that investigated healthcare providers' workload or stress during resuscitation on patients in real cardiac arrest. This could potentially limit the applicability of our findings, as certain factors may have different effects in real clinical settings compared to simulation. Additionally, workload and stress tolerance vary amongst individuals, cultures, or resource settings but are based on experience, personality factors, and intrinsic capacity. The included studies employed tools like NASA-TLX which did not account for individual tolerability or resilience, limiting our understanding toward how external factors translate into perceived workload or stress on different individuals.

While certain factors may influence providers' perceived workload in simulation, the extent of this influence may vary in real clinical settings. More real-world studies on this topic are needed, as are studies specifically examining the relationship between workload and/or stress and resuscitation performance and patient outcome for further synthesis of knowledge. Influence of personal factors, contextual factors, and clinical experience in mitigating the impact of external stressors and perceived workload is also of interest.

Conclusion

This scoping review provides a comprehensive overview of factors influencing workload and/or stress for healthcare providers during resuscitation in both real-world situation and simulated scenarios. We identified multiple factors affecting stress or workload including team composition, telemedicine, workflow adjustments, and the use of specific tools (e.g. cognitive aids). These findings highlight the imperative for developing and implementing interventions that specifically address identified factors to modulate the workload and stress healthcare providers experience during resuscitation.

CRediT authorship contribution statement

Cheng-Heng Liu: Writing – original draft, Methodology. Chih-Wei Yang: Writing – review & editing, Methodology. Andrew Lockey: Writing – review & editing, Validation. Robert Greif: Writing – review & editing, Supervision. Adam Cheng: Validation, Supervision.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix 1. Search strategies

Concept	Resuscitation		Stress and Workload Measure
Search Terms	(resuscitation? or basic cardiac life support? or basic life support? or code blue? or advanced cardiac life support? or ACLS or CPR or cardi* arrest* or heart arrest* or “return of circulation” or “return of spontaneous circulation” or ROSC or chest compression*) OR (emergency care or emergency health service? or emergency medical service? or emergicenter? or medical emergency service? or prehospital emergency care)	AND	(workload? or work load? or psychological impact or psychological load? or psychological outcome? or psychological skill? or cognitive process* or cognitive error? or reasoning error? or heuristic? or decision fatigue or problem solving intuit* or distraction? or high stake?) OR (stress* adj3 (trainee? or medical team? or medical staff or hospital staff or medical team? or patient care team? or resuscitation team? or resuscitation staff or code team? or code blue team? or health care professional? or health* professional? or health care provider? or health* provider? or health personnel or health care worker? or health* worker? or physician? or doctor? or nurse? or occupational or job or work))
Sample search strategy on Embase	('resuscitation'/exp OR 'basic cardiac life support' OR 'basic life support'/exp OR 'code blue' OR 'advanced cardiac life support'/exp OR 'acls' OR 'cpr' OR 'heart arrest'/exp OR 'return of spontaneous circulation'/exp OR 'rosc' OR 'chest compression'/exp OR resuscitation*:ti,ab,kw OR ((basic NEXT/1 cardiac NEXT/1 life NEXT/1 support*):ti,ab,kw) OR ((basic NEXT/1 life NEXT/1 support*):ti,ab,kw) OR ((code NEXT/1 blue*):ti,ab,kw) OR ((advanced NEXT/1 cardiac NEXT/1 life NEXT/1 support*):ti,ab,kw) OR acls:ti,ab,kw OR cpr:ti,ab,kw OR ((cardi* NEXT/1 arrest*):ti,ab,kw) OR ((heart NEXT/1 arrest*):ti,ab,kw) OR 'return of circulation':ti,ab,kw OR 'return of spontaneous circulation':ti,ab,kw OR rosc:ti,ab,kw OR ((chest NEXT/1 compression*):ti,ab,kw) OR 'emergency care'/exp OR 'emergency health service'/exp OR 'emergency medical service'/exp OR 'emergicenter' OR 'medical emergency service' OR 'prehospital emergency care' OR ((emergency NEXT/1 care):ti,ab,kw) OR ((emergency NEXT/1 health NEXT/1 service*):ti,ab,kw) OR ((emergency NEXT/1 medical NEXT/1 service*):ti,ab,kw) OR emergicenter*:ti,ab,kw OR ((medical NEXT/1 emergency NEXT/1 service*):ti,ab,kw) OR ((prehospital NEXT/1 emergency NEXT/1 care):ti,ab,kw)) AND (stress*:ti,ab,kw AND (trainee*:ti,ab,kw OR 'medical staff'/exp OR ((medical NEXT/1 staff):ti,ab,kw) OR 'hospital staff'/exp OR ((hospital NEXT/1 staff):ti,ab,kw) OR ((medical NEXT/1 team*):ti,ab,kw) OR ((patient NEXT/1 care NEXT/1 team*):ti,ab,kw) OR ((resuscitation NEXT/1 team*):ti,ab,kw) OR 'resuscitation staff':ti,ab,kw OR ((resuscitation NEXT/1 staff):ti,ab,kw) OR ((code NEXT/1 team*):ti,ab,kw) OR ((code NEXT/1 blue NEXT/1 team*):ti,ab,kw) OR ((health NEXT/1 care NEXT/1 professional*):ti,ab,kw) OR ((health* NEXT/1 professional*):ti,ab,kw) OR 'health care provider'/exp OR ((health* NEXT/1 provider*):ti,ab,kw) OR 'health personnel'/exp OR ((health NEXT/1 personnel):ti,ab,kw) OR ((health NEXT/1 care NEXT/1 worker*):ti,ab,kw) OR ((health* NEXT/1 worker*):ti,ab,kw) OR physician*:ti,ab,kw OR doctor*:ti,ab,kw OR nurse*:ti,ab,kw OR occupational:ti,ab,kw OR job:ti,ab,kw OR work:ti,ab,kw) OR workload*:ti,ab,kw OR ((work NEXT/1 load*):ti,ab,kw) OR 'psychological impact'/exp OR ((psychological NEXT/1 impact):ti,ab,kw) OR ((psychological NEXT/1 load*):ti,ab,kw) OR ((psychological NEXT/1 outcome*):ti,ab,kw) OR ((psychological NEXT/1 skill*):ti,ab,kw) OR ((cognitive NEXT/1 process*):ti,ab,kw) OR ((cognitive NEXT/1 error*):ti,ab,kw) OR ((reasoning NEXT/1 error*):ti,ab,kw) OR heuristic*:ti,ab,kw OR 'decision fatigue' OR 'problem solving'/exp OR ((problem NEXT/1 solving):ti,ab,kw) OR intuit*:ti,ab,kw OR distraction*:ti,ab,kw OR ((high NEXT/1 stake*):ti,ab,kw))