https://orcid.org/0000-0002-1486-4446
Abstract
Introduction
Globally, very few settings have undertaken prehospital randomized controlled trials. Given this lack of experience, there is a risk that such trials in these settings may result in protocol deviations, increased prehospital intervals, and increased cognitive load, leading to error. Ultimately, this may affect patient safety and mortality. The aim of this study was to assess the effect of trial-related procedures on simulated scene interval, self-reported cognitive load, medical errors, and time to action.
Methods
This was a prospective simulation study. Using a cross-over design, ten teams of prehospital clinicians were allocated to three separate simulation arms in a random order. Simulations were: (1) Eligibility assessment and administration of freeze-dried plasma (FDP) and a hemoglobin-based oxygen carrier (HBOC), (2) Eligibility assessment and administration of HBOC, (3) Eligibility assessment and standard care. All simulations also required clinical management of hemorrhagic shock. Simulated scene interval, error rates, cognitive load (measured by NASA Task Load Index), and competency in clinical care (assessed using the Simulation Assessment Tool Limiting Assessment Bias (SATLAB)) were measured. Mean differences between simulations with and without trial-related procedures were sought using one-way ANOVA or Kruskal-Wallis test. A p-value of <0.05 within the 95% confidence interval was considered significant.
Results
Thirty simulations were undertaken, representing our powered sample size. The mean scene intervals were 00:16:56 for Simulation 1 (FDP and HBOC), 00:17:22 for Simulation 2 (HBOC only), and 00:14:24 for Simulation 3 (standard care). Scene interval did not differ between the groups (p = 0.27). There were also no significant differences in error rates (p = 0.28) or cognitive load (p = 0.67) between the simulation groups. There was no correlation between cognitive load and error rates (r = 0.15, p = 0.42). Competency was achieved in all the assessment criteria for all simulation groups.
Conclusion
In a simulated environment, eligibility screening, performance of trial-related procedures, and clinical management of patients with hemorrhagic shock can be completed competently by prehospital advanced life support clinicians without delaying transport or emergency care. Future prehospital clinical trials may use a similar approach to help ensure graded and cautious implementation of clinical trial procedures into prehospital emergency care systems.
Introduction
Current evidence to support prehospital emergency care is mostly derived from in-hospital trials or observational studies (Citation1, Citation2). Performing randomized controlled trials (RCTs) in the prehospital setting is difficult due to complex logistics (Citation3, Citation4), significant time and cost requirements (Citation2) and challenges related to protocol training and compliance (Citation3, Citation4) and obtaining informed consent (Citation4). It is also important to balance the effects of trial procedures on patient care and prehospital delays, especially in time-dependent conditions, such as hemorrhagic shock.
Hemorrhagic shock is a leading cause of preventable death after severe trauma in both military and civilian settings (Citation5, Citation6). Prehospitally, early hemorrhage control and targeted hemostatic resuscitation with blood products may improve outcomes (Citation7–11). However, there is a shortage of blood globally (Citation12, Citation13), with Africa, Oceania, and south Asia being most affected (Citation13).
Crystalloid solutions are currently used for resuscitating hemorrhagic shock patients in the prehospital setting. However, aggressive crystalloid resuscitation may lead to hemodilution, coagulopathy, and increased mortality (Citation7), while a reduction in crystalloid use decreases mortality (Citation8). Other alternatives are therefore essential in order to improve mortality following traumatic hemorrhagic shock. The ideal solution should mimic blood by improving oxygen carrying capacity, maintaining or increasing circulating volume, and restoring clotting factors while minimizing coagulopathies.
Where blood transfusion is not available or not an option, we propose the sequential administration of freeze-dried plasma (FDP) and a hemoglobin-based oxygen carrier (HBOC). A prehospital RCT, named the REsuscitation of Shock from Traumatic HemORrhagE (RESTORE) trial, is being planned to investigate the safety and efficacy of this approach on short term mortality, when compared to crystalloids. There are currently no studies that have investigated the efficacy of this potentially life-saving approach in the prehospital setting. This is especially relevant to South Africa (SA), which has high injury burden (Citation14) and critical shortages of blood (Citation15, Citation16).
Given that the planned RCT will be the first such large-scale trial ever conducted in (South) Africa, we needed to determine whether the trial interventions might delay transport or distract prehospital clinicians, which could result in increased errors and harm to potential trial participants. Similarly, it was essential to determine whether prehospital clinicians can comply with trial-related procedures as per protocol, following training. Pilot studies (Citation17–20) have been used in other prehospital trials to inform trial planning and execution. We performed a prospective, randomized, simulation study to assess the effect of trial-related procedures on simulated scene interval as the main outcome. This simulation study represents the first phase of the pilot study for RESTORE (named PRESTORE).
Methods
Study Design
This prospective simulation study assessed scene interval, cognitive load, medical errors, and time to action in simulations performed by prehospital clinicians with and without trial-related procedures. Competency in the prehospital treatment of hemorrhagic shock was also assessed. Using a cross-over design, each participant group was allocated to three separate simulation arms in random order.
Ethical approval was obtained from the Human Research Ethics Committee of Stellenbosch University (Reference Nr. N20/09/099), as Committee of record. This study is reported according to the simulation-based research extensions for the Consolidated Standards of Reporting Trials statement (Citation21).
Setting
South Africa has a formalized emergency medical services (EMS) system. All ambulances are staffed by at least two (basic or intermediate life support) EMS clinicians, and advanced life support (ALS) personnel often work as secondary back-up in rapid response vehicles. Training for prehospital clinicians ranges from 6-week vocational training (registering as a basic life support clinician) to a 2-year national diploma (registering as an ALS clinician). Personnel may also complete 4-year degrees and register as emergency care practitioners with more advanced and critical care skillsets. All South African prehospital clinicians are taught and assessed in simulated environments.
The study was conducted at the Simulation and Clinical Skills Unit of the University of Stellenbosch, South Africa, which provides a high-fidelity simulation environment using state of the art simulation manikins. Simulation exercises may be observed by instructors or assessors from behind a one-way mirror, thus not interfering with nor intimidating participants.
Participants were required to perform clinical assessments, procedures, and clinical interventions in an environment that mimics the real world as much as possible. Actual equipment that prehospital clinicians used during the simulations reflected the equipment that they would have access to during everyday practice. The trial products were placed in a grabber bag 50 m away from the simulation room to simulate a situation where this would be left in the ambulance on an emergency scene.
Sample and Sampling
A convenience sample was constructed of ten teams consisting of three prehospital clinicians each. At least one of the clinicians on each of the ten teams was qualified as an ALS practitioner registered with the Health Professions Council of South Africa (HPCSA). Only the ALS clinician stayed constant for each of the three sim groups. This composition best represents the team in real-world emergency care practice and reflects the requirement of ALS clinicians to work with multiple different teams on emergency scenes. Eligibility criteria were participants from private and public EMS systems in the Western Cape, who are prehospital clinicians in the province and are registered with the HPCSA as such. All participants volunteered their time and provided written informed consent prior to participation.
No data to determine minimally important effect sizes were available to calculate a powered sample size; however, we proposed a minimal sample size of between 30 and 33 simulations (10–11 iterations per simulation). This sample size is based on a previous simulation study that assessed procedural time, error rate, and cognitive load (Citation22). It was determined that eleven simulations in each of the simulation arms of the study were required to have a 90% chance of detecting a difference in means of 4 min at a level of significance of 5% (two-sided). We deemed, a priori, 4 min to be an acceptable minimum change, as a previous prehospital study from South Africa described an approximate 4-min increase in scene interval for every additional clinical intervention (Citation23).
Interventions
The primary intervention was prehospital clinician assessment of patient eligibility for clinical trial inclusion and enrollment, and the implementation of the clinical trial protocol once eligibility was established. Participants received training on the study protocols, medications, and the management of hemorrhagic shock prior to the simulations. The simulations with trial-related procedures were divided into two simulations: one where eligibility was assessed and the patient was deemed eligible and required the administration of both Bioplasma FDP and Hemopure HBOC (Simulation 1), and one where eligibility was assessed, the patient was deemed eligible, and the administration of only HBOC was required (Simulation 2). Simulations with no trial-related procedures other than eligibility assessment (Simulation 3) were deemed standard care and were used as reference simulations. A detailed description of eligibility criteria can be found in the supplementary material, Table S1. Participants were permitted to use a checklist to assess eligibility for trial inclusion.
Outcomes
The primary outcome was scene interval. Secondary outcomes were error rates, cognitive load, time to action, and competency. Scene interval was defined as from the time the simulation started to the time that the ALS clinician stated that he or she would load the patient into the ambulance.
Error was determined in real time by a panel of experts who observed the participants during the simulations. The expert panel consisted of an emergency care practitioner (degree paramedic, WS) with experience in research, one prehospital emergency care educator with expertise in simulation (WC), and one emergency medicine specialist (CvK). An error was defined as an incorrect decision or action without consideration of the outcome. A basic taxonomy of errors was used that categorized errors into diagnostic and management (assessment, algorithm, diagnostic, management, decision-making) errors; medication (wrong medication, dose, route, sequence, type, intravenous fluids) errors; technical (wrong device, size, or incorrect technical skill) errors; and scene and environment (scene safety, scene interval, positioning, patient/family) errors (Citation24).
Cognitive load was assessed using the NASA Task Load Index (TLX), which consists of six domains, each measured by a seven-point Likert scale (Citation25). Dimensions measured as part of the NASA TLX include three dimensions relating to the demands imposed on a participant (mental demand, physical demand, and temporal demand), and three dimensions that relate to the interaction of the participant with the task (effort, performance, and frustration level) (Citation25). The six domains were weighted according to the participant’s subjective importance of each domain, and a total cognitive load score was calculated. After each simulation, the participants completed the TLX questionnaire. The NASA TLX has previously been used to assess cognitive load during simulation for prehospital clinicians (Citation26).
Time to action was defined as the interval from the start of the simulation until an action was completed. Actions that were recorded were aligned to variables that were considered of procedural importance in a clinical trial (time to eligibility assessment) or in the prehospital care of hemorrhagic shock (time to hemorrhage control, time to vital signs, time to intravenous access and fluid resuscitation, and scene interval).
Competency was assessed using the Simulation Assessment Tool Limiting Assessment Bias (SATLAB) (Citation27) to ensure consistency. Competency assessment was completed by the same expert panel that assessed error. Competency was assessed with a six-point Likert scale (3 = best practice, 2 = competent, 1 = not competent, 0 = action omitted, −1 = action results in minor potential harm, −2 = action results in significant potential harm). Qualitative criteria that denote each result were developed to minimize inconsistencies across nine domains that were deemed important in the management of hemorrhagic shock following trauma. These domains were patient assessment, immediate hemorrhage control, eligibility for enrollment, crystalloid infusion, administration of HBOC, administration of FDP, fracture management, pelvic splinting, and patient packaging.
Data obtained from the simulation studies were reviewed by the clinicians in the study team to determine whether these differences may represent clinically significant differences in error, cognitive load, and time to action. Clinical significance was determined through consensus and was defined as any error or delay that was deemed to result in or might result in harm to the simulated patient. Harm was defined in accordance with SATLAB and was deemed significant if the potential harm could have life-threatening consequences. Results were further compared to available literature to inform the assessment of clinical significance.
Randomization
In an attempt to address any unanticipated effects of learning or fatigue as the teams progressed through the simulations, the order in which the simulations were undertaken was assigned randomly using the Microsoft Excel randomization feature (“=RAND()”). This process of randomization was repeated with every new team. Teams were blinded to which simulation they were allocated to. Further, to avoid any learning that might take place between teams, teams were not able to observe the simulations of other teams.
Data Analysis
For scene interval, data were analyzed descriptively. Mean differences between simulations with and without trial-related procedures were reported. The numbers of errors were counted and are reported in whole numbers. For cognitive load we report the individual cognitive load scores per domain, and the overall weighted scores. Time to action is descriptively reported. Competency was assessed by averaging the SATLAB scores across the different assessors and is presented descriptively. A combined result of 2 or greater denotes competence in each domain, or the overall simulation.
Outliers for continuous variables (scene interval, cognitive load, errors) were determined by inspection of box plots. Next, data were assessed for normality using the Shapiro-Wilk test. As data were normally distributed (Shapiro-Wilk p > 0.05), mean differences between simulations with and without trial-related procedures were sought using one-way ANOVA. This test was applied to scene interval and cognitive load. Homogeneity of variances was assessed by Levene’s test, and this was satisfied, p > 0.05 across all variables. Error rate was not normally distributed and as such differences were sought using the Kruskal-Wallis test. For all inferential statistics, a p-value of <0.05 within the 95% confidence interval was considered significant.
We also explored the association between self-reported cognitive load and error rates. The relations between error and cognitive load were determined using Spearman’s correlation, two-tailed. All statistical analyses were done using IBM SPSS v. 28.
Results
A total of 30 simulations (10 per arm) with three crew members each were completed and represents our powered sample size.
The mean (standard deviation, SD) scene (simulation) intervals were as follows: Simulation 1 (Sequential administration of FDP Bioplasma and Hemopure HBOC) 00:16:56 (SD: 00:05:01), Simulation 2 (HBOC only) 00:17:22 (SD: 00:03:38), and Simulation 3 (Standard of care) 00:14:24 (00:04:09). Scene interval did not differ between the simulation groups (p = 0.27).
shows the errors that were made across the different simulations. There were no statistically significant differences (p = 0.28) in error rates across the different simulation arms. Common errors across each of the different simulations are outlined in Table S2 (supplementary material).
Table 1. Errors in simulations with and without trial-related procedures.
provides the mean cognitive load scores for each individual domain and overall. There were no statistically significant differences in cognitive load (p = 0.67) between simulation arms. Cognitive load was not correlated with error rate (r = 0.15, p = 0.42).
Table 2. Cognitive Load in simulations with and without trial-related procedures.
The time to action for important clinically significant interventions is noted in . Clinically important interventions such as hemorrhage control, vital sign assessment, and the administration of a fluid bolus occurred quickly across all simulations.
Table 3. Time to clinically important actions in simulations with and without trial-related procedures.
The aggregated competency scores, rounded to the closest whole number, are provided descriptively in . Competency was achieved in all the criteria and in each of the different simulations overall. There was also good agreement (Fleiss’ kappa = 0.60, CI 95% 0.42–0.78) between the assessors.
Table 4. Competency scores in simulations with and without trial-related procedures.
Discussion
Scene Interval
This study showed no statistically significant effects on scene interval when EMS clinicians implemented trial-related procedures. A few recent studies have challenged a long-held belief that increased scene (or prehospital) interval increases mortality. A prospective Pan-Asian observational study among 24,365 trauma patients failed to demonstrate an effect of prolonged prehospital interval in 30-day mortality (aOR 1.03 (95% CI 0.98–1.09, p = 0.236)) per 10-minute delay. However, every 10-minute increase in total prehospital delay did result in decreased functional survival, i.e., increased morbidity (aOR 1.06 (95% CI 1.04–1.08, p < 0.001)) (Citation28).
Another multicenter, retrospective study (n = 5 499) found that prehospital intervals had no effect on mortality (aOR 1.00 (95% CI 0.98–1.01) per 10 min increments (Citation29). This study has particular relevance to the patient population of interest in RESTORE and other applications for hemorrhagic shock, given that it investigated outcomes in hypotensive (SBP <90 mmHg) adult trauma patients (Citation29).
Similarly, a study from South Africa failed to show an effect of prehospital interval on mortality (Citation30). Out of 532 trauma patients presenting to the facility, 322 (60.5%) were deemed unstable/severely injured. Mortality was not influenced by prehospital transport intervals (p > 0.09) (Citation30). Interestingly, EMS scene intervals were similar to those found in this simulation study (IQR 11–28.3 min).
Conversely, another large study (n = 164 471) found increased mortality (OR 1.21; 95% CI 1.02–1.44; p = 0.03) with prolonged scene interval of >50 min. However, this effect was mitigated when corrected for mediators such as prehospital intubation or patient extrication (Citation31). Similarly, a large systematic review found that prolonged total prehospital intervals resulted in increased mortality for hypotensive patients following penetrating injury and patients with traumatic brain injury. Total prehospital intervals ranging from 30 to 90 min were shown to be significant (Citation32). One study reported increased odds of mortality for patients with penetrating injury should scene interval be longer than 20 min (Citation33).
Taken together, data reported in the literature support our conclusion that trial-related procedures resulting in no statistically significant changes in scene interval, and our projected scene intervals will not constitute any clinically significant risk beyond that of standard prehospital care.
Time to Clinically Important Interventions
Clinically important interventions such as hemorrhage control, vital sign assessment, and the administration of fluid boluses occurred quickly, without differences in time, across all simulations. South Africa data have shown that it takes approximately 4 min for every additional clinical intervention performed, and this does not change patient stability (Citation23). In this simulation study, all interventions took well below this time and thus an effect on the clinical stability of the patient is unlikely. Further, the time spent to determine eligibility was minimal (<2 min), but less in the standard care simulation. This might be because participants only continued until exclusion criteria were encountered. It was also observed that participants generally completed many interventions simultaneously by delegating standard care tasks to crew members based on their scope of practice. This might minimize time to action in general, and reflects what prehospital clinicians do in real-world practice.
Error
Data on errors in prehospital emergency care are extremely limited. However, in the South Africa context it has been reported as occurring “commonly” (Citation34). The errors described in Table S2 (as well as their explanations) are not likely to result in any patient harm, especially considering the inherent limitations of a simulation-based study. Further, given the lower absolute number of errors in the simulations with trial-related procedures, study participation may result in more attentive focus, thus reducing error.
Further supporting that these errors may not be related to the trial procedures per se, a recent study also noted that error correlated to “insufficient clinical knowledge and experience” and “insufficient education and training” (Citation34).
To minimize general clinical error, it might therefore be argued that prehospital clinicians should receive training in managing the clinical condition under study in a clinical trial when receiving training for clinical trial procedures. This not only results in a standard approach and benefits to the control arm, but might result in improved clinical care for other similar patients who are not eligible for the study.
Cognitive Load
While no specific literature could be found on the effect of cognitive load on patient outcomes, it has been demonstrated that increased cognitive load decreases clinical performance (Citation35). A simulation study on laparoscopic surgery found higher levels of error when the cognitive load (as measured by the NASA TLX) was higher, but that cognitive load decreased and performance increased with training (Citation36). Notwithstanding the obvious differences in setting between this study and our simulation study, the cognitive load in the laparoscopic surgery (which resulted in more error) was much higher than what we report (45.0–49.8 vs 59–86). Importantly, there does not currently exist a standardized cognitive load score that is considered high or low, and for this reason it is best to compare cognitive load to some baseline task in a given study with a given sample (Citation37). In the case of this study, it was the standard care simulation.
It may be argued that new procedures could result in more attentive focus (and higher cognitive demand) from participants, which results in less error. This is supported in our results, where a lower absolute error rate and higher cognitive load was found in simulations with trial-related procedures. Given that error (Citation34) and cognitive load (Citation36) seem to be amenable to training and that most of the errors that occurred in the simulations seem to be related to training, we do not believe that the added cognitive load from trial-related procedures reported, would result in any clinically significant effect in real-world practice.
Recommendations for Future Prehospital Clinical Trials
Our results support the idea that prehospital ALS clinicians in South Africa may competently and safely engage in and administer prehospital clinical trials. This is despite the emergency care system having very little previous experience with clinical trial research. The performance of high quality prehospital RCTs is essential for enhancing the evidence base for out-of-hospital emergency care. In settings that have not attempted clinical trials in the prehospital or emergency care setting before, we believe our simulation-based approach may be useful in future studies to help ensure a graded, cautious approach to determine the effect of trial-related procedures not only on protocol compliance, but also on variables that are important to patient safety or outcome.
Limitations
The study was conducted in a simulated environment, which may not fully transfer to real-world clinical practice. This is especially true for other on-scene stressors such as bystanders, limited light, and noise, all of which may affect cognitive load and error rates. In this study, we did not record the level of experience of the prehospital clinicians, which might have affected the outcome measures. However, this would not have affected any potential differences of results by the same team under different conditions.
All three simulations for a specific team occurred on the same day, and this could have resulted in some learning, affecting the performance of subsequent simulations. Attempts were made to mitigate bias arising from this by randomizing the order of the simulations and presenting different scenarios (with the same outcomes) across the different simulations. The randomization schedule is contained in Table S3 (supplementary material).
We did not have specific data from which to power this study. The sample size was based on data from a previous simulation study that assessed procedural time and another retrospective study that assessed scene interval. One may criticize our study as underpowered, considering that it did not detect a difference for the primary outcome measure, scene interval. However, we calculated that repowering of a similar study based on the data obtained would require participation of approximately 70 teams, rather than ten. While this is technically possible, it is not feasible.
Conclusion
We demonstrated that in a simulated environment, trial-related procedures and clinical management of patients with hemorrhagic shock can be completed competently by prehospital ALS clinicians without prolonging scene interval to any statistically or clinically significant levels. Time to action of clinically important interventions and trial procedures were also not affected in any meaningful manner. Neither cognitive load nor error rates were statistically different. The errors that occurred commonly related to inherent limitations with simulation-based assessments or to hemorrhagic shock emergency care in general, rather than the trial-related procedures.
Supplemental Material
Download MS Word (18.8 KB)Acknowledgments
The authors would like to acknowledge Bronwen Espen, Regina Mlobeli, and Nozizwe Makola for their assistance in arranging the simulation events. We also acknowledge Jocelyn Park-Ross for her assistance in facilitating some of the simulations.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Additional information
Funding
References
- Sayre MR, Gausche-Hill M. Conducting randomized trials in the prehospital setting. Prehosp Emerg Care. 2002;6(2 Suppl):S38–S47. doi:10.3109/10903120209102681.
- Osterwalder JJ. Insufficient quality of research on prehospital medical emergency care–where are the major problems and solutions? Swiss Med Wkly. 2004;134(27–28):389–94. doi:10.4414/smw.2004.10357.
- Lerner EB, Weik T, Edgerton EA. Research in prehospital care: overcoming the barriers to success. Prehosp Emerg Care. 2016;20(4):448–53. doi:10.3109/10903127.2014.980480.
- Stassen W, Wallis L, Castren M, Vincent-Lambert C, Kurland L. A prehospital randomised controlled trial in South Africa: challenges and lessons learnt. Afr J Emerg Med. 2019;9(3):145–9. doi:10.1016/J.AFJEM.2019.02.002.
- Cannon JW. Hemorrhagic shock. N Engl J Med. 2018;378(4):370–9. doi:10.1056/NEJMra1705649.
- Wilson MH, Habig K, Wright C, Hughes A, Davies G, Imray CHE. Pre-hospital emergency medicine. Lancet. 2015;386(10012):2526–34. doi:10.1016/S0140-6736(15)00985-X.
- Kasotakis G, Sideris A, Yang Y, de Moya M, Alam H, King DR, Tompkins R, Velmahos G. Aggressive early crystalloid resuscitation adversely affects outcomes in adult blunt trauma patients: an analysis of the Glue Grant database. J Trauma Acute Care Surg. 2013;74(5):1215–21. doi:10.1097/TA.0b013e3182826e13.
- Simmons JW, White CE, Eastridge BJ, Holcomb JB, Perkins JG, Mace JE, Blackbourne LH. Impact of improved combat casualty care on combat wounded undergoing exploratory laparotomy and massive transfusion. J Trauma. 2011;71(1 Suppl):S82–S86. doi:10.1097/TA.0b013e3182218ddb.
- Dries DJ. The contemporary role of blood products and components used in trauma resuscitation. Scand J Trauma Resusc Emerg Med. 2010;18(1):63. doi:10.1186/1757-7241-18-63.
- Miller TE. New evidence in trauma resuscitation-is 1: 1: 1 the answer? Perioper Med. 2013;2(1):13. doi:10.1186/2047-0525-2-13.
- Pusateri AE, Moore EE, Moore HB, Le TD, Guyette FX, Chapman MP, Sauaia A, Ghasabyan A, Chandler J, McVaney K, et al. Association of prehospital plasma transfusion with survival in trauma patients with hemorrhagic shock when transport times are longer than 20 minutes. JAMA Surg. 2020;155(2):e195085. doi:10.1001/jamasurg.2019.5085.
- Raykar NP, Makin J, Khajanchi M, Olayo B, Munoz Valencia A, Roy N, Ottolino P, Zinco A, MacLeod J, Yazer M, et al. Assessing the global burden of hemorrhage: the global blood supply, deficits, and potential solutions. SAGE Open Med. 2021;9. doi:10.1177/20503121211054995.
- Roberts N, James S, Delaney M, Fitzmaurice C. The global need and availability of blood products: a modelling study. Lancet Haematol. 2019;6(12):e606–e615. doi:10.1016/S2352-3026(19)30200-5.
- Matzopoulos R, Prinsloo M, Pillay-van Wyk V, Gwebushe N, Mathews S, Martin LJ, Laubscher R, Abrahams N, Msemburi W, Lombard C, et al. Injury-related mortality in South Africa: a retrospective descriptive study of postmortem investigations. Bull World Health Organ. 2015;93(5):303–13. doi:10.2471/BLT.14.145771.
- Hodkinson PW, Wallis LA. Cross-sectional survey of patients presenting to a South African urban emergency centre. Emerg Med J. 2009;26(9):635–40. doi:10.1136/emj.2008.063362.
- Swart L-A, Laher H, Seedat M, Gantchev G. Urban and rural differences in child injury deaths in South Africa: a one-year review. Afr Saf Promot. 2012;10(2):28–40.
- Smith IM, Crombie N, Bishop JR, McLaughlin A, Naumann DN, Herbert M, Hancox JM, Slinn G, Ives N, Grant M, et al. RePHILL: protocol for a randomised controlled trial of pre-hospital blood product resuscitation for trauma. Transfus Med. 2018;28(5):346–56. doi:10.1111/tme.12486.
- Fuller GW, Goodacre S, Keating S, Perkins G, Ward M, Rosser A, Gunson I, Miller J, Bradburn M, Thokala P, et al. The ACUTE (Ambulance CPAP: use, treatment effect and economics) feasibility study: a pilot randomised controlled trial of prehospital CPAP for acute respiratory failure. Pilot Feasibility Stud. 2018;4(1):86–10. doi:10.1186/s40814-018-0281-9.
- Chin EJ, Chan CH, Mortazavi R, Anderson CL, Kahn CA, Summers S, Fox JC. A pilot study examining the viability of a prehospital assessment with ultrasound for emergencies (PAUSE) protocol. J Emerg Med. 2013;44(1):142–9. doi:10.1016/j.jemermed.2012.02.032.
- Thomas M, Voss S, Benger J, Kirby K, Nolan JP. Cluster randomised comparison of the effectiveness of 100% oxygen versus titrated oxygen in patients with a sustained return of spontaneous circulation following out of hospital cardiac arrest: a feasibility study. PROXY: post ROSC OXYgenation study. BMC Emerg Med. 2019;19(1):16. doi:10.1186/s12873-018-0214-1.
- Cheng A, Kessler D, Mackinnon R, Chang TP, Nadkarni VM, Hunt EA, Duval-Arnould J, Lin Y, Cook DA, Pusic M, et al. Reporting guidelines for health care simulation research: extensions to the CONSORT and STROBE statements. Adv Simul. 2016;1(1):25. doi:10.1186/s41077-016-0025-y.
- Swinton P, Corfield AR, Moultrie C, Percival D, Proctor J, Sinclair N, Perkins ZB. Impact of drug and equipment preparation on pre-hospital emergency Anaesthesia (PHEA) procedural time, error rate and cognitive load. Scand J Trauma Resusc Emerg Med. 2018;26(1):82. doi:10.1186/s13049-018-0549-3.
- van Niekerk G, Welzel T, Stassen W. Clinical interventions account for scene time in a helicopter emergency medical service in South Africa. Air Med J. 2018;37(6):357–61. doi:10.1016/j.amj.2018.07.027.
- Herzberg S, Hansen M, Schoonover A, Skarica B, McNulty J, Harrod T, Snowden JM, Lambert W, Guise J-M. Association between measured teamwork and medical errors: an observational study of prehospital care in the USA. BMJ Open. 2019;9(10):e025314. doi:10.1136/bmjopen-2018-025314.
- Hart SG, Staveland LE. Development of NASA-TLX (Task Load Index): results of empirical and theoretical research. Adv Psychol. 1988;52:139–83.
- Mills BW, Carter OB-J, Rudd CJ, Claxton LA, Ross NP, Strobel NA. Effects of low-versus high-fidelity simulations on the cognitive burden and performance of entry-level paramedicine students: a mixed-methods comparison trial using eye-tracking, continuous heart rate, difficulty rating scales, video observation and interviews. Simul Healthc. 2016;11(1):10–8. doi:10.1097/SIH.0000000000000119.
- Makkink AW, Vincent-Lambert C. The development of ‘SATLAB’: a tool designed to limit assessment bias in simulation-based learning. South Afr J Pre-Hosp Emerg Care. 2020;1(1):26–34. doi:10.24213/1-1-3024.
- Chen C-H, Shin S, Do Sun J-T, Jamaluddin SF, Tanaka H, Song KJ, Kajino K, Kimura A, Huang EP-C, Hsieh M-J, et al. Association between prehospital time and outcome of trauma patients in 4 Asian countries: a cross-national, multicenter cohort study. PLoS Med. 2020;17(10):e1003360. doi:10.1371/journal.pmed.1003360.
- Shibahashi K, Ishida T, Sugiyama K, Kuwahara Y, Okura Y, Hamabe Y. Prehospital times and outcomes of patients who had hypotension at the scene after trauma: a nationwide multicentre retrospective study. Injury. 2020;51(5):1224–30. doi:10.1016/j.injury.2020.02.002.
- Möller A, Hunter L, Kurland L, Lahri S, van Hoving DJ. The association between hospital arrival time, transport method, prehospital time intervals, and in-hospital mortality in trauma patients presenting to Khayelitsha Hospital, Cape Town. Afr J Emerg Med. 2018;8(3):89–94. doi:10.1016/j.afjem.2018.01.001.
- Brown JB, Rosengart MR, Forsythe RM, Reynolds BR, Gestring ML, Hallinan WM, Peitzman AB, Billiar TR, Sperry JL. Not all prehospital time is equal: influence of scene time on mortality. J Trauma Acute Care Surg. 2016;81(1):93–100. doi:10.1097/TA.0000000000000999.
- Harmsen AMK, Giannakopoulos GF, Moerbeek PR, Jansma EP, Bonjer HJ, Bloemers FW. The influence of prehospital time on trauma patients outcome: a systematic review. Injury. 2015;46(4):602–9. doi:10.1016/j.injury.2015.01.008.
- McCoy CE, Menchine M, Sampson S, Anderson C, Kahn C. Emergency medical services out-of-hospital scene and transport times and their association with mortality in trauma patients presenting to an urban Level I trauma center. Ann Emerg Med. 2013;61(2):167–74. doi:10.1016/j.annemergmed.2012.08.026.
- Rowland M, Adefuye A. Human errors and factors that influence patient safety in the pre-hospital emergency care setting: perspectives of South African emergency care practitioners. Health SA. 2022;27:1798. doi:10.4102/hsag.v27i0.1798.
- Josephsen J. Cognitive load theory and nursing simulation: an integrative review. Clin Simul Nurs. 2015;11(5):259–67. doi:10.1016/j.ecns.2015.02.004.
- Yurko YY, Scerbo MW, Prabhu AS, Acker CE, Stefanidis D. Higher mental workload is associated with poorer laparoscopic performance as measured by the NASA-TLX tool. Simul Healthc. 2010;5(5):267–71. doi:10.1097/SIH.0b013e3181e3f329.
- Hart SG. NASA-task load index (NASA-TLX); 20 years later. Proc Hum Factors Ergon Soc Annu Meet. 2006;50(9):904–8. doi:10.1177/154193120605000909.