ABSTRACT
Background COVID-19 poses many challenges to healthcare systems and workers. Responding to medical emergencies in patients with suspected COVID-19 will require new guidelines and protocols. Simulation can support their development.
Methods We organised seven simulations involving patients with suspected COVID-19 for staff at Brighton and Sussex University Hospitals. Participants completed pre- and post-simulation questionnaires.
Results Fifty-six staff participated and they reported being significantly less prepared to respond to an emergency in a patient with suspected COVID-19 than in one in whom it is not suspected. The simulations significantly improved the participants' confidence in responding to emergencies in patients with suspected COVID-19. Numerous challenges were identified along the themes of equipment, personnel, communication and procedures.
Conclusions Low-fidelity simulation can provide relevant and timely information on how prepared health systems and their workforce are to respond to emergencies. We urge NHS trusts nationally to implement simulations to identify problems and develop effective solutions.
Introduction
Simulation is playing an increasing role in medical education globally.1 It provides the opportunity to assess not only clinical knowledge, but also the soft skills that are required to work effectively in emergency situations without compromising patient safety. There has been a large focus on high-fidelity simulation that acts to replicate the working environment as closely as possible, with the literature supporting its effectiveness as a teaching method.2,3 High-fidelity simulation is often undertaken in a simulation suite utilising video playback software and responsive mannequins. Less focus has been placed on low-fidelity simulation (which uses the minimal resources necessary to achieve its learning objectives). However, it can be used effectively to stress-test systems and to provide clinical staff with exposure to situations in their usual work context.4
The global COVID-19 pandemic is expected to challenge health systems universally, with the expected number of acutely unwell patients predicted to increase drastically.5 As a result of this, hospitals, staff and their systems will have to adapt to ever-changing conditions. This particularly relates to the use of personal protective equipment (PPE) and the approach to medical emergencies. Low-fidelity simulation not only has the potential to provide staff with more experience in these areas, but also allows the testing and improvement of protocols and guidance.6
At this NHS trust, a medical emergency team (MET) responds to emergencies across the hospital site. Depending on the nature of the emergency call placed, a multidisciplinary team responds and includes the medical registrar, several below-registrar grade junior doctors, the critical care outreach team (CCOT) and anaesthetists/intensivists.
This study aims to pilot a low-fidelity COVID-19 MET response simulation to improve our understanding of the role that simulation can play in this global crisis, to offer a safe training environment to staff, and to identify problems with the current protocols while providing workable solutions.
Methods
We organised and ran seven simulations in barrier-nursed side rooms at Brighton and Sussex University Hospitals NHS Trust. The simulation team consisted of one patient (a junior doctor), two ward nurses (one acting as initiator), one ward doctor (FY1 grade), and the medical emergency / cardiac arrest team (a three- to four-person team comprising a combination of FY1s, SHOs and SpRs). Scenarios were run as realistically as possible (eg no response until a simulated call was placed, donning and doffing of appropriate PPE, use of the resuscitation trolley, and simulation of arterial blood gases (ABG) / venepuncture / cannulation).
To ensure that we had an adequate overview of the scenarios, two facilitators observed different aspects of the simulation. One provided observations and clinical findings to the participants, on request, from a pre-designed simulation template. The other was tasked with observing and providing feedback focused on the communication and procedural issues of running an emergency and arrest call through a closed door. The simulations were often observed by non-participating members of the ward team to maximise interprofessional learning. We asked all participants to complete an online pre- and post-simulation questionnaire and feedback form, hosted on Qualtrics, reviewing perceived preparedness for MET / cardiac arrest calls in patients with suspected or confirmed COVID-19 (supplementary materials).
Scenario one
A 75-year-old male was admitted with a presumed diagnosis of decompensated heart failure, for which he required continuous positive airway pressure (CPAP); however, having also presented with fever and a cough, the COVID-19 pathway had been triggered. He was isolated in a side room, with contacts required to wear full PPE. The scenario begins with the nurse placing an emergency call as he becomes unresponsive and shows no signs of life.
Scenario two
We wanted to focus on how the team would deal with managing a MET call that did not require full aerosol-generating PPE but where the patient deteriorates, mid-simulation, into a cardiac arrest to monitor efficiency and time to don and doff PPE in a time-critical scenario. To ensure that we did not overload the participants, we used the same clinical background as scenario one. The scenario commences with an acutely unwell patient admitted with decompensated heart failure on nasal oxygen, but being barrier nursed as a query COVID-19 case. The nurse puts out an emergency call as he deteriorates, becoming hypoxic and tachypnoeic. The scenario follows an initial A–E assessment, with the team realising the need for CPAP. The patient then becomes unresponsive with no signs of life before CPAP can be set up. A cardiac arrest scenario follows.
Both scenarios were stopped by the faculty once a suitable end point had been reached (based upon the simulation template). In between each scenario, the faculty led the debrief focusing on procedural issues, as well as soft skills feedback to individual participants. These were documented and collated by the facilitators, and solutions were proposed where possible.
Data were imported and analysed in R Studio (v1.02.5033). Statistics were performed using a two-sample paired t-test.
Results
Fifty-six people participated in the simulations and completed the surveys. FY1s were the largest contingent with 22 of them participating in the simulations, followed by 16 nurses, 14 SHOs, three SpRs and one other. Of the 56 participants, 20% (n=11) had been to a previous MET call in a patient with suspected COVID-19 and 5% (n=3) had been to a cardiac arrest in a patient with suspected COVID-19.
The majority of respondents (87.5%, n=49) reported knowing where to find local trust guidelines, and the majority had reviewed updated guidance on PPE, aerosol-generating procedures (AGPs) and adult life support (91%, 84% and 91% respectively). In addition, 80% (n=45) of participants were aware that chest compressions are an AGP (as per local trust guidelines).
The following questions were assessed on a five-point Likert scale (1: low, 5: high). Participants felt significantly less prepared for a MET call in a patient with suspected COVID-19 than in a patient in whom it is not suspected; the mean Likert response was 2.4 and 3.9 respectively (p<0.001, Fig 1). The same was true for cardiac arrest calls, with a mean response for patients with suspected COVID-19 of 2.2 compared with 3.9 for non-COVID-19 patients (p<0.001, Fig 1).
The responses on a Likert scale (1: completely unprepared, 5: completely prepared) for responding to a normal/COVID medical emergency / cardiac arrest. **p<0.001. The ‘Count’ axis is aligned such that response 3 (‘Somewhat prepared (eg still unsure on aspects)’) is equally distributed either side of 0.
Prior to the simulation, participants were least confident in ‘Acting as first responder in an emergency call to a suspected COVID-19 patient’ (pre-simulation mean Likert score of 2.82) and in ‘Managing the scenario if the patient passes away’ (pre-simulation mean Likert score of 2.64). Table 1 outlines the scores for the remainder of the aspects of the emergency response assessed in Question 8 of the pre-simulation questionnaire.
Confidence in undertaking different roles in the emergency response
All participants, including the observer, felt that the simulations improved their preparedness for responding to an emergency call in a patient with suspected COVID-19. This was evidenced by the fact that the mean Likert score for responding to COVID-19 MET calls increased from 2.4 to 3.7 (p<0.001, Fig 2) after simulation. Similarly, the simulations increased the mean Likert score from 2.2 to 3.6 for COVID-19 cardiac arrest calls (p<0.001, Fig 2).
The responses on a Likert scale (1: completely unprepared, 5: completely prepared) for responding to COVID MET/arrest calls pre- and post-simulation. **** p<0.001. The ‘Count’ axis is aligned such that response 3 (‘Somewhat prepared (eg still unsure on aspects)’) is equally distributed either side of 0.
The simulations were run successfully. It was felt that the low-fidelity simulations were realistic enough to identify the challenges posed by an emergency in a patient with suspected COVID-19. This was evidenced by the fact that participants reported no concerns with the operating of the simulation apart from the fact that there was little briefing before the simulation. This was a design feature of the simulation, as one of the aims was to stress-test the current systems.
For further simulations, the participants responded with two main requests. The first was to run more simulations (of the same type) on a regular basis to improve skills. The second was to run simulations that involved different skillsets (eg setting up CPAP).
Assessment of the current emergency response
The content of the faculty-led debrief and post-simulation questionnaire identified four key themes: communication, equipment, personnel and procedures. Table 2 details the issues and their potential solutions.
Descriptions of the identified issues and proposed solutions for responding to a medical emergency or cardiac arrest call for a patient with suspected COVID-19
Discussion
COVID-19 poses an unprecedented challenge to healthcare systems and workers globally.5 Healthcare professionals report a significant degree of unfamiliarity with the novel protocols and guidance, in particular the use of PPE and responding to MET calls. This study set out to demonstrate the potential role that low-fidelity simulation can play in supporting staff to feel more prepared to respond to challenging situations involving patients with suspected COVID-19.
At baseline, the participants felt significantly less prepared to deal with an emergency call in a patient with suspected COVID-19 than in a patient in whom COVID-19 is not suspected. After just two simulated scenarios involving patients with COVID-19, the participants felt significantly more prepared to respond to a medical emergency. In addition, participants felt significantly more confident to perform all aspects of the emergency response after a simulation.
The participants reported that the simulations were well run and without any significant operational issues. This is important as it demonstrates that low-fidelity simulations can be organised and identify issues with current guidelines in a timely manner. The simulations revealed issues in four key themes: communication, equipment, personnel and procedures.
In particular, the participants reported concerns over the feasibility of running a cardiac arrest with only four people inside the room (as was local NHS trust guidance on 4 April 2020), given that the following roles need to be shared among them: compressions, defibrillation (often linked to timing), procedures (eg ABG and cannulation), team leader, airway management and in-room runner. In particular, the ‘in-room runner’ was a newly identified role and therefore it was felt that a minimum of five people were required to be inside the room (with some appropriate rotation of roles). Similarly, it was found to be crucial that there be a ‘gatekeeper’ outside the room, in addition to outside ‘runners’. Table 2 outlines the issues raised and details potential solutions. A number of the solutions are relatively simple (eg producing ‘grab bags’ and equipment lists), further illustrating that low-fidelity simulation can aid in the development of quick, high-yield solutions. These solutions were specific to this NHS trust, and it is likely that performing simulations in different NHS trusts would identify different challenges and thus require different solutions.
This study has several limitations that require addressing. We describe the results of a small number of simulation groups, and as a result we are limited by the sample size available to us. Secondly, the simulation is set up with a COVID patient isolated in a side room, as this was the case for the majority of COVID patients at the conception of the study. However, this is unlikely to remain the case, with bays of COVID patients beginning to exist.
Additionally, this study only demonstrates an improvement in the participants' self-evaluated preparedness to respond to a COVID emergency and its different aspects. It is likely that this increased preparedness relates to an improved knowledge base and awareness of the skills required. However, we have not attempted to demonstrate improvement in those measures objectively.
Despite this, our methodology achieved the study aims of improving participants' preparedness to respond to emergency calls, supporting the notion that low-fidelity simulation is a useful tool in this context, and identifying the current gaps in the emergency response. This paper is an interim review as the results of further simulations, including scenarios in different environments, will be published in a later article. Low-fidelity in situ simulation is dependent upon the availability of space and staff, which, although present now, may not persist. As such, the authors made the decision to publish an interim report as it felt necessary to disseminate the initial conclusions swiftly in order to support other NHS trusts.
The current climate demonstrates the need for clear, concise and effective guidance. This simulation study has acted as a stress test for current NHS trust protocols.
This study provides clear evidence that low-fidelity clinical simulation can be organised quickly, with minimal resources, and identify issues with current guidance while also providing the opportunity to develop workable solutions. In addition, we received very positive feedback that it improved the preparedness of the staff who will be involved in emergency calls to patients with suspected COVID-19. We recommend that NHS trusts nationwide adopt this approach in light of the COVID-19 pandemic.
Acknowledgements
We would like to thank the participants of the simulations.
- © Royal College of Physicians 2020. All rights reserved.
Jump to section
Related Articles
Cited By...
- No citing articles found.