Cardiopulmonary resuscitation missed by bystanders: Collateral damage of coronavirus disease 2019

Abstract Objective The coronavirus disease 2019 (COVID‐19) pandemic changed the time‐dependent cardiac arrest network. This study aims to understand whether the rescue standards of cardiopulmonary resuscitation (CPR) and out‐of‐hospital cardiac arrest (OHCA) were handled differently during pandemic compared to the previous year. Methods Data for the years 2019 and 2020 were provided by the records of the Lombardy office of the Regional Agency for Emergency and Urgency. We analysed where the cardiac arrest occurred, when CPR started and whether the bystanders used public access to defibrillation (PAD). Results During 2020, there was a reduction in CPRs performed by bystanders (odds ratio [OR] = 0.936 [95% confidence interval (CI95%) 0.882–0.993], p = .029) and in the return of spontaneous circulation (ROSC) (OR = 0.621 [CI95% 0.563–0.685], p < .0001), while there was no significant reduction in the use of PAD. Analysing only March, the period of the first wave in Lombardy, the comparison shows a reduction in bystanders CPRs (OR = 0.727 [CI95% 0.602–0.877], p = .0008), use of PAD (OR = 0.441 [CI95% 0.272–0.716], p = .0009) and in ROSC (OR = 0.179 [CI95% 0.124–0.257], p < .0001). These phenomena could be influenced by the different settings in which the OHCAs occurred; in fact, those that occurred in public places with a mandatory PAD were strongly reduced (OR = 0.49 [CI95%, 0.44–0.55], p < .0001). Conclusions COVID‐19 had a profound impact on the time‐dependant OHCA network. During the first pandemic wave, CPR and PAD used by bystanders decreased. The different contexts in which OHCAs occurred may partially explain these differences.


Editorial Comment
This retrospective analysis of out-of-hospital cardiac arrest cases and bystander-initiated life support in one region in Italy showed major differences during the first year of the COVID-19 pandemic compared to a matched pre-pandemic period.

| INTRODUCTION
The coronavirus disease 2019 (COVID- 19) pandemic has had an impact on the epidemiology of out-of-hospital cardiac arrest (OHCA), 1 indeed, the incidence 1-3 and mortality 4-6 of OHCAs increased in 2020. However, there is no clear evidence whether these phenomena are due to the major complications of COVID-19 or to delayed response times and late arrival of ambulances in an attempt to save patients 7,8 or even to the lack of CPR training due to the logistical complications of the pandemic. 9 Recent studies have shown that OHCAs increased during the pandemic period by up to 60% compared with 2019. 2,10 Furthermore, the increase in the number of OHCAs was followed by a decrease in the return of spontaneous circulation (ROSC) by up to 41%. 11,12 These trends are probably the consequences of a 10% lower number of CPRs performed by bystanders 13,14 prior to the arrival of ambulance and/or Advanced Life Support (ALS) vehicle, and the delay in emergency medical service (EMS) response due to reorganisation and profound stress. 15,16 In addition, the Italian Ministry of Health recommended that people should not enter Emergency Departments (EDs), but call the 112 emergency number, and patients often refused to be transported by ambulance to the hospital ED for fear of being infected. 17,18 Thus, the reluctance of patients to be transported to the ED during the pandemic period changed the epidemiology of time-related diseases, 19,20 increasing the number of cardiac arrests occurring at home. 21 Since the beginning of the pandemic, bystanders' fear of being infected while performing CPR has been a debated topic in the scientific community, especially among those involved in lay education. 22,23 Therefore, as chest compressions can produce aerosols, the chain of survival had to be modified to ensure a greater safety for bystanders.
However, there has been a reduction in the number of CPRs performed by bystanders and in the use of public access to defibrillation (PAD). 24,25 Consequences include increased mortality after OHCA and a worst neurologic outcome for hospitalised patients. 26 Lombardy was the region most affected by the COVID-19 outbreak in Italy. During the first wave, through April 30, Lombardy had more than 75,000 verified positive cases, and 13,772 deaths out of the 27,967 (49.2%) that occurred throughout Italy. 8

| METHODS
This is a retrospective observational cohort study. The study was conducted according to the principles of the declaration of Helsinki and was approved by the Regional Agency for Emergency and Urgency

| Data registry
We analysed all records saved as "cardiac arrest" in Emergency Management (EmMa) database in 2019 and 2020. The label "cardiac The data analysis process was conducted employing the Statistical Analysis System of AREU (SAS-AREU portal). The portal contains all data related to emergency calls. All types of cardiac arrest, both medical and traumatic, were selected.

| The Regional Agency for Emergency and Urgency
In the Lombardy region, AREU is responsible for pre-hospital EMS. This region was the first in Europe to face with COVID-19, [27][28][29][30] and consequently had to modify its emergency system. 30 AREU coordinates and ensures territorial first aid by means of 265 ambulances with a crew of 2-3 rescuers qualified in Basic Life Support manoeuvres, 50 Intermediate Rescue Vehicles (ambulance or car) with a nurse, 59 Advanced Rescue Vehicles with a doctor qualified in ALS and 5 helicopters with ALS crew. All interventions carried out by EMS teams are recorded on the EmMa portal. 31 Detailed description of the system can be obtained from literature. 8,16,32

| Statistical analysis
Continuous variables are presented as mean and SD while categorical variables are presented as numbers and percentages. The total number of OHCAs was analysed by means of t-test for paired data.
The total number of CPRs performed by bystanders, the num- 3 | RESULTS Table 1 shows the total number of OHCAs that occurred and were         23 We also found a reduction in OHCA in public areas where the PADs were available in 2020 compared to 2019. This reduction is particularly significant, as in 2020 the number of OHCAs occurring in places where PAD was present was 3.9%, compared to 7.7% in 2019.
The reason, most likely, is due to the fact that, especially during the pandemic waves, many workers were smart working and many com- We cannot exclude that this phenomenon was due to media awareness related to cardiac arrest and its treatment that occurred in 2020 as a result of the pandemic compared with the previous year. If this was the case, the public was sensitized to pay more attention to the problem, and it is likely that when PAD could be used, this occurred more frequently than in the past. Despite this, ROSC decreased significantly in 2020, which is consistent with the literature. 1 In March 2020, we registered the largest reduction in ROSC com- For example, the Utstein-based ROSC score was derived using the following variables: age, aetiology, location, assisted OHCA, bystander CPR, time to EMS arrival, and attachment rate. [33][34][35] The presence of all the variables described would have allowed us to use the model mentioned above and to do further investigation and analysis on the ROSC data. Unfortunately, the lack of some variables does not allow further investigation that might be done in future studies.
The study found several changes strongly associated with pandemic peaks, thus achieving the main aim of the study. However, a limitation of the study is the possible reduced accuracy in the data collecting process due to the high pressure and stress of the emergency phase. Examining the impact of COVID-19 on time-dependent networks becomes essential to fully understand the collateral damage of the pandemic.

| CONCLUSIONS
The

FUNDING INFORMATION
There are no funders to report for this submission.