key: cord-0963102-vqza05mt
authors: Rosell Ortiz, Fernando; Fernández del Valle, Patricia; Knox, Emily C.; Jiménez Fábrega, Xavier; Navalpotro Pascual, José M.; Mateo Rodríguez, Inmaculada; Ruiz Azpiazu, José I.; Iglesias Vázquez, José A.; Echarri Sucunza, Alfredo; Alonso Moreno, Daniel F.; Forner Canos, Ana B.; García-Ochoa Blanco, María J.; López cabeza, Nuria; Mainar Gómez, Belén; Batres Gómez, Susana; Cortés Ramas, José A.; ceniceros Rozalén, María I.; Guirao Salas, Francisco A.; Fernández martínez, Begoña; Daponte Codina, Antonio
title: Influence of the Covid-19 pandemic on out-of-hospital cardiac arrest. A Spanish nationwide prospective cohort study
date: 2020-10-10
journal: Resuscitation
DOI: 10.1016/j.resuscitation.2020.09.037
sha: 740628794e2e01200b5faa08e2e0e57540291de1
doc_id: 963102
cord_uid: vqza05mt

AIMS: The influence of the COVID-19 pandemic on attendance to out-of-hospital cardiac arrest (OHCA) has only been described in city or regional settings. The impact of COVID-19 across an entire country with a high infection rate is yet to be explored. METHODS: The study uses data from 8629 cases recorded in two time-series (2017/2018 and 2020) of the Spanish national registry. Data from a non-COVID-19 period and the COVID-19 period (February 1st to April 30th 2020) were compared. During the COVID-19 period, data a further analysis comparing non-pandemic and pandemic weeks (defined according to the WHO declaration on March 11th, 2020) was conducted. The chi-squared analysis examined differences in OHCA attendance and other patient and resuscitation characteristics. Multivariate logistic regression examined survival likelihood to hospital admission and discharge. The multilevel analysis examined the differential effects of regional COVID-19 incidence on these same outcomes. RESULTS: During the COVID-19 period, the incidence of resuscitation attempts declined and survival to hospital admission (OR = 1.72; 95%CI = 1.46–2.04; p < 0.001) and discharge (OR = 1.38; 95%CI = 1.07–1.78; p = 0.013) fell compared to the non-COVID period. This pattern was also observed when comparing non-pandemic weeks and pandemic weeks. COVID-19 incidence impinged significantly upon outcomes regardless of regional variation, with low, medium, and high incidence regions equally affected. CONCLUSIONS: The pandemic, irrespective of its incidence, seems to have particularly impeded the pre-hospital phase of OHCA care. Present findings call for the need to adapt out-of-hospital care for periods of serious infection risk. STUDY REGISTRATION NUMBER: ISRCTN10437835.

Since its outbreak, the pandemic due to the Sars-CoV-2 virus has resulted in high morbidity and mortality all over the world. The effects of the pandemic, together with the containment measures adopted by the majority of countries, has led to modifications in how health services function and in the care provided to patients. Such modifications have affected health care and cardiovascular emergencies, amongst other aspects. [1] [2] [3] Scientific evidence on the impact of COVID-19 on out-of-hospital cardiac arrest (OHCA) care is still limited. Some local level studies have been published on the incidence and outcomes of OHCA during the COVID-19 pandemic, specifically in New York, Paris, and the Italian region of Lombardy. [4] [5] [6] [7] However, whilst existing studies focus on heavily hit settings they are limited to defined regions or cities and there are, therefore no national or global studies. This prevents the examination and comparison of areas with significant differences in the impact of the pandemic.

Spain is one of the countries with the highest COVID-19 incidence (455/100,000), mortality (52.5/100,000) and excess mortality (59%) rates worldwide. [8] [9] [10] [11] J o u r n a l P r e -p r o o f On January the 31st 2020 the first positive COVID-19 case was reported in Spain. 12 By March the 3rd there were a total of 150 cases, with the country reaching its daily peak on March the 20th, with 10,786 cases being diagnosed. In response, the government decreed a state of alarm on March the 14th, enforcing strict confinement and other social distancing measures throughout the country.

The objective of the present study is to analyze the influence of the COVID-19 pandemic on OHCA response and survival in Spain, whilst also comparing differences between regions based on their infection incidence. Inclusion criteria: all consecutive OHCA cases in which an emergency team performed resuscitation manoeuvres or post-resuscitation care following cardiopulmonary resuscitation (CPR) attempts by a first responder. Cases were excluded if the emergency team suspended resuscitation on-site due to confirmation of futility criteria during resuscitation. An attempt at CPR was considered futile when new data during resuscitation showed that it was not indicated (terminal disease, prolonged arrest time prior to EMS arrival, "do not resuscitate" orders).

OHSCAR records variables relating to the patient, event, care factors prior to emergency team arrival, treatment carried out by emergency team, final on-site state, hospital treatment and survival. All variables were recorded according to Utstein definitions. 14 Study period and data sources.

The analysis used data collected from two time periods. The COVID-19 period was defined as events between February 1 st and April 30 th 2020. This spanned the first documented infection in Spain, peak growth, and flattening and decline of the incidence curve. The OHSCAR does not collect data continuously. Therefore, the control period Tables S2 and S3 in 

The dependent variables used in analyses were response to OHCA (frequency or incidence where relevant), hospital admission with return of spontaneous circulation (ROSC) and survival to hospital discharge. The independent variable was COVID-19 J o u r n a l P r e -p r o o f which was examined according to the following two different comparisons. The first compared data collected during the non-COVID period with that from the COVID period.

In response to the second objective, we classified regions into low, medium and high incidence groups, according to cumulative COVID-19 incidence tertiles, up-to-date as of April 30 th , 2020.

Descriptive statistics are reported as mean (standard deviation), median (interquartile range) or frequency (percent), where relevant. Between-group comparisons were made for general patient characteristics, events, and pre-and in-hospital care. The Kruskal-Wallis test or ANOVA was used to make comparisons between continuous variables depending on the distribution of the variable under analysis. χ 2 analyses were used for categorical comparisons. All statistical tests were two-tailed with significance set at p < 0.05. The statistical software SPSS version 26.0 was used for all analyses.

In order to analyze whether a change existed in the profile of cases attended to during the study period, the 4 sub-groups recommended by Utstein 14 were compared.

Official population census data for 2018 and 2020 16 were used to calculate resuscitation attempts per 10 5 inhabitants for the non-COVID period and COVID period, respectively.

Given the assumption that accumulated COVID-19 incidence could impact OHCA response over time, changing trends in resuscitation attempts and survival over the 13 weeks between February the 1st and April the 30th 2020 (COVID period) were examined using joinpoint regression. The 13 weeks of the non-COVID period were also examined as a form of control.

Overall survival from OHCA to hospital admission and discharge was examined using logistic regression. Odds ratios for survival per treatment group were adjusted for age and gender. The influence of resuscitation characteristics on survival rates before and during the COVID-19 pandemic, and during NPW and PW, was examined via stratification.

Finally, given the implications of the current COVID-19 pandemic on health resources, a multilevel logistic regression model adjusted for age and gender was developed to examine the influence of COVID-19 incidence on resuscitation attempts and survival, clustered according to regions. Regional analysis was carried out in two ways. In an initial multilevel analysis, the cumulative incidence of each region was used. Subsequently, regions were categorically analyzed according to their aforementioned incidence classification (Table S1 in the Supplementary Appendix).

J o u r n a l P r e -p r o o f

A flowchart of included cases is shown in Figure 1 . There were more futile resuscitations in the COVID period in comparison with the non-COVID period (9.8% vs 6.4%; p < 0.001).

A total of 3169 resuscitation attempts were made between February 1 st and April 30 th in the two periods studied. More cases were attended to during the non-COVID period (n = 1723) than the COVID period (n = 1446; p < 0.001). The number of OHCAs resuscitation attempts performed during the non-COVID period and COVID period are shown in Figure 2 , together with the evolution of the incidence of the COVID-19

pandemic. Further, no significant change in resuscitation trends occurred over time during the non-COVID period but a significant drop in resuscitation attempts was seen from week 4 onwards during the COVID period (t = -2.52; SE = 0.06; p < 0.05) (Figure 3, 3A) .

For the comparison of non-pandemic weeks and pandemic weeks, a sample of 2275 resuscitation attempts was used. More resuscitation attempts were made during nonpandemic weeks (n=1652; standardized incidence = 0.50) than pandemic weeks (n = 623; standardized incidence = 0.30; p < 0.001) ( Table 1) .

Participant and OHCA characteristics and differences between the non-COVID period and COVID period, and NPW and PW are presented in table 1. No baseline gender differences were found, with average representation across time-periods being 70.6% male and 29.4% female. Patients from the COVID period were significantly younger than those from the non-COVID period. Pandemic week patients were also younger than nonpandemic weeks patients, with fewer patients aged 75 and older being treated during PW.

Relative to the non-COVID period, OHCA was more likely to occur at home and bystander CPR was less likely during the COVID period, with these differences remaining even when cardiac arrest was witnessed. More OHCAs also occurred at home during PW than NPW. The time-interval between call placement and ET arrival was longer during PW than NPW, with ambulances also being less likely to arrive within 8 and 15 minutes, respectively.

Airway isolation was less often performed during the COVID period than the non-COVID period, with an increase in the use of supraglottic devices and a decline in orotracheal intubation. This same pattern was seen when comparing non-pandemic weeks and pandemic weeks. Only one difference emerged in the profile of cases treated regarding Utstein subgroups with more OHCAs being witnessed by emergency team during pandemic weeks than non-pandemic weeks.

No differences regarding in-hospital care, percutaneous coronary intervention, hypothermia treatment, implantable cardioverter-defibrillator implant or thrombolysis treatment use were found between any of the groups Survival following OHCA Odd ratios, confidence intervals and p-values are provided in Table 2 . Odds of ROSC upon hospital admission were more favourable in the non-COVID period than the COVID period and in non-pandemic weeks than in pandemic weeks. These differences remained when only arrests witnessed by emergency teams or by a bystander but without a shockable initial rhythm were considered. Survival to discharge was more likely in the non-COVID period than the COVID period both overall for all patients and when cardiac arrest was witnessed by ET. Relative to pandemic weeks, survival to hospital discharge was more J o u r n a l P r e -p r o o f likely in non-pandemic weeks when OHCA was witnessed and when no shockable initial rhythm was present.

Further, no change in survival trends at hospital admission and at hospital discharge when comparing Covid period versus Non-Covid period (figure 3B and 3 C respectively).

Regional comparisons of all participant and OHCA characteristics are presented in table 3 with few meaningful differences emerging.

The multilevel model pertaining to accumulated COVID-19 incidence showed that higher incidence within different regions overall (Z = 2.05; p = 0.040; SE = 0.33) led to fewer resuscitation attempts being made (B = -0.004; t = -55.84; p = 0.000). Thus, another model was developed with the aim of identifying whether there were differential effects within regions with low, medium and high Covid-19 incidence. This model showed that all three COVID-19 incidence groups (low, medium and high) predicted fewer resuscitation attempts relative to the absence of COVID-19, however, similar coefficients were seen for each of the three groups (low: B = -2.51; p = 0.000; medium: B = -2.45; p = 0.000; high: B = -2.51; p = 0.000). Further, overall regional effects were no longer significant.

Resuscitation attempts and survival according to low, medium and high COVID-19 incidence regions are presented in Figures 

With regards to the profile of treated patients, they were somewhat younger and fewer resuscitations were performed within the over 75s age group during PW. No meaningful changes were found according to Utstein subgroups.

A change in attitudes of emergency teams is also suggested regarding airway management, with a clear increase in the use of supraglottic devices in accordance with guidelines laid out by leading scientific societies. [19] [20] [21] Similarly, during the pandemic, emergency teams did not include cases in asystolic organ donation programs.

The final results offer cause for concern when comparing both periods. Hospital admission with ROSC fell to 22.0 % during the COVID period, a decrease from Non-COVID period levels of 32.1%. Survival at discharge was also found to be at just 7.5%, declining from 9.8% prior to the pandemic. Similar outcomes emerged when comparing PW and NPW, with drops of more than 10% in ROSC at hospital and 2% in survival at discharge, although differences in survival at hospital discharge were not significant.

The present results are expected given the importance of the location of CPR at home 21,22 and adequate execution of the first links in the chain of survival 23 . Indeed, survival in patient groups with shockable initial rhythm or CPR witnessed by emergency team members were similar. This was not the case in those who were bystander witnessed and without a shockable initial rhythm, with this sub-group demonstrating much worse survival rates during pandemic weeks. This group accounts for a large number of patients, approximately half of all cases, has a worse overall prognosis and is even more susceptible to the timing of intervention and bystander CPR. Actual survival achieved within this group, 1%, would lead it to be considered a futile action if it did not incorporate so many cases 24 .

The most relevant aspect of our study is the influence that the pandemic has had on all participating communities, regardless of infection level. A significant fall is seen in the number of resuscitation attempts and in successful outcomes. Significant differences did not always emerge, however, clear trends did emerge and the lack of significance was likely due to loss of statistical power due to the decline in patients treated. In some instances, greater detriment was seen in communities with low and medium infection rates.

This suggests that the presence of the disease per se, rather than its intensity, in addition to the strict social distancing measures adopted have dictated this negative influence on health care.

Beyond the formal aspects of each of the examined variables, the COVID-19 pandemic has, overall, led to important set-backs in the main determinants of survival following OHCA. Key initiatives are required to increase the number of resuscitation attempts. Citizen engagement to promote immediate initiation of resuscitation, public access to early defibrillation and quick response capacity of emergency services have all been significantly affected, regardless of the level of infection. OHCA treatment during times of a health pandemic must reconsider specific measures for each patient, alongside the most appropriate health and social strategies for this public health issue. Studies and close monitoring will be needed to see if this impact remains over time and successful strategies must be modified 26,27 .

The COVID-19 pandemic has also had a negative impact on processes favouring a network approach, in other words, processes running from the pre-hospital stage to inhospital treatment. Response networks for ST-elevation myocardial infarction and stroke have seen a decline in the number of patients attended and a worsening of the outcomes obtained for response times and complications [28] [29] [30] . Since there was no difference in hospital treatment, despite the disparity in hospital resource saturation in some of the J o u r n a l P r e -p r o o f regions, the present data makes it apparent that the COVID-19 pandemic has had a particularly negative influence on pre-hospital care.

Although other data are not yet available to verify these effects on emergency services in other countries and in other health models, our findings call for the need to adapt out-of-hospital care for citizens and health professionals during periods of serious infection risk.

Our results reflect experiences and implications of the COVID-19 pandemic in relation to OHCA in a country with a specific physician-led pre-hospital care model. It would be enlightening to collect data from other countries with other care models in relation to OHCA. OHSCAR registers reanimation attempts for OHCA and so we cannot know, as has been reported in other registers, know whether the number of OHCAs for which it was decided not to initiate resuscitation manoeuvres increased. We are also unable to establish the reasons for stopping advanced life support when cases were considered futile. The data available within OSCHAR precluded us from a direct comparison with a pre-COVID period. Finally, it is possible that with a larger sample the negative trend in survival at discharge would have been significant. 

Rapid assessment of service delivery for NCDs during the COVID-19 pandemic

The untold toll-the pandemic's effects on patients without Covid-19

The Covid-19 Pandemic and the Incidence of Acute Myocardial Infarction

Out-of-hospital cardiac arrest during the Covid-19-19 outbreak in Italy

Out-of-hospital cardiac arrest during the COVID-19-19 pandemic in Paris, France: a population-based, observational study

COVID-19-19 kills at home: the close relationship between the epidemic and the increase of out-of-hospital cardiac arrests

Characteristics Associated With Out-of-Hospital Cardiac Arrests and Resuscitations During the Novel Coronavirus Disease

COVID-19-19 map

Impact of bystander-focused public health interventions on cardiopulmonary resuscitation and survival: a cohort study

Reduction of hospitalizations for myocardial infarction in Italy in the COVID-19 era

Collateral Effect of Covid-19 on Stroke Evaluation in the United States

Acute Stroke Care Is at Risk in the Era of COVID-19: Experience at a Comprehensive Stroke Center in Barcelona