Abstract
Background:
Recent reports on challenges in resuscitation care at hospitals severely affected by the novel coronavirus disease 2019 (COVID-19) pandemic raise questions about how the pandemic affected outcomes for in-hospital cardiac arrest (IHCA) throughout the U.S.
Methods:
Within Get-With-The-Guidelines-Resuscitation, we conducted a retrospective cohort study to compare IHCA survival during the pre-surge (January 1-February 29), surge (March 1-May 15) and immediate post-surge (May 16-June 30) periods in 2020 compared to 2015–2019. Monthly COVID-19 mortality rates for each hospital’s county were categorized, per 1,000,000 residents, as low (0–10), moderate (11–50), high (51–100), or very high (>100). Using hierarchical regression models, we compared rates of survival to discharge in 2020 vs. 2015–2019 for each period.
Results:
Of 61,586 IHCAs, 21,208 (4309 in 2020), 26,459 (5949 in 2020), and 13,919 (2686 in 2020) occurred in the pre-surge, surge, and post-surge periods, respectively. During the pre-surge period, 24.2% survived to discharge in 2020 vs. 24.7% in 2015–2019 (adjusted OR, 1.12 [95% CI: 1.02–1.22]). In contrast, during the surge period, 19.6% survived to discharge in 2020 vs. 26.0% in 2015–2019 (adjusted OR, 0.81 [0.75–0.88]). Lower survival was most pronounced in communities with high (28% lower survival) and very high (42% lower survival) monthly COVID-19 mortality rates (interaction P-value<0.001). Resuscitation times were shorter (median: 22 vs. 25 minutes; P<0.001) and delayed epinephrine treatment was more prevalent (11.3% vs. 9.9%; P=0.004) during the surge period. Survival was lower even when patients with confirmed/suspected COVID-19 infection were excluded from analyses. During the post-surge period, survival rates were similar in 2020 vs. 2015–2019 (22.3% vs. 25.8%; adjusted OR, 0.93 [0.83–1.04]), including communities with high COVID-19 mortality (interaction P-value=0.16).
Conclusions:
Early during the pandemic, rates of survival to discharge for IHCA decreased, even among patients without COVID-19 infection, highlighting the early impact of the COVID-19 pandemic on in-hospital resuscitation.
Keywords: in-hospital cardiac arrest, COVID-19, survival
INTRODUCTION
In the wake of the novel coronavirus 2019 (COVID-19) pandemic, hospitals may have modified resuscitation response to protect healthcare staff. New guidance recommending donning personal protective equipment prior to entering patient rooms, avoiding bag-mask ventilation prior to intubation to reduce aerosol generation, and limiting the number of responders in a patient’s room,1 may have been implemented broadly irrespective of patients’ underlying COVID-19 infection status. Yet, the extent to which the COVID-19 pandemic has affected processes-of-care and survival outcomes for all patients with an in-hospital cardiac arrest (IHCA), including those without COVID-19 infection, and whether these changes were sustained beyond the initial pandemic months, is unknown and critical to understand as COVID-19 will likely be endemic in the U.S. with recurrent surges over time.
To date, there have been a handful of single-center and registry reports of IHCA in patients with active COVID-19 infection, with variable survival rates ranging from 0% to 41%.2–6 However, these studies have generally not compared IHCA survival with prior years or examined whether lower survival was accompanied by delays in initiation of cardiopulmonary resuscitation (CPR), defibrillation or epinephrine treatment or shorter duration of resuscitative efforts. Moreover, if overall survival for IHCA decreased during the pandemic, a number of key questions remain. First, was lower IHCA survival only seen during the initial months of the COVID-19 pandemic or were these differences sustained over time? Second, was the association of the pandemic on IHCA survival greater in communities with a higher COVID-19 disease burden? Finally, if IHCA survival was lower during the pandemic months of 2020, was survival also lower in patients without COVID-19 infection, suggesting a wide-reaching effect of the pandemic on in-hospital resuscitation outcomes?
To address these gaps in knowledge, we linked IHCA data from a national registry to county-level COVID-19 mortality data in the U.S to provide a more comprehensive assessment of the association of the COVID-19 pandemic on IHCA outcomes. We examined whether IHCA survival was lower during the initial surge months as well as during the immediate post-surge period and if such decreases were confined only to communities with high COVID-19 mortality rates. We compared rates of delayed CPR, defibrillation and epinephrine treatment, as well as duration of resuscitation efforts, during the pandemic with prior years and evaluated whether IHCA survival during the pandemic was lower even in patients without COVID-19 infection.
METHODS
The study was approved by Saint Luke’s Hospital’s institutional review board, which waived the requirement for informed consent as the study involved deidentified data. The data and analytical code that support the findings of this study are available from the GWTG-Resuscitation registry and the authors, respectively, upon approval from the registry’s Research and Publications committee.
Main Data Source
Get With The Guidelines (GWTG)-Resuscitation® is a large, prospective, national quality-improvement registry of IHCA. Its design has been previously described.7 In brief, trained hospital personnel identify all patients without do-not-resuscitate orders with a cardiac arrest (defined as absence of a palpable central pulse, apnea, and unresponsiveness) who undergo cardiopulmonary resuscitation. Cases are identified through multiple methods, including centralized collection of cardiac arrest flow sheets, reviews of hospital paging system logs, and routine checks of code carts and pharmacy tracer drug records.7 Standardized Utstein-style definitions are used for all patient variables and outcomes to facilitate uniform reporting across hospitals.8, 9 Data accuracy in GWTG-Resuscitation is supported by certification of research staff and use of standardized software with data checks for completeness and accuracy.
Study Population
This retrospective cohort study was conducted after a special mid-year 2020 data extraction performed in September of 2020 to examine resuscitation practices and outcomes during COVID-19. Therefore, we included patients enrolled in GWTG-Resuscitation between January 1 to June 30 during the years of 2015–2020. We restricted the comparison years to 2015–2019, given no temporal change in IHCA survival during these five years (Supplemental Figure S1). There were 75,210 patients with IHCA at hospitals with at least 5 cases during 2020 and the comparison period of 2015–2019. We excluded 9852 patients from hospitals which did not have at least one IHCA case in 2020 in each of the pre-surge, surge, and post-surge periods (see below). We further excluded 3772 patients at hospitals for which we could not link to county-level COVID-19 mortality data. Our final cohort comprised 61,586 patients with IHCA from 178 hospitals. Notably, the 13,624 patients excluded from the last two exclusions were similar in most characteristics as those in the study cohort (Supplemental Table S1).
Independent Variable and Study Outcomes
The primary outcome was survival to hospital discharge. The secondary outcome was sustained return of spontaneous circulation for at least 20 minutes (ROSC). To place survival findings into a clinical context, we evaluated delays from time of IHCA detection for three resuscitation process-of-care quality measures: initiation of cardiopulmonary resuscitation >1 minute, defibrillation time >2 minutes for shockable IHCAs, and epinephrine administration >5 minutes for non-shockable IHCAs. We also evaluated two resuscitation practices: duration of resuscitative efforts in those without ROSC and rates of do-not-attempt resuscitation (DNAR) orders within 24 hours in those with ROSC to identify changes in resuscitation practice during the pandemic. In prior work, we have shown that resuscitation time in those without ROSC is a surrogate measure of resuscitation effort intensity, with higher resuscitation time associated with higher rates of survival to discharge.10
The independent variable was time period of arrest: 2020 vs. control time period of 2015–2019. We divided the months of January to June into three COVID-19 comparison periods: pre-surge (January 1 to February 28 or 29); surge (March 1 to May 15); and post-surge (May 16 to June 30). The post-surge period was defined based on when COVID-19 case and death rates initially abated in 2020 and when the U.S. relaxed ‘stay-at-home’ orders. For this study, we focused on the comparison of IHCA outcomes during the surge and post-surge periods in 2020 vs. 2015–2019 but used the pre-surge period comparison as an additional control.
Besides an overall comparison of IHCA outcomes between 2020 and 2015–2019 during the surge and post-surge periods, we also examined whether differences in IHCA survival during the pandemic were confined to communities with high COVID-19 mortality rates. Daily county-level COVID-19 mortality data were obtained from publicly available sources.11 For each county, we calculated the monthly COVID-19 mortality rate per 1,000,000 residents by dividing the total number of deaths due to COVID-19 occurring every month by the number of residents in the county. Each GWTG-Resuscitation hospital was geocoded to a U.S. county based on its zip code location through a crosswalk file from the Department of Housing and Urban Development, which links each zip code to a U.S. county.12 In this manner, the county-level COVID-19 mortality rate for patients at each hospital during each time period was determined.
Statistical Analysis
Given the large sample size, baseline differences for patients with IHCA during the pre-surge, surge, and post-surge periods in 2020 vs. the same time period in 2015–2019 were compared using standardized differences. Standardized differences of >10% were considered clinically important.13
To examine whether rates of survival to discharge were different in 2020 vs. 2015–2019 during each of the three time periods, we constructed separate multivariable hierarchical logistic regression models, with hospital as a random effect to account for clustering of patients within hospitals.14, 15 Models adjusted for age (as continuous variable), sex, self-identified race by patients or families (categorized as white, black, and other), initial cardiac arrest rhythm (asystole, pulseless electrical activity, ventricular fibrillation, pulseless ventricular tachycardia), location of cardiac arrest, illness category (medical cardiac, medical non-cardiac, surgical cardiac, surgical non-cardiac, other), comorbid conditions (prior heart failure or myocardial infarction, index admission heart failure or myocardial infarction, diabetes mellitus, baseline depression in central nervous system function, acute stroke, pneumonia, and metastatic or hematologic malignancy), medical conditions present within 24 hours of cardiac arrest (renal insufficiency, hepatic insufficiency, respiratory insufficiency, hypotension, septicemia, and metabolic or electrolyte abnormality) and interventions in place at the time of cardiac arrest (continuous intravenous vasopressor, assisted or mechanical ventilation, and hemodialysis). Model also adjusted for the time of day (work hours [7:00 am to 10:59 pm] vs. after hours [11:00 pm to 6:59 am]), and day of the week (weekday vs. weekend) of the IHCA.16
For models during the surge and post-surge periods, we then included an interaction term between year period (2020 vs. 2015–2019) and the hospital’s county-level monthly COVID-19 mortality rate to determine if rates of survival to discharge varied by the severity of the COVID-19 pandemic. Rates of COVID-19 mortality were calculated per 1,000,000 residents per month and categorized as low (0–10 COVID-19 deaths), moderate (11–50), high (51–100), or very high (>100). We repeated all analyses for the secondary outcome of ROSC. Additionally, we constructed separate regression models with each aforementioned process-of-care measure and resuscitation practice as the dependent variable and examined whether rates of these measures differed during the surge and post-surge periods by the severity of the COVID-19 pandemic by evaluating an interaction term between time period and COVID-19 mortality strata. Finally, as the variable ‘suspected or confirmed COVID-19 status’ was added to the GWTG-Resuscitation registry in late March of 2020, we excluded patients with confirmed/suspected COVID-19 infection and repeated all analyses above for the surge and pre-surge periods to assess whether changes in IHCA survival were seen in patients not suspected to have COVID-19 infection.
For each analysis, the null hypothesis was evaluated at a 2-sided significance level of 0.05 and calculated 95% confidence intervals (CIs) using robust standard errors. All statistical analyses were conducted using SAS Version 9.1.3 (SAS Institute, Cary, NC) and R Version 3.6.2 (Free Software Foundation, Boston, MA).
RESULTS
Of 61,586 IHCAs between 2015–2020, 21,208 arrests (4309 in 2020) occurred during the pre-surge period of January to February, 26,459 (5949 in 2020) in the surge period of March 1 to May 15, and 13,919 (2686 in 2020) in the post-surge period of May 16 to June 30. Supplemental Table S2 compares baseline characteristics of IHCAs during 2020 vs. 2015–2019 for each of these three periods. Generally, during the pre-surge period, IHCAs in 2020 had similar characteristics as those in 2015–2019 except for higher rates of hypotension and use of continuous intravenous vasopressors at the time of IHCA. During the surge and post-surge periods, patients with IHCAs were more likely to be of Black race; have an initial non-shockable cardiac arrest rhythm; and have hypotension, a non-cardiac medical admitting diagnosis, and pneumonia prior to arrest. Moreover, during the surge period, IHCAs during 2020 vs. 2015–2019 were more likely to be from the Northeast region of the U.S. and be on mechanical ventilation and continuous vasopressors at the time of arrest. Notably, 1505 of 5949 (25.3%) patients in the 2020 surge period and 418 of 2686 (15.6%) patients in the 2020 post-surge period were confirmed or suspected to have COVID-19 infection at the time of IHCA.
Survival Outcomes
Between January to June of 2020 in the entire U.S., there were 127,462 probable or confirmed deaths from COVID-19,11 with 1 death in the pre-surge period, 87,498 in the surge period, and 39,963 in the post surge period (Figure 1a), which translates to 0.002, 105.2, and 80.1 COVID-19 deaths per 1,000,000 residents per month in the three periods, respectively, based on U.S. Census estimates of 332.6 million Americans in 2020.17 The trajectory of survival rates to discharge for IHCA in our study cohort from January to June in 2020 vs. 2015–2019 is depicted in Figure 1b. During the pre-surge period (January and February), rates of survival to discharge were 24.2% in 2020 and 24.7% in 2015–2019. After multivariable adjustment, overall survival was higher during the pre-surge period in 2020 vs. 2015–2019 (adjusted Odds Ratio [OR], 1.12 [95% Confidence Interval [CI]: 1.02. 1.22). Rates of ROSC during the pre-surge period in 2020 were similar to those in 2015–2019 (67.7% vs. 66.6%; adjusted OR, 1.08 [95% CI: 0.99, 1.16). Notably, delays in CPR and epinephrine administration were less prevalent during the pre-surge period, whereas delays in defibrillation, duration of resuscitation in those without ROSC, and DNAR rates within 24 hours of ROSC were similar (Table 1).
Figure 1. COVID-19 deaths in the U.S. Between January to June of 2020 (Figure 1A) and Crude Rates of Survival to Discharge for In-Hospital Cardiac Arrest During the First 6 Months of 2020 vs. 2015–2019 (Figure 1B).
Data are shown per half month period.
Data points for Figure 1A are derived from https://github.com/nytimes/covid-19-data/blob/master/us-counties.csv11, and data for Figure 1B are from the GWTG-Resuscitation study cohort.
Table 1.
Survival Outcomes, Processes of Care, and Resuscitation Practices During the Pre-Surge, Surge, and Post-Surge Periods.
| Pre-Surge | Surge | Post-Surge | |||||||
|---|---|---|---|---|---|---|---|---|---|
| 2015–2019 | 2020 | Adjusted OR† (95% CI) | 2015–2019 | 2020 | Adjusted OR† (95% CI) | 2015–2019 | 2020 | Adjusted OR† (95% CI) | |
| Survival Outcomes * | |||||||||
|
| |||||||||
| Survival to Discharge | 24.7% | 24.2% | 1.12 (1.02, 1.22) |
26.0% | 19.6% | 0.81 (0.75, 0.88) |
25.8% | 22.3% | 0.93 (0.83, 1.04) |
| ROSC | 66.6% | 67.7% | 1.08 (0.99, 1.16) |
67.4% | 59.4% | 0.73 (0.69, 0.78) |
67.4% | 65.0% | 0.93 (0.85, 1.02) |
|
| |||||||||
| Processes-of-Care and Practices | P value | P value | P value | ||||||
|
| |||||||||
| Delayed Defibrillation >2’ | 35.5% | 36.3% | 0.69 | 33.4% | 31.5% | 0.30 | 34.4% | 32.7% | 0.53 |
| Delayed Epinephrine >5’ | 9.8% | 8.5% | 0.02 | 9.9% | 11.3% | 0.004 | 10.1% | 9.8% | 0.70 |
| Delayed Initiation of CPR >1’ | 2.5% | 1.8% | 0.01 | 2.6% | 2.1% | 0.06 | 2.7% | 2.3% | 0.19 |
| Resuscitation Time, median (IQR)‡ | 25 (16, 36) |
24 (16, 36) |
0.84 | 25 (16, 37) |
22 (13, 33) |
<0.001 | 25 (16, 36) |
23 (14, 35) |
0.07 |
| % DNAR in those with ROSC | 24.4% | 25.7% | 0.13 | 24.6% | 26.0% | 0.08 | 24.2% | 28.4% | <0.001 |
Survival outcomes were compared using hierarchical logistic regression models, adjusted for the variables listed in Supplemental Table II
The adjusted ORs represent the odds of survival in 2020 during the pre-surge, surge and post-surge periods as compared with corresponding periods in 2015–2019
Resuscitation times are for patients who did not achieve ROSC (i.e., died during the resuscitation)
Abbreviations: CPR, cardiopulmonary resuscitation; DNAR, do-not-attempt resuscitation orders; IQR, imter-quartile range; OR, odds ratio; ROSC, return of spontaneous circulation
In contrast, during the pandemic surge between March 1 to May 15, overall rates of survival to discharge were lower in 2020 vs. 2015–2019 even after accounting for patients’ higher illness severity (19.6% vs. 26.0%; adjusted OR, 0.81 [95% CI: 0.75, 0.88]). A significant interaction (P<0.001) was found between year and county-level COVID-19 mortality rate, with patients in communities with high (51–100 deaths per 1,000,000) and very high (>100 deaths per 1,000,000) monthly COVID-19 mortality rates experiencing, respectively, 28% and 42% lower survival during the surge period in 2020 as compared with the same time period in 2015–2019 (Table 2). For the secondary outcome of ROSC, rates were also lower during the surge period in 2020 vs. 2015–2019 (59.4% vs. 67.4%; adjusted OR, 0.73 [95 CI: 0.69, 0.78]). Although most pronounced in communities with very high monthly COVID-19 mortality, rates of ROSC were significantly lower in all communities, even in those with low COVID-19 mortality rates (Table 2). The median duration of resuscitation efforts in those without ROSC was shorter during the surge period (22 minutes in 2020 vs. 25 minutes in 2015–2019; P<0.001), and this difference was most pronounced in communities with very high (>100 deaths per 1,000,000) monthly COVID-19 mortality, where a decrease of nearly eight minutes in resuscitation duration was noted (interaction p-value <0.001; (Table 3). Additionally, delays in epinephrine treatment were more prevalent during the surge period in 2020 vs. the same period in 2015–2019, whereas rates of delayed defibrillation and CPR initiation, as well as DNAR within 24 hours of ROSC, were similar (see Table 1). For these measures, there was no significant interaction between year and COVID-19 mortality rate (Supplemental Table S3).
Table 2. Rates of Survival to Discharge and ROSC During the 2020 Surge Period vs. 2015–2019.
Unadjusted and model-adjusted comparison of survival outcomes for IHCA during the surge period of March 1 to May 15, 2020 vs. the same time period in 2015–2019. Results are shown for the overall cohort and when stratified by the county-level COVID-19 mortality rate.
| 2015–2019 | 2020 | Adjusted OR | Interaction | ||
|---|---|---|---|---|---|
| n = 20,510 | n = 5949 | (95% CI) | P | P | |
|
SURVIVAL TO DISCHARGE
| |||||
| Overall | 5342/20,510 (26.0%) | 1164/5949 (19.6%) | 0.81 (0.75, 0.88) | <0.001 | |
| Monthly COVID-19 Mortality Rate | |||||
| 0–10 per 1,000,000 | 1459/5794 (25.2%) | 353/1493 (23.6%) | 1.04 (0.90, 1.20) | 0.61 | <0.001 |
| 11–50 per 1,000,000 | 1654/6172 (26.8%) | 344/1502 (22.9%) | 0.97 (0.84, 1.12) | 0.67 | |
| 51–100 per 1,000,000 | 1130/3834 (29.5%) | 217/1060 (20.5%) | 0.72 (0.60, 0.86) | <0.001 | |
| >100 per 1,000,000 | 1099/4710 (23.3%) | 250/1894 (13.2%) | 0.58 (0.49, 0.68) | <0.001 | |
|
| |||||
|
RETURN OF SPONTANEOUS CIRCULATION
| |||||
| Overall | 13,833/20,510 (67.4%) | 3534/5949 (59.4%) | 0.73 (0.69, 0.78) | <0.001 | |
| Monthly COVID-19 Mortality Rate | |||||
| 0–10 per 1,000,000 | 3928/5794 (67.8%) | 931/1493 (62.4%) | 0.81 (0.71, 0.91) | 0.001 | <0.001 |
| 11–50 per 1,000,000 | 4166/6172 (67.5%) | 951/1502 (63.3%) | 0.84 (0.74, 0.95) | 0.005 | |
| 51–100 per 1,000,000 | 2624/3834 (68.4%) | 674/1060 (63.6%) | 0.83 (0.72, 0.97) | 0.02 | |
| >100 per 1,000,000 | 3115/4710 (66.1%) | 978/1894 (51.6%) | 0.56 (0.50, 0.63) | <0.001 | |
Abbreviations: OR, odds ratio; ROSC, return of spontaneous circulation
Table 3. Duration of Resuscitation Efforts During the Surge Period in 2020 vs. 2015–2019.
During the surge period in 2020, duration of resuscitative efforts in those who did not achieve return of spontaneous circulation was shorter vs. the same time period in 2015–2019. Decreases in resuscitation time were most pronounced in communities with high (51–100 per 1,000,000) and very high (>100 per 1,000,000) monthly COVID-19 mortality rates.
| 2015–2019 | 2020 | P value | |
|---|---|---|---|
| Resuscitation Time, minutes median (IQR) | 25 (16, 37) | 22 (13, 33) | <0.001 |
|
| |||
| Monthly COVID-19 Mortality Rate | Difference in Time of Resuscitation, minutes 2020 vs. 2015–2019 | P | P for interaction |
|
| |||
| 0–10 per 1,000,000 | −0.8 (−2.6, 0.9) | 0.36 | <0.001 |
| 11–50 per 1,000,000 | −0.5 (−2.3, 1.3) | 0.57 | |
| 51–100 per 1,000,000 | −2.0 (−4.2, 0.7) | 0.059 | |
| >100 per 1,000,000 | −7.8 (−9.3, −6.3) | <0.001 | |
Abbreviation: IQR, inter-quartile range
During the post-surge period between May 16 to June 30, rates of survival to discharge were not significantly different in 2020 vs. 2015–2019 (22.3% vs. 25.8%; adjusted OR, 0.93 [95% CI: 0.83, 1.04]), and rates were not significantly different in communities with high COVID-19 mortality rates (interaction P-value of 0.16; Table 4). Rates of ROSC were also not different in 2020 vs. 2015–2019 (65.0% vs. 67.4%; adjusted OR, 0.93 [95 CI: 0.0.85, 0.1.02]), overall and when stratified by a community’s COVID-19 mortality rate. Rates of delayed treatment and resuscitation times in those without ROSC were similar during the post-surge period in 2020 vs. 2015–2019, overall and across COVID-19 mortality strata, whereas rates of DNAR within 24 hours in those with ROSC increased (see Table 1 and Supplemental Table S3).
Table 4. Rates of Survival to Discharge and ROSC During the 2020 Post-Surge Period vs. 2015–2019.
Unadjusted and model-adjusted comparison of survival outcomes for IHCA during the post-surge period of May 16 to June 30, 2020 vs. the same time period in 2015–2019. Results are shown for the overall cohort and when stratified by the county-level COVID-19 mortality rate.
| 2015–2019 | 2020 | Interaction | |||
|---|---|---|---|---|---|
| n = 11,233 | n = 2686 | OR (95% CI) | P | P | |
|
SURVIVAL TO DISCHARGE
| |||||
| Overall | 2898/11,233 (25.8%) | 599/2686 (22.3%) | 0.93 (0.83, 1.04) | 0.18 | |
| Monthly COVID-19 Mortality Rate | |||||
| 0–10 per 1,000,000 | 600/2577 (23.3%) | 150/612 (24.5%) | 1.16 (0.93, 1.46) | 0.19 | 0.16 |
| 11–50 per 1,000,000 | 920/3355 (27.4%) | 177/750 (23.6%) | 0.91 (0.74, 1.11) | 0.35 | |
| 51–100 per 1,000,000 | 583/2296 (25.4%) | 108/527 (20.5%) | 0.84 (0.65, 1.09) | 0.19 | |
| >100 per 1,000,000 | 795/3005 (26.5%) | 164/797 (20.6%) | 0.85 (0.69, 1.04) | 0.12 | |
|
| |||||
|
RETURN OF SPONTANEOUS CIRCULATION
| |||||
| Overall | 7573/11,233 (67.4%) | 1745/2686 (65.0%) | 0.93 (0.85, 1.02) | 0.12 | |
| Monthly COVID-19 Mortality Rate | |||||
| 0–10 per 1,000,000 | 1750/2577 (67.9%) | 396/612 (64.7%) | 0.89 (0.74, 1.08) | 0.24 | 0.72 |
| 11–50 per 1,000,000 | 2256/3355 (67.2%) | 498/750 (66.4%) | 0.98 (0.83, 1.17) | 0.85 | |
| 51–100 per 1,000,000 | 1560/2296 (67.9%) | 356/527 (67.6%) | 0.99 (0.81, 1.22) | 0.94 | |
| >100 per 1,000,000 | 2007/3005 (66.8%) | 495/797 (62.1%) | 0.88 (0.75, 1.04) | 0.14 | |
Abbreviations: OR, odds ratio; ROSC, return of spontaneous circulation
When patients with confirmed or suspected active COVID-19 infection were excluded from the analyses, rates of survival to discharge remained lower during the surge period in 2020 vs. 2015–2019, particularly for patients in communities with high and very high monthly COVID-19 mortality rates (interaction p-value of 0.004; Table 5). Patterns for ROSC paralleled the main study findings during the surge period, with lower ROSC rates for patients in all communities, including those with low monthly COVID-19 mortality rates. Findings for survival to discharge and ROSC were similar in the post-surge period when patients with confirmed or suspected COVID-19 infection were excluded.
Table 5. Survival Outcomes During the 2020 Surge and Post-Surge Periods vs. 2015–2019 for Patients Without COVID-19 Infection.
Survival outcomes for IHCA during the surge and post-surge periods in 2020 vs. 2015–2019 are shown when patients with confirmed or suspected COVID-19 infection are excluded from the 2020 cohort.
| Surge Period | 2015–2019 | 2020 | Adjusted OR | Interaction | |
|---|---|---|---|---|---|
| n = 20,510 | n = 4444 | (95% CI) | P | P | |
|
SURVIVAL TO DISCHARGE
| |||||
| Overall | 5342/20,510 (26.0%) | 991/4444 (22.3%) | 0.90 (0.82, 0.98) | 0.01 | |
|
| |||||
| Monthly COVID-19 Mortality Rate | |||||
| 0–10 per 1,000,000 | 1459/5794 (25.2%) | 315/1310 (24.1%) | 1.05 (0.90, 1.22) | 0.57 | 0.004 |
| 11–50 per 1,000,000 | 1654/6172 (26.8%) | 306/1300 (23.5%) | 0.96 (0.83, 1.12) | 0.62 | |
| 51–100 per 1,000,000 | 1130/3834 (29.5%) | 172/822 (20.9%) | 0.67 (0.55, .82) | <0.001 | |
| >100 per 1,000,000 | 1099/4710 (23.3%) | 198/1012 (19.6%) | 0.83 (0.69, 1.01) | 0.058 | |
|
| |||||
|
RETURN OF SPONTANEOUS CIRCULATION
| |||||
| Overall | 13,833/20,510 (67.4%) | 2770/4444 (62.3%) | 0.82 (0.76, 0.88) | <0.001 | |
|
| |||||
| Monthly COVID-19 Mortality Rate | |||||
| 0–10 per 1,000,000 | 3928/5794 (67.8%) | 822/1310 (62.8%) | 0.82 (0.72, 0.93) | 0.002 | 0.34 |
| 11–50 per 1,000,000 | 4166/6172 (67.5%) | 824/1300 (63.4%) | 0.85 (0.74, 0.96) | 0.01 | |
| 51–100 per 1,000,000 | 2624/3834 (68.4%) | 535/822 (65.1%) | 0.88 (0.75, 1.03) | 0.12 | |
| >100 per 1,000,000 | 3115/4710 (66.1%) | 589/1012 (58.2%) | 0.73 (0.63, 0.84) | <0.001 | |
|
| |||||
| Post-Surge Period | 2015–2019 | 2020 | Interaction | ||
| n = 11,233 | n = 2288 | OR (95% CI) | P | P | |
|
| |||||
|
SURVIVAL TO DISCHARGE
| |||||
| Overall | 2898/11,233 (25.8%) | 530/2288 (23.2%) | 0.96 (0.85, 1.07) | 0.45 | |
| Monthly COVID-19 Mortality Rate | |||||
| 0–10 per 1,000,000 | 600/2577 (23.3%) | 137/577 (24.6%) | 1.18 (0.93, 1.48) | 0.17 | 0.25 |
| 11–50 per 1,000,000 | 920/3355 (27.4%) | 153/623 (24.6%) | 0.91 (0.73, 1.14) | 0.41 | |
| 51–100 per 1,000,000 | 583/2296 (25.4%) | 101/434 (23.3%) | 0.93 (0.72, 1.22) | 0.61 | |
| >100 per 1,000,000 | 795/3005 (26.5%) | 139/654 (21.3%) | 0.86 (0.69, 1.08) | 0.19 | |
|
| |||||
|
RETURN OF SPONTANEOUS CIRCULATION
| |||||
| Overall | 7573/11,233 (67.4%) | 1496/2288 (65.4%) | 0.98 (0.89, 1.08) | 0.65 | |
| Monthly COVID-19 Mortality Rate | |||||
| 0–10 per 1,000,000 | 1750/2577 (67.9%) | 365/577 (65.5%) | 0.92 (0.76, 1.13) | 0.45 | 0.63 |
| 11–50 per 1,000,000 | 2256/3355 (67.2%) | 417/623 (66.9%) | 1.02 (0.84, 1.23) | 0.82 | |
| 51–100 per 1,000,000 | 1560/2296 (67.9%) | 302/434 (69.6%) | 1.09 (0.87, 1.37) | 0.46 | |
| >100 per 1,000,000 | 2007/3005 (66.8%) | 412/654 (63.0%) | 0.92 (0.77, 1.11) | 0.39 | |
DISCUSSION
Understanding the impact of the COVID-19 pandemic on IHCA care and outcomes is essential to better plan for future pandemics. We found that during the initial surge period of the COVID-19 pandemic in the U.S., rates of survival to discharge after an IHCA decreased by 19% and rates of ROSC decreased by 27%, when compared with the same time period during 2015–2019. Lower rates of survival to discharge during the surge period was most pronounced in communities with very high COVID-19 mortality rates, which accounted for 1 in 5 IHCAs and where overall survival decreased by 42%. These findings during the surge period persisted even when patients without suspected or confirmed COVID-19 infection were excluded from analyses. In contrast, during the post-surge period starting in mid-May, rates of ROSC and survival to discharge were not lower in 2020 as compared with 2015–2019, even in communities with high COVID-19 mortality rates. Collectively, our findings provide important insights into the association between the COVID-19 pandemic and IHCA outcomes during the initial surge and immediate post-surge periods.
To date, there have been limited reports on the association between the COVID-19 pandemic and cardiac arrest survival outcomes in hospitals. Most prior studies have focused on survival outcomes of patients with COVID-19 infection and IHCA.2–6 However, there is scant information on the association of the pandemic on processes-of-care, resuscitation practices, and survival outcomes for all patients with IHCA as compared with prior years. Our study extends the existing literature in several important ways. First, we examined the association of the pandemic on processes of care and resuscitation practices. Second, we evaluated heterogeneity in the association between the pandemic and IHCA survival as a function of COVID-19 disease burden. Third, we evaluated the association of the pandemic during a pre-surge, surge, and post-surge period. The pre-surge period provided an additional control period and highlights the immediate and substantial decrease in IHCA survival coinciding with the initial COVID-19 surge in the U.S. Moreover, the stabilization of IHCA survival during the post-surge period suggests that hospitals adapted their resuscitation response after the initial surge period, especially in communities with high COVID-19 mortality rates. And lastly, we evaluated the effect of the pandemic in patients with IHCA who were not suspected to have active COVID-19 infection to determine if there were ‘spillover’ effects on unaffected patients.
Several findings deserve further mention. First, it is notable that rates of ROSC were lower during the COVID-19 surge period, regardless of a community’s COVID-19 mortality rate (including those with few COVID-19 deaths) and even when analyses excluded patients with COVID-19 infection. This suggests that many hospitals may have modified resuscitation protocols during the initial months of the pandemic when there was tremendous uncertainty and fear regarding the risk to responders in performing aerosolizing procedures such as chest compressions and bag-mask ventilation, and these changes may have affected hospital-wide responses to resuscitation regardless of a patient’s COVID-19 infection status. Pervasively lower rates of ROSC did not directly translate into lower rates of survival to discharge in all COVID-19 strata, as overall survival was significantly lower only in communities with monthly COVID-19 mortality rates of >50 per 1,000,000 residents. Nevertheless, this impact was substantial as IHCAs in communities with high and very high monthly COVID-19 mortality rates represented 49.7% (2954/5949) of all IHCAs during the surge period.
Despite lower rates of ROSC and survival to discharge during the surge period, we did not detect an increase in rates of delayed defibrillation or CPR initiation. Similarly, rates of DNAR orders in those who achieved ROSC did not increase during the surge period, even in communities with high COVID-19 mortality rates. We did find higher rates of delayed epinephrine treatment during this period, which may partly account for our survival findings as prompt epinephrine treatment is associated with survival18 and as non-shockable IHCAs comprise over 80% of all IHCAs. Perhaps the most dramatic change was in duration of resuscitative efforts during the surge period, particularly in communities with high and very high monthly COVID-19 mortality. In prior work, we have found that resuscitation time in those without ROSC is a surrogate measure of resuscitation effort intensity for IHCA and is associated with survival to discharge.10 Compared with a median resuscitation time of 25 minutes in 2015–2019, a decrease of 2.0 to 7.8 minutes in resuscitation time during the surge period represents a relative decrease of 8% to 31% in duration of resuscitative efforts for IHCA in communities with high and very high monthly COVID-19 mortality rates, respectively, and may help explain lower rates of survival to discharge in these strata during the surge period.
Somewhat surprisingly, rates of ROSC and survival to discharge during the immediate post-surge period in 2020 were not significantly different from the same time period in 2015–2019, even as the monthly COVID-19 mortality rate nationally was only modestly lower at 80.5 deaths per 1,000,000 residents during the post-surge period as compared with 105.7 deaths per 1,000,000 during the surge period. This was the case overall, and even in communities with the highest COVID-19 mortality rates during the post-surge period. This may reflect that hospitals throughout the U.S. had adapted their initial pandemic response to IHCA to avoid delivering lower quality resuscitation care even with new requirements for personal protective equipment and the potential of CPR and bag-mask ventilation to be aerosol-generating procedures. Notably, duration of resuscitation efforts during the post-surge period in 2020 was similar to those in 2015–2019, even in communities with the highest COVID-19 mortality. Further studies are needed to identify what modifications to resuscitation care and procedures were implemented in hospitals during the surge and post-surge periods to better understand our findings.
Our study should be interpreted in the context of the following limitations. First, we used county-level COVID-19 mortality rates to approximate the severity of the COVID-19 pandemic at hospitals. Although COVID-19 mortality is a more accurate marker of the pandemic’s severity than incidence (given geographic variability in testing), COVID-19 mortality is a lagging indicator of the pandemic’s severity in communities; therefore, our interaction analyses by COVID-19 mortality strata in the post-surge period should be interpreted with this limitation in mind. Second, GWTG-Resuscitation does not routinely collect data on a number of hospital resuscitation practices, such as quality (depth and rate) of chest compressions, ventilation adequacy and interruptions of CPR. Although we found shorter resuscitation times and higher rates of delayed epinephrine treatment during the surge period, other reasons for lower ROSC and survival rates during the surge period deserve study. Third, we examined the association of the pandemic on survival during the first half of 2020. Additional studies are needed to better understand if IHCA survival was lower during the summer and winter COVID-19 surges in 2020. Since limited availability and accuracy of COVID-19 testing during the first months of the pandemic may have led to underestimation of county-level COVID-19 mortality rates, the COVID-19 mortality strata results from this study should be interpreted with this limitation in mind especially if they are compared to later periods during the pandemic. Follow-up studies on the association of the COVID-19 pandemic on IHCA survival during subsequent surges in the U.S., as well as explicit survival comparisons between those with and without COVID-19 and IHCA in 2020, are currently ongoing and will provide additional insights. Fourth, GWTG-Resuscitation collects data only on patients with IHCAs who undergo cardiopulmonary resuscitation. If rates of DNAR orders had increased during the surge period,19 our study may have underestimated the effect of the pandemic on IHCA survival. Finally, our study was limited to hospitals participating in GWTG-Resuscitation and our findings may not be generalizable to non-participating hospitals.
Conclusion
In a large national registry, we found that survival rates for IHCA were lower during the initial COVID-19 pandemic surge in 2020, even in patients without suspected or confirmed COVID-19 infection. Although overall survival was lower in communities with high COVID-19 mortality rates, rates of ROSC were lower in all communities, including those with low COVID-19 disease burden. Our findings highlight the substantial impact of COVID-19 on in-hospital resuscitation outcomes during the initial months of the pandemic.
Supplementary Material
What is Known:
In-hospital cardiac arrest (IHCA) is common and has a low survival rate, and recent reports of the novel coronavirus 2019 (COVID-19) pandemic suggests that IHCA survival worsened during the pandemic.
What the Study Adds:
We found that rates of survival to discharge for IHCA were lower during the initial COVID-19 pandemic surge in the U.S., especially in communities with high COVID-19 mortality rates.
During the initial COVID-19 surge, the duration of resuscitation efforts was shorter and there were more frequent delays in defibrillation.
Rates of survival to discharge for IHCA decreased, even among patients without COVID-19 infection, highlighting the early impact of the COVID-19 pandemic on in-hospital resuscitation.
Acknowledgments
Funding/Support:
GWTG-Resuscitation is sponsored by the American Heart Association.
This funder had no role in the study design, data analysis or manuscript preparation and revision.
Acknowledgment and Authorship:
Dr. Chan had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Non-standard Abbreviations and Acronyms (in order of appearance in manuscript)
- COVID-19
novel coronavirus 2019
- IHCA
in-hospital cardiac arrest
- CPR
cardiopulmonary resuscitation
- GWTG
Get With The Guidelines
- ROSC
return of spontaneous circulation for ≥20 minutes
- DNAR
do not attempt resuscitation
Footnotes
Conflict of Interest Disclosures: None of the authors have any financial disclosures or conflicts of interest to disclose related to this manuscript.
REFERENCES
- 1.Edelson DP, Sasson C, Chan PS, Atkins DL, Aziz K, Becker LB, Berg RA, Bradley SM, Brooks SC, Cheng A, et al. , American Heart Association ECCICGA. Interim Guidance for Basic and Advanced Life Support in Adults, Children, and Neonates With Suspected or Confirmed COVID-19: From the Emergency Cardiovascular Care Committee and Get With The Guidelines-Resuscitation Adult and Pediatric Task Forces of the American Heart Association. Circulation 2020;141:e933–e943. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Shao F, Xu S, Ma X, Xu Z, Lyu J, Ng M, Cui H, Yu C, Zhang Q, Sun P, et al. In-hospital cardiac arrest outcomes among patients with COVID-19 pneumonia in Wuhan, China. Resuscitation 2020;151:18–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Thapa SB, Kakar TS, Mayer C, Khanal D. Clinical Outcomes of In-Hospital Cardiac Arrest in COVID-19. JAMA Intern Med 2021;181:279–281. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Miles JA, Mejia M, Rios S, Sokol SI, Langston M, Hahn S, Leiderman E, Salgunan R, Soghier I, Gulani P, et al. Characteristics and Outcomes of In-Hospital Cardiac Arrest Events During the COVID-19 Pandemic: A Single-Center Experience From a New York City Public Hospital. Circ Cardiovasc Qual Outcomes 2020;13:e007303. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Hayek SS, Brenner SK, Azam TU, Shadid HR, Anderson E, Berlin H, Pan M, Meloche C, Feroz R, O’Hayer P, et al. , Investigators S-C. In-hospital cardiac arrest in critically ill patients with covid-19: multicenter cohort study. BMJ 2020;371:m3513. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Sultanian P, Lundgren P, Stromsoe A, Aune S, Bergstrom G, Hagberg E, Hollenberg J, Lindqvist J, Djarv T, Castelheim A, et al. Cardiac arrest in COVID-19: characteristics and outcomes of in- and out-of-hospital cardiac arrest. A report from the Swedish Registry for Cardiopulmonary Resuscitation. Eur Heart J 2021;42:1094–1106. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Peberdy MA, Kaye W, Ornato JP, Larkin GL, Nadkarni V, Mancini ME, Berg RA, Nichol G, Lane-Trultt T. Cardiopulmonary resuscitation of adults in the hospital: a report of 14720 cardiac arrests from the National Registry of Cardiopulmonary Resuscitation. Resuscitation 2003;58:297–308. [DOI] [PubMed] [Google Scholar]
- 8.Cummins RO, Chamberlain D, Hazinski MF, Nadkarni V, Kloeck W, Kramer E, Becker L, Robertson C, Koster R, Zaritsky A, et al. Recommended guidelines for reviewing, reporting, and conducting research on in-hospital resuscitation: the in-hospital ‘Utstein style’. American Heart Association. Circulation 1997;95:2213–2239. [DOI] [PubMed] [Google Scholar]
- 9.Jacobs I, Nadkarni V, Bahr J, Berg RA, Billi JE, Bossaert L, Cassan P, Coovadia A, D’Este K, Finn J, et al. Cardiac arrest and cardiopulmonary resuscitation outcome reports: update and simplification of the Utstein templates for resuscitation registries: a statement for healthcare professionals from a task force of the International Liaison Committee on Resuscitation (American Heart Association, European Resuscitation Council, Australian Resuscitation Council, New Zealand Resuscitation Council, Heart and Stroke Foundation of Canada, InterAmerican Heart Foundation, Resuscitation Councils of Southern Africa). Circulation 2004;110:3385–3397. [DOI] [PubMed] [Google Scholar]
- 10.Goldberger ZD, Chan PS, Berg RA, Kronick SL, Cooke CR, Lu M, Banerjee M, Hayward RA, Krumholz HM, Nallamothu BK. Duration of resuscitation efforts and survival after in-hospital cardiac arrest: an observational study. Lancet 2012;380:1473–1481. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.New York Times COVID-19 Data https://github.com/nytimes/covid-19-data/blob/master/us-counties.csv. Acessed May, 15, 2020.
- 12.Department of Housing and Urban Development. Dataset Crosswalk API https://www.huduser.gov/portal/dataset/uspszip-api.html. Accessed June 11, 2020.
- 13.Austin PC. Using the Standardized Difference to Compare the Prevalence of a Binary Variable Between Two Groups in Observational Research. Communications in Statistics – Simulation and Computation 2009;38:1228–1234. [Google Scholar]
- 14.Goldstein H Multilevel Statistical Models, 3rd Edition. London. Free 2nd Edition: http://www.ats.ucla.edu/stat/examples/msm_goldstein/goldstein.pdf: Arnold Publishers; 2003. [Google Scholar]
- 15.Goldstein H, Browne W, Rasbash J. Partitioning variation in multilevel models. Understanding Statistics 2002;1:223–232. [Google Scholar]
- 16.Peberdy MA, Ornato JP, Larkin GL, Braithwaite RS, Kashner TM, Carey SM, Meaney PA, Cen L, Nadkarni VM, Praestgaard AH, et al. Survival from in-hospital cardiac arrest during nights and weekends. JAMA 2008;299:785–792. [DOI] [PubMed] [Google Scholar]
- 17.United States Census Bureau. Demographic Analysis Uses Birth and Death Records, International Migration Data and Medicare Records to Produce a Range of Population Estimates as of April 1, 2020 https://www.census.gov/library/stories/2020/12/census-bureau-provides-population-estimates-for-independent-evaluation-of-upcoming-census-results.html#:~:text=The%20tabulation%20of%20the%20official,1%2C%202020%20of%20332.6%20million. Accession date: March 27, 2021.
- 18.Donnino MW, Salciccioli JD, Howell MD, Cocchi MN, Giberson B, Berg K, Gautam S, Callaway C. Time to administration of epinephrine and outcome after in-hospital cardiac arrest with non-shockable rhythms: retrospective analysis of large in-hospital data registry. Bmj 2014;348:g3028. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Chan PS, Berg RA, Nadkarni VM. Code Blue During the COVID-19 Pandemic. Circ Cardiovasc Qual Outcomes 2020;13:e006779. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.