University of Southern California Dissertations and Theses

Characteristics associated with emergency medical services for cardiac arrest pre- and during COVID-19 in Los Angeles, CA

(USC Thesis Other)

Characteristics associated with emergency medical services for cardiac arrest pre- and during COVID-19 in Los Angeles, CA

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Characteristics Associated with Emergency Medical Services for Cardiac Arrest Pre- and During
COVID-19 in Los Angeles, CA

By

Xinyu Hu

A Thesis Presented to the
FACULTY OF THE KECK SCHOOL OF MEDICINE
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(BIOSTATISTICS)

May 2022

ii

TABLE OF CONTENTS
List of Tables ............................................................................................................................................... iii
List of Figures .............................................................................................................................................. iv
Abstract ........................................................................................................................................................ vi
1. Introduction ............................................................................................................................................... 1
2. Methods .................................................................................................................................................... 1
2.1 Sample................................................................................................................................................. 1
2.2 Variables ............................................................................................................................................. 2
2.2.1 Primary Outcomes........................................................................................................................ 2
2.2.2 Demographics and Comorbidities ................................................................................................ 2
2.3 Unit of Observation ............................................................................................................................. 3
2.4 Analysis............................................................................................................................................... 3
2.4.1 Weekly Trends ............................................................................................................................. 3
2.4.2 Yearly Trends ............................................................................................................................... 4
2.4.3 Multivariable Association ............................................................................................................ 4
3. Results ....................................................................................................................................................... 5
3.1 Weekly Trends .................................................................................................................................... 5
3.2 Yearly Trends .................................................................................................................................... 12
3.3 Multivariable Regression Models ..................................................................................................... 17
4. Discussion ............................................................................................................................................... 20
References ................................................................................................................................................... 23

iii

List of Tables
Table 1. Associations between EMS measures and year. ............................................................. 12
Table 2. Demographics and comorbidities across years. .............................................................. 15
Table 3. Multivariable logistic regression models in EMS-treated
events and bystander CPR delivered events among EMS-attended events. ................................. 18
Table 4. Multivariable logistic regression model
a
in bystander CPR
delivered events among EMS-treated events. ............................................................................... 19

iv

List of Figures
Figure 1. Weekly counts and Loess smoothed curves for
the EMS-attended events. ............................................................................................................... 6
Figure 2. Weekly counts and Loess smoothed curves for
the bystander CPR delivery events. ................................................................................................ 6
Figure 3. Weekly counts and Loess smoothed curves for
the EMS-treated events. .................................................................................................................. 7
Figure 4. Weekly counts and Loess smoothed curves for
the EMS-untreated events. .............................................................................................................. 7
Figure 5. Weekly counts and Loess smoothed curves for
the cardiac arrest witnessed events. ................................................................................................ 8
Figure 6. Weekly counts and Loess smoothed curves for
the cardiac arrest unwitnessed events. ............................................................................................ 8
Figure 7. Weekly proportions and Loess smoothed curves
of EMS-treated events among EMS-attended events. .................................................................... 9
Figure 8. Weekly proportions and Loess smoothed curves
of bystander CPR delivered events among EMS-attended events. ............................................... 10
Figure 9. Weekly proportions and Loess smoothed curves
of bystander CPR delivered events among EMS-treated events. ................................................. 10
Figure 10. Weekly proportions and Loess smoothed curves
of bystander CPR delivered events among arrest witnessed events. ............................................ 11
Figure 11. Weekly proportions and Loess smoothed curves
of bystander CPR delivered events among arrest unwitnessed events. ........................................ 11

v

Figure 12. Boxplots of age across four individual years. ............................................................. 13
Figure 13. Boxplots of age before and during the COVID-19 pandemic. .................................... 14

vi

Abstract
Patients experiencing cardiac arrest often seek help from the 9-1-1 Emergency Medical
Service (EMS) Management, yet the EMS Management in Los Angeles, CA was impacted by the
challenging situation of the COVID-19 pandemic. This study analyzed the relationship between
patients’ demographics and comorbidities and the EMS performance including EMS attendance,
treatment, arrest witness, and bystander Cardiopulmonary resuscitation (CPR) delivery, and
whether the COVID-19 pandemic was associated with changes to the EMS performance metrics.
When compared to years 2017 through 2019, more patients during COVID-19 experienced cardiac
arrest and called EMS for help but not so many underwent the EMS treatment. The proportion of
EMS-treated among EMS-attended events decreased during COVID-19, and COVID-19 was also
associated with lowering the proportion of bystander CPR delivery among EMS-attended events.
Among EMS-attended events, when compared to the baseline age group (< 52 y/o), COVID-19
was associated with the lower odds of being EMS-treated for older age groups (52-66 y/o: 0.90
(95%CI: 0.68, 1.20), 66-80 y/o: 1.12 (95%CI: 0.84, 1.51), and over 80 y/o: 0.69 (95%CI: 0.51,
0.92)). When compared to non-homeless people, the odds of being EMS-treated (OR: 0.59, 95%CI:
0.47, 0.76) and having bystander CPR delivered (OR: 0.32, 95%CI: 0.22, 0.48) for homeless
people were lower. Among EMS-treated events, when compared to non-homeless people, the odds
of having bystander CPR delivered for homeless people was 0.37 (95%CI: 0.24, 0.58). We
conjecture that patients infected by COVID-19 were likely to have more serious outcomes related
to the infection during cardiac arrest and that residents were less likely to help deliver CPR for the
fear of COVID-19 infections. More attention is needed in getting residents in Los Angeles to
participate in CPR delivery for homeless people experiencing cardiac arrest.

1

1. Introduction

Since the declaration of the COVID-19 pandemic in mid-March, 2020, the number of cases
and deaths due to COVID-19 greatly increased [1]. In large cities, when compared to pre-COVID-
19 years, the number of people experiencing out-of-hospital cardiac arrest increased, while the
proportion of those who survived from cardiac arrest decreased [2]- [3]. To aid in challenging
situations during the COVID-19 pandemic and standardize the treatment protocols for high
COVID-19 prevalence areas, interim guidance on Emergency Medical Services (EMS) for cardiac
arrest incidents was published [4]. The EMS management in the City of Los Angeles also
collaborated with local health care providers to adopt pre-hospital strategies in identifying patients
in need of acute treatment [5]. This article further investigates the EMS performance in the City
of LA in relation to the demographics and comorbidities of the patients, and whether the EMS
performance metrics changed between the COVID-19 pandemic and pre-COVID-19 years.

2. Methods
2.1 Sample
In the City of Los Angeles, the Los Angeles Fire Department (LAFD) takes the sole
responsibility in response to over one million 9-1-1 calls annually, and dispatch units through the
LAFD Metropolitan Fire Communications Center to over 400,00 incidents every year [6]. Upon
the receipt of 9-1-1 calls, professionally trained and certified emergency medical technicians or
paramedics provide medical guidance to the caller or deploy units for direct medical assistance in
the field. After the unit deployment, the dispatch data is logged with a timestamp and synchronized

2

to the dispatch system at LAFD. The data set for this analysis contains information for all cardiac
arrest events attended by LAFD from January 2017 to March 2021. The sample includes 30,549
phone calls received by LAFD, demographic information and medical conditions for the associated
patients and the EMS attendance decisions after each phone call.
2.2 Variables
2.2.1 Primary Outcomes
From the data set, we computed the number of cardiac arrest events that 1) were attended
by LAFD EMS staff (“EMS-attended”), 2) were well enough to be treated by LAFD EMS staff
(“EMS-treated”), 3) were witnessed by the EMS personnel (“EMS-witnessed”), and 4) had a
bystander deliver Cardiopulmonary resuscitation (CPR). From these counts, we calculated rates
and proportions for our primary outcomes. Using the annual population estimates in Los Angeles
from the US Census Bureau, we calculated the citywide annual rate of EMS-attended events and
the annual rate of EMS-treated events. Additionally, we computed the proportion of EMS-attended
events that 1) were EMS-treated, and 2) had a bystander deliver CPR. We also computed the
proportion of EMS-treated events that 1) were EMS-witnessed, and 2) had a bystander deliver
CPR. The proportion of events with bystander CPR was additionally computed within all EMS-
witnessed and within all EMS-unwitnessed events.
2.2.2 Demographics and Comorbidities
For each EMS-attended event, EMS responders reported several demographic
characteristics about the patient to the best of their ability. These characteristics included age,
gender, race, homelessness, medical conditions (e.g., asthma, emphysema), and medical response
by a layperson (e.g., applied AED, defibrillation). EMS staff were able to choose the units for
patient age (i.e., “minutes”, “hours”, “days”, “months”, and “years”). All values for age were

3

converted to years for consistency in the analysis. Gender was recorded either as “male” or as
“female”. Race was recorded by the EMS staff to the best of their knowledge, with the option to
choose multiple race categories. Therefore, race was regrouped by the first listed race category of
White, Hispanic, Black or African American or Other. Homelessness, asthma, emphysema,
layperson applied AED and layperson applied defibrillation were recorded as “Yes” or “NA”; all
values recorded as “NA'' were conservatively assumed to be “No”.
2.3 Unit of Observation
Event counts from 2017 to 2021 were aggregated both yearly and weekly. To examine the
effect of the COVID-19 pandemic, each year was designated as starting in the second week in
March, which corresponds to when COVID-19 was declared as a global pandemic in 2020. Yearly
data were aggregated within each of four years, ranging from 2017 to 2020. We also examined
year as a binary variable that compared 2020 (post-pandemic) to 2017-2019 (pre-pandemic).
Counts were additionally aggregated at the week level, within year groupings defined above, with
weeks defined as Sunday through Saturday.
2.4 Analysis
2.4.1 Weekly Trends
Aggregated weekly plots were generated for the EMS-related measures (weekly counts in
EMS-attended events, EMS-treated events, EMS-untreated events, arrest witnessed events, arrest
unwitnessed events, and bystander CPR performed events, and proportions of EMS-treated among
EMS-attended, bystander CPR performed among attended, bystander CPR performed among
treated, bystander CPR performed among arrest witnessed, and bystander CPR performed among
arrest unwitnessed) stratified by year. A Loess smoothed curve was fitted for each individual year.

4

Unlike a simple linear regression line, a Loess smooth curve can highlight weekly fluctuations by
taking local data into account and is still able to show the direction of change in the data when
generating the plot [7]-[8]. Therefore, general trends and fluctuations in weekly data were
examined visually.
2.4.2 Yearly Trends
From the yearly data, a χ
2
-test was performed to evaluate the univariate association
between each of the EMS-related measures (EMS attendance rate, EMS treatment rate, proportion
of EMS-witnessed over EMS-treated, proportion of bystander CPR over EMS-attended,
proportion of bystander CPR over EMS-treated, proportion of bystander CPR over EMS-witnessed,
and proportion of bystander CPR over EMS-unwitnessed) and year (individual years and COVID-
19 vs. pre-COVID-19). A χ
2
-test was also used to evaluate the relationship between year and
categorical demographic factors (gender, race, homelessness), and the relationship between year
and medical conditions (asthma, emphysema, layperson applied AED and layperson applied
defibrillation). To evaluate the association between the continuous variable age and year
(individual years and COVID-19 vs pre-COVID-19), an ANOVA test was performed to compare
the mean patient age over years. The equal variance assumption was checked using Levene’s test.
For the evaluation of association between categorical measures and year, a χ
2
-test was used.
2.4.3 Multivariable Association
To further investigate the relationship between each of the EMS related measures (EMS
treatment, bystander CPR performance among EMS-attended, and bystander CPR performance
among EMS-treated) and demographic factors, comorbidities and year (COVID-19 vs pre-
COVID-19), multivariable logistic regression models were developed. When modeling age as a
continuous variable in the logistic regression models, the linearity assumption was checked using

5

fractional polynomials. If the linearity assumption failed, age was transformed into a categorical
variable based on its quartiles. In the multivariable regression model, all possible interaction
effects between year and the independent variables were tested and kept in the final model if the
interaction effects of interest yielded statistical significance. For those independent variables that
did not have significant interaction effect with year, their overall odds ratios were reported.

3. Results
3.1 Weekly Trends
Figure 1 - Figure 6 shown below are the graphs of weekly counts, Loess smoothed curves
and the 95% confidence level intervals for predictions of the EMS related measures from 2017 to
2020. From Figure 1 and Figure 4, when COVID-19 started, the weekly counts in EMS-attended
events and EMS-untreated events were remarkably greater than in 2017 through 2019. In Figure 3
and Figure 6, the weekly counts in EMS-treated events and arrest unwitnessed events do not exhibit
much difference through much of the year yet increased during the winter season of 2020, after
COVID-19 started.

6

Figure 1. Weekly counts and Loess smoothed curves for EMS-attended events.

Figure 2. Weekly counts and Loess smoothed curves for bystander CPR performed events.

7

Figure 3. Weekly counts and Loess smoothed curves for EMS-treated events.

Figure 4. Weekly counts and Loess smoothed curves for EMS-untreated events.

8

Figure 5. Weekly counts and Loess smoothed curves for the cardiac arrest EMS-witnessed events.

Figure 6. Weekly counts and Loess smoothed curves for the cardiac arrest EMS-unwitnessed events.

9

Figure 7 - Figure 11 are the graphs of weekly proportions of EMS related measures from
2017 to 2020. In Figure 7, the proportion of EMS-treated events among EMS-attended events
increased from 2017 to 2019 but was lower in 2020, when COVID-19 started. Similarly, the
proportion of bystander CPR performed events among EMS-attended events in Figure 8 increased
from 2017 to 2019 but remarkably decreased in 2020 when COVID-19 started.

Figure 7. Weekly proportions and Loess smoothed curves of EMS-treated events among EMS-attended events.

10

Figure 8. Weekly proportions and Loess smoothed curves of bystander CPR delivered events among EMS-attended events.

Figure 9. Weekly proportions and Loess smoothed curves of bystander CPR performed events among EMS-treated events.

11

Figure 10. Weekly proportions and Loess smoothed curves of bystander CPR delivered among arrest EMS-witnessed events.

Figure 11. Weekly proportions and Loess smoothed curves of bystander CPR delivered events among arrest unwitnessed events.

12

3.2 Yearly Trends

Table 1. Associations between EMS measures and year.
Variable
a

4-Year Group Comparison
(N = 30549)
COVID-19 Comparison
(N = 30549)
2017 2018 2019 2020 p-value
Pre-Pandemic
c

(N = 20516)
During Pandemic
c

(N = 10033)
p-value
Annual Rates
(per 100,000 people)

EMS-attended 176.5 164.9 172.9 252.4 < 0.001 171.4 252.4 < 0.001

EMS-treated 60.1 69.9 74.9 88.9 < 0.001 68.3 88.9 <0.001

Proportions

EMS-treated
b
0.34 0.42 0.43 0.35 < 0.001 0.40 0.35 < 0.001

Arrest witnessed
b
0.45 0.45 0.43 0.48 0.002 0.44 0.48 < 0.001

Bystander CPR in
EMS-attended 0.20 0.24 0.22 0.15 < 0.001 0.22 0.15 < 0.001

Bystander CPR in
EMS-treated 0.49 0.52 0.50 0.42 < 0.001 0.50 0.42 < 0.001

Bystander CPR in
arrest witnessed 0.48 0.49 0.44 0.41 < 0.001 0.47 0.41 < 0.001

Bystander CPR in
arrest unwitnessed 0.50 0.54 0.54 0.43 < 0.001 0.53 0.43 < 0.001

a. A χ
2
-test was performed to evaluate the association between EMS-related measures and year.
b. In EMS-attended events.
c. Rates are calculated based on the annual population of the city of LA by US Census Bureau [9].

Table 1 summarizes the statistical tests conducted to evaluate the association between EMS
related measures and year (4 individual years and COVID-19 vs pre-COVID-19). With a
significance level of 0.05, all EMS-related measures were significantly associated with year (p ≤
0.002). When compared to the pre-pandemic years, EMS-attended annual rate and EMS-treated
annual rate increased during COVID-19. On the other hand, when compared to years before
COVID-19, the proportions of EMS-treated over EMS-attended, arrest EMS-witnessed over EMS-

13

attended, bystander CPR delivered over EMS-attended, bystander CPR delivered over EMS-
treated, bystander CPR delivered over arrest EMS-witnessed, and bystander CPR delivered over
arrest EMS-unwitnessed decreased during the pandemic.

Figure 12. Boxplots of age across years.

14

Figure 13. Boxplots of age before and during the COVID-19 pandemic.

15

Table 2. Demographics and comorbidities across years.
Variable
a,b

4-Year Group Comparison
(N = 30549)
COVID-19 Comparison
(N = 30549)
2017 2018 2019
2020 p-value Pre-COVID-19 During COVID-19 p-value

Age 65.0 ± 20.3 64.7 ± 20.0 64.1 ± 20.1 64.1 ± 20.4 0.080 64.5 ± 20.1 64.1 ± 20.4 0.075

Gender

0.001

0.014

Male 61.4 62.5 64.9 64.8 63.4 64.8

Homelessness

< 0.001

0.001

Yes 10.0 5.7 6.1 7.5 6.5 7.5

Race

< 0.001

< 0.001

Black or African American 7.3 6.1 4.4 4.3 5.6 4.3

Hispanic 6.9 5.6 4.4 5.1 5.2 5.1

White 8.4 6.9 5.1 4.5 6.3 4.5

Other 4.6 4.2 2.1 2.1 3.3 2.1

Missing 72.8 77.2 84.1 84.0 79.6 84.0

Asthma

0.001

0.005

Yes 3.9 3.1 4.1 3.0 3.7 3.0

Emphysema

0.053

0.025

Yes 1.4 1.6 1.3 1.1 1.4 1.1

Layperson applied AED

< 0.001

< 0.001

Yes 4.5 4.4 4.7 2.9 4.5 2.9

Layperson applied
defibrillation

< 0.001

< 0.001
Yes 2.3 1.9 1.6 0.7 1.8 0.7
a. Values are mean ± SD for age and percent for the remaining variables
b. An ANOVA test was used to evaluate the association between age and year. A χ
2
-test was performed to evaluate the association between categorical demographics,
comorbidities and year.

16

Table 2 summarizes statistical tests used to evaluate the relationship between demographic
characteristics, comorbidities and year (individual years and COVID-19 vs pre-COVID-19).
Testing the relationship between age and year, the Levene’s test was used to check for the equal
variance assumption (4-year group: p = 0.018, COVID-19 comparison: p = 0.002). The assumption
was violated primarily due to the large sample size. There was no significant difference in age over
year (p=0.08). Figure 12 and Figure 13 are the box plots for age vs year (left: pre-COVID-19 vs
COVID-19, right: individual years). By visually comparing the median and interquartile ranges in
the two plots above, no remarkable difference was found, which was consistent with the ANOVA
test result for difference in mean age across years.
From the χ
2
-test, gender, homelessness, race, asthma, layperson applied AED and
layperson applied defibrillation statistically significant differ over year (individual years and
COVID-19 vs pre-COVID-19) (p < 0.02). Emphysema was statistically significantly associated
with year in COVID-19 comparison (p = 0.025) and was not statistically significantly associated
with year in the 4-year group comparison (p = 0.053). There was no statistically significant
association between age and year (4-year group: p = 0.080, COVID-19 comparison: 0.075).
Compared to pre-pandemic years, the proportions of male patients and homeless patients increased
during COVID-19. On the other hand, the proportions of patients with asthma, emphysema,
experienced a layperson applied AED and experienced a layperson applied defibrillation were
lower during COVID-19 comparing to pre-pandemic years. For the racial composition of patients,
since the percentage of missing data increased during COVID-19, all specified race levels had
lower percentages comparing to pre-pandemic years.

17

3.3 Multivariable Regression Model

Table 3 summarizes two adjusted logistic regression models. The models were based on
the EMS-attended events, and the outcomes were EMS-treated events and Bystander CPR
performed events. When the interaction between year (pre- vs. during COVID-19) and an
independent variable was not significant, the result for that variable was reported as an overall
odds ratio and its associated 95% confidence interval. The linearity assumption for age failed;
therefore, the covariates tested in the models were age as a categorical variable, gender,
homelessness, race, asthma and emphysema. In the regression model for EMS-treated events, the
interaction effect between year (COVID-19 vs pre-COVID-19) and age was statistically significant
(p = 0.025). There was no statistically significant interaction effect between race and year
(COVID-19 vs pre-COVID-19) overall (p = 0.072). In the regression model for Bystander CPR
performed events, there was a statistically significant interaction effect between year (COVID-19
vs pre-COVID-19) and age overall (p < 0.001). Among the EMS-attended events, when compared
to the baseline category of individuals less than 52 years old, the three age categories (52-66 y/o,
66-80 y/o and over 80 y/o) had lower odds of being EMS-treated when comparing the pre- versus
during COVID-19. When compared to the baseline category of Black or African Americans, White
individuals also had decreased odds of being EMS-treated (OR: 0.57, 95%CI: 0.43, 0.75) during
COVID-19, compared to the pre-pandemic years (p = 0.02).

18

Table 3. Multivariable logistic regression models in EMS-treated events and bystander CPR delivered events among EMS-attended events.
Variable
a,b,c

Proportions Outcome: EMS-Treated (N = 25669) Outcome: Bystander CPR (N = 25669)
Pre-
COVID-19
During
COVID-19
Pre-COVID-19 During COVID-19
Interaction
p-value
Pre-COVID-19 During COVID-19
Interaction
p-value
Odds ratio
(95% CI)
Odds ratio
(95% CI)
Odds ratio
(95% CI)
Odds ratio
(95% CI)

Age

<52
0.239 0.258 ref. ref. ref. ref. ref. ref.

52-66
0.250 0.242 1.30 (1.06, 1.59) 0.90 (0.68, 1.20)
0.03
1.25 (0.97, 1.59) 1.02 (0.70, 1.47)
<0.001

66-80
0.250 0.239 1.69 (1.37, 2.08) 1.12 (0.84, 1.51) 1.60 (1.25, 2.03) 1.13 (0.78, 1.65)

≥8 0
0.258 0.260 1.17 (0.95, 1.45) 0.69 (0.51, 0.92) 1.43 (1.11, 1.83) 0.86 (0.59, 1.25)

Missing
0.003 0.001 - - - - - -

Gender

Male
0.634 0.648 0.98 (0.86, 1.10) n.s. 0.98 (0.85, 1.13) n.s.

Homelessness

Yes
0.065 0.075 0.59 (0.47, 0.76) n.s. 0.32 (0.22, 0.48) n.s.

Race

Black or
African
American
0.056 0.043 ref. ref. ref. ref. ref. ref.

Hispanic
0.052 0.051 1.21 (0.99, 1.46) 1.14 (0.88, 1.48)
0.07
1.23 (1.02, 1.48) n.s.

White
0.063 0.045 0.86 (0.71, 1.03) 0.57 (0.43, 0.75) 0.99 (0.82, 1.20) n.s.

Other
0.033 0.021 0.96 (0.77, 1.20) 0.75 (0.53, 1.06) 0.99 (0.79, 1.23) n.s.

Missing
0.796 0.840 - - - - - -

Asthma

Yes
0.037 0.030 1.03 (0.80, 1.32) n.s. 0.83 (0.61, 1.13) n.s.

Emphysema

Yes
0.014 0.011 0.79 (0.54, 1.18) n.s. 1.01 (0.64, 1.59) n.s.
a. The table is based on EMS-attended events.
b. Multivariable logistic regression models were created to evaluate the relationship between all demographics, comorbidities and EMS-treated, bystander CPR delivered events.
c. For independent variables that did not have significant interaction effect with year, their overall odds ratios were reported.

19

In the regression model for Bystander CPR performed events, when compared to the
baseline category of individuals less than 52 years old, the three age categories (52-66 y/o, 66-80
y/o and over 80 y/o) had lower odds of being EMS-treated when comparing the pre- versus during
COVID-19. Among the three age categories, the reduction in the odds ratios for categories of
individuals between 66 and 80 years old and individuals over 80 years old were more outstanding.

Table 4. Multivariable logistic regression model
a
in bystander CPR delivered events among EMS-treated events.
Variable
Outcome: Bystander CPR (N = 10274)
Odds ratio 95% CI

Age (ref: <52)

52-66
1.14 (0.89, 1.47)

66-80
1.29 (1.00, 1.65)

≥8 0
1.48 (1.14, 1.92)

Gender

Male
0.99 (0.83, 1.19)

Homelessness

Yes
0.37 (0.24, 0.58)

Race (ref: African
American)

Hispanic 1.07 (0.85, 1.35)

White 1.31 (1.03, 1.66)

Other 1.09 (0.82, 1.43)

Asthma

Yes
0.72 (0.49, 1.04)

Emphysema

Yes
1.29 (0.70, 2.38)
a. The odds ratios for all variables were not statistically significantly different between pre- and during COVID
(p> 0.05). Thus, the overall odds ratios were reported.

Table 4 summarizes the results of the logistic regression model based on all EMS-treated
events. The outcome was Bystander CPR performed events and the covariates of interest were the

20

demographic factors and comorbidities, along with the interaction effect between year (COVID-
19 vs pre-COVID-19) and those covariates. From the regression model, no statistically significant
interaction effect between year and covariates was observed. Among all age categories, when
comparing to the baseline category of the individuals less than 52 years old, the individuals over
80 years old had the greatest odds of experiencing bystander CPR when they were treated in an
EMS event (OR: 1.48, 95% CI: 1.14, 1.92). On the other hand, when compared to non-homeless
individuals, homeless people had the lower odds of experiencing bystander CPR when they were
treated in an EMS event (OR: 0.37, 95% CI: 0.24, 0.58).

4. Discussion

During COVID-19, the number of both EMS-attended and EMS-treated events increased
but the proportion of EMS-treated over EMS-attended decreased. When comparing EMS measures
across years, all p-values were found much smaller than the significance level. In the regression
models, COVID-19 was found to be associated with lowering the odds of being EMS-treated or
delivered with bystander CPR for older patients and White patients. Homeless people had lower
odds of having EMS-treated or bystander CPR delivered comparing to non-homeless people but
COVID-19 did not exhibit significant association with lowering the odds.
The number of EMS-attended events and the number of EMS-treated events increased
during COVID-19. After the pandemic started, the higher number of patients affected by COVID-
19 might need EMS, but the symptoms during cardiac arrest might be too severe for EMS treatment.
This could explain the increase in the number of the EMS-attended events and in the number of
the EMS-untreated events. During the winter season after COVID-19 started, the number of

21

COVID-19 cases greatly increased [10]-[11], accompanied by an increase in the number of
patients having COVID-related cardiac complications [12] that might need EMS treatment.
Therefore, the remarkable increase in COVID-19 cases during the winter season may partly
explain the increase in the number of the EMS-treated events.
During COVID-19, the proportion of EMS-treated among EMS-attended decreased. This
finding might have been due to COVID-19 infection having more serious outcomes during cardiac
arrest. The proportion reduction also coincides with the increase in both EMS-attended events and
EMS-untreated events during COVID-19. COVID-19 was also associated with lowering the
proportion of Bystander CPR performed among EMS-attended, presumably because citizens were
less likely to help deliver CPR for the fear of being infected by COVID-19.
When comparing EMS measures across individual years and between pre- and during
COVID-19, the p-values were found much more significant. Since the sample size was large, the
statistical tests became more powerful to detect the difference in the EMS measures across years.
Therefore, p-values were much smaller than the significance level of 0.05.
In the regression model of EMS-attended events, COVID-19 was found to be associated
with lowering the odds of being EMS-treated or delivered with bystander CPR for older patients.
When compared to the baseline age group of younger than 52 years old, COVID-19 was associated
with lowering the odds of being EMS-treated for the other age groups (p = 0.04 for 52-66 y/o, p =
0.03 for 66-80 y/o, and p < 0.01 for over 80 y/o). The reason may be because elder patients with
COVID-19 infection were having too serious outcomes during cardiac arrest to receive EMS
treatment. When compared to the baseline age group of younger than 52 years old, COVID-19 was
associated with lowering the odds of being delivered with bystander CPR for patients over 80 years

22

old (p = 0.03). This may be due to the citizens' fear of being infected by COVID-19 and elder
patients' serious outcomes during cardiac arrest.
From the regression model of EMS-attended events, homeless people tend to have lower
odds of being EMS-treated or delivered with bystander CPR during cardiac arrest. When compared
to non-homeless people, the odds of being EMS-treated and performed with bystander CPR for
homeless people did not change significantly between pre- and during COVID-19. However, when
compared to non-homeless people, the odds of being EMS-treated (OR: 0.59, 95%CI: 0.47, 0.76)
and delivered with bystander CPR (OR: 0.32, 95%CI: 0.22, 0.48) for homeless people were lower;
both odds ratios were well below 1. This suggests that because of their poor living condition and
possibly more severe symptoms during the cardiac arrest, homeless people might have had lower
odds of being EMS-treated or delivered with the bystander CPR.
In the EMS-attended regression model, when compared to Black or African Americans,
White people had decreased odds (OR: 0.57, 95%CI: 0.47, 0.75) of being EMS-treated during
COVID-19 (p = 0.02). The reduced odds ratio may not reflect the actual situation and needs further
validation, because around 80% of the data were missing racial information.
In the regression model of EMS-treated events, similar to the finding from EMS-attended
regression model, homeless people also had lower odds of having bystander CPR delivered (OR:
0.37, 95%CI: 0.24, 0.58) comparing to non-homeless people. Therefore, much more attention is
needed in getting more residents to help deliver CPR to homeless people that are experiencing
cardiac arrest.

23

References

[1] “United States - COVID-19 overview - Johns Hopkins,” Johns Hopkins Coronavirus
Resource Center. [Online]. Available: https://coronavirus.jhu.edu/region/united-states.
[Accessed: 27-Feb-2022].

[2] P. H. Lai, E. A. Lancet, M. D. Weiden, M. P. Webber, R. Zeig-Owens, C. B. Hall, and D. J.
Prezant, “Characteristics associated with out-of-hospital cardiac arrests and resuscitations during
the novel Coronavirus Disease 2019 pandemic in New York City,” JAMA Cardiology, vol. 5, no.
10, pp. 1154–1163, Oct. 2020.

[3] E. Marijon, N. Karam, D. Jost, D. Perrot, B. Frattini, C. Derkenne, A. Sharifzadehgan, V.
Waldmann, F. Beganton, K. Narayanan, A. Lafont, W. Bougouin, and X. Jouven, “Out-of-
hospital cardiac arrest during the COVID-19 pandemic in Paris, France: A population-based,
observational study,” The Lancet Public Health, vol. 5, no. 8, pp. 437–443, May 2020.

[4] J. M. Goodloe, A. Topjian, A. Hsu, R. Dunne, A. R. Panchal, M. Levy, M. McEvoy, C.
Vaillancourt, J. G. Cabanas, M. S. Eisenberg, T. D. Rea, P. J. Kudenchuk, A. Gienapp, G. E.
Flores, S. Fuchs, K. M. Adelgais, S. Owusu-Ansah, M. Terry, K. N. Sawyer, P. Fromm, M.
Panczyk, M. Kurz, G. Lindbeck, D. K. Tan, D. P. Edelson, and M. R. Sayre, “Interim guidance
for emergency medical services management of out-of-hospital cardiac arrest during the
COVID-19 pandemic,” Circulation: Cardiovascular Quality and Outcomes, vol. 14, no. 7, Jun.
2021.

[5] S. Sanko and M. Eckstein, “Mobile Integrated Health Care in Los Angeles: Upstream
solutions to mitigate the COVID-19 pandemic,” NEJM Catalyst, vol. 2, no. 2, Jan. 2021.

[6] S. Sanko, S. Kashani, C. Lane, and M. Eckstein, “Implementation of the Los Angeles tiered
dispatch system is associated with an increase in telecommunicator-assisted CPR,”
Resuscitation, vol. 155, pp. 74–81, Jun. 2020.

[7] “Local Polynomial Regression Fitting,” R: Local polynomial regression fitting, n.d..
[Online]. Available: https://stat.ethz.ch/R-manual/R-devel/library/stats/html/loess.html.
[Accessed: 27-Feb-2022].

[8] S. Berg, “Loess smoothing,” Meticulous Data Science, 05-Aug-2018. [Online]. Available:
https://meticulousdatascience.com/journal/loess-smoothing.html. [Accessed: 27-Feb-2022].

24

[9] “U.S. Census Bureau quickfacts: Los Angeles City, California,” U.S. Census Bureau.
[Online]. Available: https://www.census.gov/quickfacts/losangelescitycalifornia. [Accessed: 28-
Feb-2022].

[10] X. Liu, J. Huang, C. Li, Y. Zhao, D. Wang, Z. Huang, and K. Yang, “The role of seasonality
in the spread of covid-19 pandemic,” Environmental Research, vol. 195, p. 110874, Feb. 2021.

[11] S. Chen, K. Prettner, M. Kuhn, P. Geldsetzer, C. Wang, T. Bärnighausen, and D. E. Bloom,
“Climate and the spread of covid-19,” Scientific Reports, vol. 11, Apr. 2021.

[12] L. Ma, K. Song, and Y. Huang, “Coronavirus disease-2019 (COVID-19) and cardiovascular
complications,” Journal of Cardiothoracic and Vascular Anesthesia, vol. 35, no. 6, pp. 1860–
1865, May 2020.

Abstract (if available)

Abstract Patients experiencing cardiac arrest often seek help from the 9-1-1 Emergency Medical Service (EMS) Management, yet the EMS Management in Los Angeles, CA was impacted by the challenging situation of the COVID-19 pandemic. This study analyzed the relationship between patients’ demographics and comorbidities and the EMS performance including EMS attendance, treatment, arrest witness, and bystander Cardiopulmonary resuscitation (CPR) delivery, and whether the COVID-19 pandemic was associated with changes to the EMS performance metrics. When compared to years 2017 through 2019, more patients during COVID-19 experienced cardiac arrest and called EMS for help but not so many underwent the EMS treatment. The proportion of EMS-treated among EMS-attended events decreased during COVID-19, and COVID-19 was also associated with lowering the proportion of bystander CPR delivery among EMS-attended events. Among EMS-attended events, when compared to the baseline age group (< 52 y/o), COVID-19 was associated with the lower odds of being EMS-treated for older age groups (52-66 y/o: 0.90 (95%CI: 0.68, 1.20), 66-80 y/o: 1.12 (95%CI: 0.84, 1.51), and over 80 y/o: 0.69 (95%CI: 0.51, 0.92)). When compared to non-homeless people, the odds of being EMS-treated (OR: 0.59, 95%CI: 0.47, 0.76) and having bystander CPR delivered (OR: 0.32, 95%CI: 0.22, 0.48) for homeless people were lower. Among EMS-treated events, when compared to non-homeless people, the odds of having bystander CPR delivered for homeless people was 0.37 (95%CI: 0.24, 0.58). We conjecture that patients infected by COVID-19 were likely to have more serious outcomes related to the infection during cardiac arrest and that residents were less likely to help deliver CPR for the fear of COVID-19 infections. More attention is needed in getting residents in Los Angeles to participate in CPR delivery for homeless people experiencing cardiac arrest.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Evaluation of factors influencing Los Angeles Tiered-Dispatch System’s improvement on bystander CPR rate and inter reliability between electronic patient care report (ePCR) and 911 call review on...

PDF

Association of neighborhood characteristics with bystander CPR and out-of-hospital cardiac arrests in the city of Los Angeles

PDF

Racial/ethnic differences in patient experiences with health care in association with the Gleason’s score level at prostate cancer diagnosis: findings from the SEER-CAHPS data

PDF

The impact of the COVID-19 pandemic on cancer care delivery

PDF

Contemporary outcomes for adult congenital heart surgery in an adult tertiary care hospital

PDF

Expansion of emergency medical service regional routing for stroke patient care in the United States from 2000-2010

PDF

Comparative study of laparoscopy vs. laparotomy for ovarian mass removal

PDF

The evaluation of the long-term effectiveness of zero/low fluoroscopy workflow in ablation procedures for the treatment of paroxysmal and persistent atrial fibrillation

PDF

Relationship between L.A. County residents' demographics and willingness to take the COVID-19 vaccine

PDF

Mixed-method analysis for school-aged children with access and functional needs during the Covid-19 pandemic

PDF

Surgical aortic arch intervention at the time of extended ascending aortic replacement is associated with increased mortality

PDF

Preeclampsia and occurrence of neurological outcomes in the child: a meta-analysis

PDF

Survival of children and adolescents with low-risk non-rhabdomyosarcoma soft tissue sarcomas (NRSTS) treated with surgery only: an analysis of 234 patients from the Children’s Oncology Group stud...

PDF

Factors influencing the decision and timing to undergo breast reconstruction after mastectomy for breast cancer in public hospital vs. private medical center from 2007 to 2013: a retrospective co...

PDF

Fine-grained analysis of temporal and spatial differences of behavior patterns and their correlation with the spread of COVID-19 in Los Angeles County

PDF

The association of asthma and asthma-related medications on subclinical atherosclerosis: a cross-sectional analysis of four randomized clinical trials

PDF

Spatiotemporal studies of out-of-hospital cardiac arrests and bystander cardiopulmonary resuscitation in Los Angeles

PDF

Carboplatin and vincristine chemotherapy for progressive low grade gliomas in pediatric patients with or without neurofibromatosis type 1 (NF1)

PDF

Effectiveness of individual- and household-level protective actions in reducing symptoms associated with hydrogen sulfide chronic low-level exposure

PDF

Pathogenic variants in cancer predisposition genes and risk of non-breast multiple primary cancers in breast cancer patients

Asset Metadata

Creator Hu, Xinyu (author)

Core Title Characteristics associated with emergency medical services for cardiac arrest pre- and during COVID-19 in Los Angeles, CA

School Keck School of Medicine

Degree Master of Science

Degree Program Biostatistics

Degree Conferral Date 2022-05

Publication Date 04/12/2022

Defense Date 03/10/2022

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Cardiac Arrest,COVID-19,emergency medical services,OAI-PMH Harvest

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Pickering, Trevor (committee chair), Mack, Wendy (committee member), Sanko, Stephen (committee member)

Creator Email Xinyuhu@usc.edu,Xyhu97@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC110936904

Unique identifier UC110936904

Document Type Thesis

Format application/pdf (imt)

Rights Hu, Xinyu

Type texts

Source 20220412-usctheses-batch-922 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu