Keywords
cardiopulmonary resuscitation - pediatric - survival - pediatric intensive care unit
- Saudi
Introduction
The prevalence of cardiopulmonary resuscitation (CPR) in hospitalized children in
pediatric wards and pediatric intensive care units (PICU) varies from 2 to 6.[1] Alten et al reported a cardiac arrest prevalence of 3.1% among 15,908 cardiac PICU
encounters.[2] Over the past decade, a significant increase in CPR events has occurred in the PICU
compared with the wards.[3] Nevertheless, pediatric resuscitation represents an underappreciated subject, as
reported by a recent systematic analysis.[4] Moreover, the outcomes of cardiac arrest in pediatric patients depend on multiple
factors, including the quality of the CPR and postresuscitation care. Improvements
in these factors are associated with improvements in the rates of return of spontaneous
circulation (ROSC) by up to 75%.[5]
[6]
Survival to hospital discharge is an area of intense interest in the resuscitation
literature. For example, Topjian et al, in their systematic review, reported a wide
range of 25 to 50% survival and discharge from hospital in pediatric patients after
CPR.[7] Higher rates of CPR survival to discharge have also been reported in pediatric cardiac
ICUs. One report from a cardiac center showed a rate of survival to discharge as high
as 50%.[8] A collective review of pediatric CPR cases reported better outcomes for respiratory
cases, with 75% of these patients surviving to discharge compared with only 13% of
cardiac cases.[9]
[10]
CPR duration is one key factor affecting the CPR outcome. The longer the CPR is performed,
the less chance the patient has for survival. For instance, a CPR duration of more
than 15 to 20 minutes was predicted to yield a poor outcome, while a CPR duration
of less than 10 minutes was predicted to yield a better outcome.[11]
[12] However, performing CPR for more than 20 minutes is not always futile. Thus, other
key factors, such as CPR quality and postresuscitation care, are of similar importance.
Matos et al illustrated that 12% of patients undergoing CPR for more than 35 minutes
survived.[12] Therefore, continuing CPR after the 20 minutes mark may yield a favorable neurological
outcome for patients who would have died had CPR not been continued.[12] All these studies indicate that the duration of CPR is an important consideration.
However, CPR duration is not independent of other factors that also affect outcomes,
such as the CPR quality and the setting of the CPR events.
The observed differences in survival rates suggest that the outcome of CPR is a moving
target, with various factors that require further exploration. In this study, we assessed
the prevalence and outcomes of CPR in the PICU and explored the factors affecting
CPR outcomes, including sex, age, causes of cardiac arrest, diagnoses, cardiac arrest
rhythm, working hours, duration of the CPR time, and medications.
Materials and Methods
Patients and Data Collection
This retrospective cohort study was conducted in the PICU of King Khalid University
Hospital. Inclusion criteria were patients ranging in age from 0 to 14 years old who
were admitted to the PICU, underwent CPR after PICU admission, and had CPR for at
least 2 minutes.[13] Patients with incomplete data were excluded. We included all consecutive patients
who met the inclusion criteria during the study period from January 1, 2011 to January
1, 2018. Outcomes of interest were a return of spontaneous circulation (ROSC) lasting
more than 20 minutes, survival for 24-hour post-CPR, and survival to hospital discharge.
The data were retrospectively collected from the PICU CPR registry and electronic
medical records. The collected data included demographic variables such as age, weight,
and sex, and the clinical variables such as type of arrest, diagnosis, first documented
rhythm, duration of CPR, time of arrest, and medications used during CPR. The working
hours were defined as 07:00 A.M. to 16:00 P.M. during weekdays, while the remaining
hours were considered on-call hours.
Statistical Analysis
SPSS IBM v24 was used to statistically analyze the data. Means and standard deviations
were used to describe continuous variables such as age and weight, while frequencies
and percentages were employed to describe categorical variables such as sex and treatments.
The Chi-square test was used to assess the association between categorical variables.
The independent samples t-test was used to assess statistical differences in continuous variables across the
patients' dichotomous survival levels. The Cochran's Q Chi-square test was also used
to assess the association between children's repeated survival measures over time
(immediate, at 24 hours, and at discharge).
Because of the repeated survival measures and the declining survival of children in
the PICU after cardiopulmonary arrest over time, we analyzed the repeated binary survival
measures using the generalized estimating equation (GEE) logistic regression analysis.
The GEE logistic regression analysis is a repeated measures analysis of survival over
time. We employed the GEE quasi-likelihood estimation method as a default. However,
the model was further evaluated for goodness of fit by using the quasi-information
criterion and Akaike's information criterion compared with an empty model with no
predictors. The association between the analyzed predictors of children's survival
over time was expressed as an exponential distribution (β). Figures were created by
using SPSS and Excel programs. The α significance level was set to 0.05 throughout
the analysis.
Ethical Considerations
The study was initiated upon receiving approval from the institutional review board
(IRB) of the College of Medicine at King Saud University, Kingdom of Saudi Arabia.
Confidentiality of patient identifications was maintained by converting the medical
record numbers to coded numbers with no identifiers. Therefore, patient consent was
waived by the IRB.
Results
[Table 1] summarizes the demographic and other characteristics of the study cohort. A total
of 70 cases that fit the inclusion criteria were found in the PICU CPR registry. We
excluded five cases that had missing data. Thus, a total of 65 cases were included
in the analysis. Although the mean patient age was 32.7 ± 51.1 months, 40 (61.6%)
patients were younger than 12 months. The study cohort consisted of 37 (56.9%) males
and 23 (43.1%) females. Most of the presenting illnesses were respiratory (23 cases,
35.4%), and a respiratory cause of arrest was attributed to 39 (60%) cases. Furthermore,
55 (84.6%) patients were intubated and mechanically ventilated before the cardiopulmonary
arrest. The time of arrest was almost equally distributed, with 34 (52.3%) and 31
(47.7%) arrests occurring during working hours and on-call hours, respectively.
Table 1
Demographics and characteristics of children admitted to the pediatric intensive care
unit and requiring cardiopulmonary resuscitation
|
Variable
|
Frequency
|
Percentage
|
|
Sex
|
|
Male
|
37
|
56.9
|
|
Female
|
28
|
43.1
|
|
Age (mo)
|
|
Mean (±SD)
|
32.7 (51.1)
|
|
Median (Q25–Q75)
|
6 (3–35)
|
|
< 6 mo
|
33
|
50.8
|
|
> 6–12 mo
|
7
|
10.8
|
|
13–48 mo
|
11
|
16.9
|
|
> 48 mo
|
14
|
21.5
|
|
Weight (kg), mean (±SD)
|
11.10 (±13.7)
|
|
Causes of cardiac arrest
|
|
Cardiac
|
26
|
40
|
|
Respiratory
|
39
|
60
|
|
Diagnoses
|
|
CVS
|
5
|
7.7
|
|
Respiratory
|
23
|
35.4
|
|
Sepsis
|
14
|
21.5
|
|
Others
|
23
|
35.4
|
|
Pre-arrest endotracheal tube
|
55
|
84.6
|
|
Pre-arrest noninvasive ventilation
|
2
|
3.1
|
|
Time of arrest
|
|
Working hours
|
31
|
47.7
|
|
On-call hours
|
34
|
52.3
|
Abbreviations: CVS, cardiovascular system; SD, standard deviation.
The CPR details, medications used, and outcomes are shown in [Table 2]. Bradycardia was the most common cardiac rhythm during the arrest, occurring in
58 (89.2%) cases. The mean duration of CPR was 18.34 ± 15.9 minutes. While 38 (58.5%)
patients achieved ROSC post-CPR, only 30 (46.2%) patients survived for more than 24 hours,
and only 21 (32.3%) patients survived to discharge, as shown in [Fig. 1].
Table 2
Cardiopulmonary resuscitation events, medications, and outcomes of children who underwent
cardiopulmonary resuscitation in the pediatric intensive care unit
|
Variables
|
Frequency
|
Percentage
|
|
Cardiac arrest rhythm
|
|
Sinus bradycardia
|
58
|
89.2
|
|
Shockable tachyarrhythmia
|
3
|
4.6
|
|
Nonshockable dysrhythmias
|
4
|
6.2
|
|
Duration of the CPR time (min)
|
|
Mean (±SD)
|
18.3 (15.9)
|
|
Median (IQR)
|
12 (61)
|
|
Immediate CPR outcome; ROSC
|
38
|
58.5
|
|
First 24 h outcome; survived
|
30
|
46.2
|
|
Discharge final outcome; survived
|
21
|
32.3
|
|
Medication (IV bolus)
|
|
Epinephrine
|
59
|
90.8
|
|
Sodium bicarbonate
|
21
|
32.3
|
|
Atropine
|
17
|
26.2
|
|
Normal saline
|
16
|
24.6
|
|
Calcium
|
12
|
18.5
|
|
Lidocaine
|
3
|
4.6
|
|
Adenosine
|
2
|
3.1
|
|
Albumin 5%
|
2
|
3.1
|
|
Amiodarone
|
1
|
1.5
|
|
Magnesium sulfate
|
1
|
1.5
|
|
Medication (IV infusion)
|
|
Epinephrine
|
38
|
58.5
|
|
Norepinephrine
|
20
|
30.8
|
|
Dopamine
|
18
|
27.7
|
|
Dobutamine
|
13
|
20
|
|
Vasopressin
|
6
|
9.2
|
|
Phenylephrine
|
6
|
9.2
|
|
Amiodarone
|
3
|
4.6
|
|
Sodium bicarbonate
|
1
|
1.5
|
Abbreviations: CPR, cardiopulmonary resuscitation; IQR, interquartile range; ROSC,
return of spontaneous circulation; SD, standard deviation.
Fig. 1 Survival after CPR, including immediate return of spontaneous circulation, survival
for the first 24 hours after CPR, and survival at time of discharge (Chi-square = 25.5,
p-value < 0.001).
The bivariate analysis and comparison between patients with and without ROSC after
CPR revealed that a younger age was significantly associated with a better ROSC outcome
after CPR. By contrast, deceased children were significantly older (51.8 ± 60.9 months)
than children who survived (19.1 ± 38.1 months) and successfully achieved ROSC (p = 0.018). For further analysis, the patients were classified into different age intervals
to account for outliers. This classification revealed that children older than 48
months were predicted to have worse outcomes after CPR than were children within younger
age intervals (p = 0.015).
[Table 3] shows the detailed bivariate analysis and comparisons between patients with and
without ROSC after CPR. ROSC outcomes were better for patients with a respiratory
cause than with a cardiac cause of arrest; 30 (78.9%) patients who survived had a
respiratory cause of the arrest (p < 0.001). By contrast, those with sepsis were marginally less likely to survive CPR
than children with cardiovascular diseases, respiratory diseases, and other diseases.
The CPR duration was longer for deceased patients (31.2 ± 15 minutes) than for patients
who reached ROSC (9.2 ± 8.5 minutes). Treatments with a sodium bicarbonate bolus or
with norepinephrine or vasopressin infusions during CPR were significantly associated
with higher mortality and less ROSC.
Table 3
Bivariate analysis of clinical characteristics associated with an immediate revival
after cardiopulmonary resuscitation in the pediatric intensive care unit
|
(Outcome) ROSC
|
|
|
|
Deceased
n = 27
|
Revived
n = 38
|
Test statistic
|
p-Value
|
|
Sex
|
|
Male
|
15 (55.6%)
|
22 (57.9%)
|
X2(1) = 0.04
|
0.851
|
|
Female
|
12 (44.4%)
|
16 (42.1%)
|
|
|
|
Age (mo), mean (±SD)
|
51.8 (60.9)
|
19.1 (38.1)
|
t (40.3) = 2.5
|
0.018
|
|
≤6 mo
|
11 (40.7%)
|
22 (57.9%)
|
X2(3) = 10.41[*]
|
0.015
|
|
> 6–12 mo
|
2 (7.4%)
|
5 (13.2%)
|
|
|
|
13–48 mo
|
3 (11.1%)
|
8 (21.1%)
|
|
|
|
> 48 Months
|
11 (40.7%)
|
3 (7.9%)
|
|
|
|
Weight (kg), mean (±SD)
|
16.5 (19)
|
7.3 (6.5)
|
t(28.9) = 2.40
|
0.025
|
|
Causes of cardiac arrest
|
|
Cardiac
|
18 (66.7%)
|
8 (21.1%)
|
X2(1) = 13.7[*]
|
<0.001
|
|
Respiratory
|
9 (33.3%)
|
30 (78.9%)
|
|
|
|
Diagnoses
|
|
CVS
|
2 (7.4%)
|
3(7.9%)
|
X2(3) = 7.60[*]
|
0.056
|
|
Respiratory
|
6 (22.2%)
|
17 (44.7%)
|
|
|
|
Sepsis
|
10 (37%)
|
4 (10.5%)
|
|
|
|
Others
|
9 (33.3%)
|
14 (36.8%)
|
|
|
|
Cardiac arrest rhythm
|
|
Sinus bradycardia
|
22 (81.5%)
|
36 (94.7%)
|
X2(2) = 2.93[*]
|
0.231
|
|
Shockable tachyarrhythmia
|
2 (7.4%)
|
1 (2.6%)
|
|
|
|
Nonshockable dysrhythmias
|
3 (11.1%)
|
1 (2.6%)
|
|
|
|
Pre-arrest intubation
|
26 (96.3%)
|
29 (76.3%)
|
χ2(1) = 3.43[**]
|
0.064
|
|
Pre-arrest noninvasive ventilation
|
1 (3.7%)
|
1 (2.6%)
|
χ2(1) <0.001[**]
|
1 = NS
|
|
Time of arrest
|
|
Working hours
|
10 (3%)
|
21 (55.3%)
|
χ2(1) = 2.10
|
0.147
|
|
On-call hours
|
17 (63%)
|
17 (44%)
|
|
|
|
Duration of the CPR time (min)
|
31.2 (15)
|
9.2 (8.5)
|
t (37.9) = 6.9
|
<0.001
|
|
Medications
|
|
Epinephrine IV bolus
|
25 (92.6%)
|
34 (89.5%)
|
χ2(1) < 0.001[**]
|
NS
|
|
Lidocaine IV bolus
|
1 (3.7%)
|
2 (5.3%)
|
χ2(1) < 0.001[**]
|
NS
|
|
Sodium bicarbonate IV bolus
|
14 (51.9%)
|
7 (18.4%)
|
χ2(1) = 8.1
|
0.005
|
|
Calcium chloride/gluconate IV bolus
|
8 (29.6%)
|
4 (10.5%)
|
χ2(1) = 2.70
|
0.103
|
|
Bolus amiodarone
|
1 (3.7%)
|
0
|
χ2(1) = 0.030[**]
|
0.862
|
|
Atropine IV bolus
|
7 (25.9%)
|
10 (26.3%)
|
χ2(1) < 0.001[**]
|
NS
|
|
Adenosine bolus
|
1 (3.7%)
|
1 (2.6%)
|
χ2(1) < 0.001
|
NS
|
|
Magnesium sulfate IV bolus
|
1 (3.7%)
|
0
|
χ2(1) = 0.030
|
0.863
|
|
Norepinephrine infusion
|
15 (55.6%)
|
5 (13.2%)
|
χ2(1) = 13.3
|
<0.000
|
|
Dopamine infusion
|
10 (37%)
|
8 (21.1%)
|
χ2(1) = 2.01
|
0.156
|
|
Vasopressin infusion
|
6 (22.2%)
|
0
|
χ2(1) = 6.004[**]
|
0.009
|
Abbreviations: CVS, cardiovascular system; CPR, cardiopulmonary resuscitation; ROSC,
return of spontaneous circulation; SD, standard deviation.
* Likelihood ratio adjusted Chi squared test.
** Chi squared test with Yates correction.
The GEE logistic regression analysis was conducted by setting the patients as subjects
and the time as a repeated measure, after transposing the data into the long format
yielding 65 × 3 = 195 patient records. The analysis revealed that bradycardia, duration
of CPR, and absence of norepinephrine infusion were significantly associated with
higher accumulative survival until discharge ([Table 4]). Moreover, the analysis model showed that CPR duration predicted significantly
shorter survival: an increase in the CPR duration of 1 minute decreased the children's
cumulative predicted survival by an average of 8.7% (p = 0.001). The GEE analysis also showed that an increase in the CPR duration from
3 to 6 minutes resulted in a nearly steady predicted survival; however, survival declined
for CPR durations of 7 to 9 minutes and plunged more rapidly for CPR durations of
10 minutes or more. This negative association between CPR duration and survival was
statistically significant when considering the other factors in the model ([Fig. 2]).
Table 4
Generalized estimating equation model explaining repeated survival rate from immediate
post-CPR until discharge time for pediatric intensive care unit patients who underwent
CPR (n = 195)
|
Parameter
|
B
|
Standard error
|
Adjusted odds ratio
|
95% CI
|
p-Value
|
|
Lower
|
Upper
|
|
(Intercept)
|
−2.635
|
0.8718
|
0.072
|
0.013
|
0.396
|
0.003
|
|
Female (sex)
|
1.161
|
0.6213
|
3.192
|
0.945
|
10.788
|
0.062
|
|
Age (mo)
|
0.004
|
0.0065
|
1.004
|
0.991
|
1.017
|
0.552
|
|
Cause of arrest
|
−0.697
|
0.5135
|
0.498
|
0.182
|
1.363
|
0.175
|
|
Bradycardia rhythm
|
1.618
|
0.6859
|
5.045
|
1.315
|
19.353
|
0.018
|
|
Shockable tachycardia
|
−0.756
|
0.8912
|
0.470
|
0.082
|
2.693
|
0.396
|
|
Time working arrest
|
0.228
|
0.5591
|
1.256
|
0.420
|
3.758
|
0.684
|
|
Duration of CPR
|
−0.091
|
0.0282
|
0.913[a]
|
0.864
|
0.965
|
0.001
|
|
No NaHCO3 bolus
|
−0.367
|
0.6318
|
0.693
|
0.201
|
2.390
|
0.561
|
|
No norepinephrine infusion
|
2.955
|
0.5162
|
19.211
|
6.986
|
52.833
|
<0.001
|
Abbreviations: CI, confidence interval; CPR, cardiopulmonary resuscitation; ROSC,
return of spontaneous circulation; SD, standard deviation.
a (0.913 times less accumulated survival = (1 − 0.913) × 100 = 8.7% less; as the CPR
duration increased by 1 minute, the children's cumulative predicted survival declined
by an average of 8.7%.
Fig. 2 The association between CPR duration (minutes) and the generalized estimating equation
model predicted survival rate (probability) over time. CPR, cardiopulmonary resuscitation.
Discussion
This retrospective study involved PICU patients and included 65 documented instances
of at least 2 minutes of CPR. The purpose of the study was to assess the frequency,
survival, and variables affecting CPR outcomes at a PICU. In our study, 38 (58.5%)
patients achieved ROSC, which is consistent with two previous studies.[14]
[15] Our examination of 24-hour survival also revealed that 30 (46.2%) patients reached
24-hour survival. This value was similar to the 58.4% 24-hour survival rate reported
in a multicenter study investigating survival trends on weekends compared with weekdays,
but it was slightly lower than the 64% survival in PICU patients 24 hours after CPR
reported in another multicenter study conducted in the United States.[12]
[13] The contrasting results between the survivals to discharge in these studies may
be due to multiple factors. One explanation might be the lack of reporting data related
to CPR quality and postcardiac arrest care. Differences in the inclusion criteria
could also contribute to these varying outcomes. For instance, we excluded patients
who had undergone CPR for less than 2 minutes, and this exclusion might have lowered
the survival to discharge reported by us and by Bhanji et al,[13] compared the rate reported by Matos et al.[12]
In our study, 21 (32.3%) patients survived to discharge. This rate was slightly below
the rates reported in three previous studies, which showed survival to discharge rates
of 45, 36.2, and 34.8% after CPR in PICUs.[1]
[11]
[13] Variations in survival rates to discharge are expected, as differences in diagnoses,
severity, comorbidity, and age can influence the survival to discharge rates. Once
again, the differences in inclusion criteria and the lack of consistent reporting
of CPR quality could contribute to the discrepancy in results seen in the reported
survival rates.[16]
[17]
[18] High-quality CPR is the first take-home message of the American Heart Association
(AHA) in their 2020 Pediatric Basic and Advanced Life Support update.[19] The literature confirms that the key components of high-quality CPR are applying
adequate chest compression rate and depth, minimizing CPR interruptions, allowing
full chest recoil between compressions, and preventing excessive ventilation.[19]
We noticed that ROSC was higher among our patients who were less than 6 months of
age, as 22 (57.9%) patients under 6 months of age achieved ROSC versus 16 (42.1%)
patients older than 6 months. This is consistent with a significantly higher survival
rate in infants and neonates (53.2 and 62%, respectively) compared with older children
(29.8–41.2%) reported in a recent study conducted in the United States.[20] Another recently published study on CPR epidemiology in PICUs across England revealed
similar results.[21] By contrast, worse outcomes were reported for infants than for older children following
out-of-hospital CPR.[22]
[23]
[24] Infants have less respiratory reserve and are therefore likely to have a respiratory
cause for CPR. Thus, any delay in recognizing or providing respiratory support during
out-of-hospital CPR can be detrimental, whereas patients in the PICU are carefully
monitored and are less likely to have a delay in respiratory support. Therefore, lower
survival rates were correlated with the advancing age of patients who underwent CPR
in the PICU setting compared with out-of-hospital CPR.
Our study showed that outcomes were significantly better when the cause of arrest
was respiratory related. A majority of patients (78.9%) who underwent CPR and survived
had a respiratory cause of arrest. This is compatible with results from another study
showing a better survival outcome for pediatric patients with respiratory arrest (64.9%)
than with cardiac-related arrest.[25] The better outcomes in patients with respiratory-related arrest may be due to a
shorter CPR duration or a better tolerance for hypoxia compared with a cessation of
perfusion. Moreover, better survival may result from airway maneuvers that enhance
oxygenation by recruiting functional alveolar segments.[25]
The use of medication during CPR was an important aspect of our study. Our aim was
to identify the most frequently used medications and their effects on CPR outcomes.
Only vasopressin, sodium bicarbonate, and norepinephrine were significantly associated
with inferior outcomes, similar to the findings of other studies.[15]
[26] This was partially due to the use of these medications as a last attempt at saving
a patient, as well as the ambiguous indications for these medications during CPR.
However, the use of epinephrine and atropine boluses was reported in an earlier study
to improve the survival rate of patients.[9]
Our results demonstrated a mean CPR duration of 9.2 minutes for patients who achieved
ROSC, while the mean CPR duration for patients that did not survive was 31.2 minutes.
A previous study showed that the first 15 minutes of CPR had the best outcomes, with
a 41% probability of survival. Alternatively, prolonged CPR that extended beyond 35 minutes
had only a 12% probability of survival, indicating that survival decreased by 2.1%
per 1 minute of CPR.[12] Thus, CPR duration is inversely related to favorable outcomes. Our study showed
similar findings, as the CPR duration was independently associated with survival to
discharge.
The GEE model suggests a downward trend in ROSC after a CPR duration of 7 to 9 minutes,
which may reflect prolonged oxygen deprivation and low perfusion during the arrest.
Another explanation for the inverse relationship between CPR duration and survival
is the waning effect of epinephrine after three to five doses, as some preclinical
models have suggested.[27] Furthermore, in contrast to adult and pediatric out-of-hospital studies, the time
until defibrillation, when a shockable rhythm was detected, did not correlate with
outcomes in the setting of in-hospital cardiac arrest.[28] CPR performed in a perioperative setting shares similar CPR contexts with the PICU
setting.[29]
The diagnosis of pediatric patients also had a significant impact on the outcome of
CPR. In our research, favorable CPR outcomes were significantly associated with respiratory
diagnoses, as 17 (73.9%) patients with a respiratory diagnosis on PICU admission achieved
ROSC. Cardiac diagnoses are usually due to congenital diseases and are more chronic
in nature than respiratory diagnoses, which are commonly associated with infectious
causes. Furthermore, our sample size for cardiac diagnosis was small, as only five
(7.69%) patients had a cardiac diagnosis. Other studies have demonstrated better survival
outcomes after CPR for pediatric patients with cardiac diagnoses. The difference between
our study and others may reflect the larger sample sizes and higher numbers of cardiac
arrest cases with cardiac diagnoses in the other studies.[9]
[30]
A recent study compared CPR survival rates after pediatric cardiac arrest during nights
and weekends to CPR survival during day shifts and weekdays in a large database that
included 12,404 children who received CPR in 354 hospitals. The survival rate for
pediatric patients undergoing CPR was lower during the night shifts than in the day
shifts, although no clear difference was noted between weekends and weekdays.[13] Other recent studies have also demonstrated a marked drop-off in immediate ROSC
and survival to discharge in patients who underwent CPR on weekends and during night
shifts compared with weekdays and day shifts.[31]
[32] However, our research found no statistically significant differences between patients
who underwent CPR during regular working hours and patients who underwent CPR during
on-call hours. The differences in our study and others may be due to each hospital's
staff management, which includes adequate staff numbers during both shifts, and the
24-hour in-house presence of a senior PICU team member. However, the decline in CPR
quality during the night, which is attributed to fatigue, has been an area of concern.
Although reports evaluating children's blood pressure during chest compression suggested
a decline in CPR quality at night, these findings were not consistent across reports.[33]
[34]
[35]
We found that the presence of an endotracheal tube and mechanical ventilation pre-arrest
were marginally correlated with a lower chance of survival, in line with other published
papers.[12]
[31]
[36] Patients who develop cardiac arrest despite receiving respiratory assistance may
have a worse prognosis due to their underlying disease severity. Moreover, the concept
of reperfusion injury due to free oxygen radicals with the application of 1.0 FiO2 deserves further study.[37]
[38]
The decreased survival rate post-ROSC needs to be integrated into parental counseling.
During a child's resuscitation, the parents may perceive overwhelming chaos, so having
a more informative prognosis during this stressful period could improve the parental
expectations.[39] While emotional support was appreciated by parents who witnessed their child's resuscitation,
the most important factor was that they received real-time clinical information from
their health care providers. Better expectations could help the parents make more
realistic decisions and could even change their attitudes toward end-of-life from
life support preferences.[40]
This study reflects the importance of continuous assessment of the progress and impact
of changes in CPR practices and outcomes. More accumulation and sharing of reports
would aid in the development of better guidelines that are generalizable to all international
communities.
Limitations
Studies on pediatric resuscitation are difficult to perform and frequently include
small numbers of patients.[29] They are mostly retrospective and based on registries or administrative data, rather
than being prospective and well-designed studies, so they are vulnerable to the challenges
of missing data and potential bias.[11] However, the pediatric CPR is relatively a rare event and the accumulation of more
reports, despite aforementioned limitation, is important for further progress in pediatrics
CPR studies. The results of our study should be interpreted with caution due to the
limitations of the study, including its retrospective and single-center design. The
multivariate analysis in our study revealed interesting findings; however, the fairly
large number of comparisons relative to the sample size needs to be interpreted with
precaution. The assessment of CPR quality components was difficult to retrieve retrospectively
and was beyond the study scope, but it warrants validation in future research. Our
PICU adheres to the AHA Guidelines for CPR for pediatric advanced life support (PALS),
and it undergoes ongoing revisions to target key questions related to pediatric resuscitation.
The PALS Guidelines, which were revised by the AHA in 2015, were published in the
“American Heart Association Guidelines Update for Cardiopulmonary Resuscitation and
Emergency Cardiovascular Care.” These updates are usually applied in real-life settings
within a year of their announcement and may affect CPR studies that overlap with these
updates.[1] We also believe that additional prospective data regarding the neurological outcomes
of surviving patients could have added more insights to the outcome. However, the
neurological outcome was not considered in the study scope. Nevertheless, this study
considered relevant and meaningful primary outcomes, with potential expansion to include
neurological outcomes in future updates. Although the duration of a CPR event could
be difficult to validate in the context of an emergency event, the documentation of
CPR in a specifically designed form was quite helpful for tracking, in combination
with monitoring the timeline print attached to the form.
Conclusion
In the PICU setting, better CPR outcomes were associated with younger age, a respiratory
cause of arrest, a bradycardia rhythm, and a shorter CPR duration. The duration of
CPR remains one of the crucial factors for CPR outcomes and needs to be considered
in parallel with the guidelines emphasizing CPR quality. The decreased survival rate
post-ROSC needs to be carefully considered during parental counseling, while better
anticipation and prevention of CPR remain ongoing challenges. A multicenter national
registry of CPR outcomes would further improve and widen the research scope.
