Filipino CPR Outcomes Study 2015
Filipino CPR Outcomes Study 2015
2 75
July 2015 to December 2015
Abstract Cardiopulmonary resuscitation (CPR) saves lives, but proper techniques must be taught and
practiced. Outcomes have to be assessed to improve patient survival.
Methods This is an observational, analytical, prospective study of CPR in adult cardiac arrest cases from
selected tertiary hospitals in the Philippines. It describes the characteristics and clinical factors
associated with CPR survival in the Philippines based on the American Heart Association 2010
guidelines. Validated Philippine Heart Association hospital cardiac arrest collection forms were
used to gather data on the demographics, pre-arrest and CPR/code characteristics of the cases.
The investigators analyzed survival outcomes from first arrest to initial return of spontaneous
circulation (ROSC) to hospital discharge. Epi Info 6 and STATA 11 statistical software were used
for descriptive and inferential statistics, while the chi-square test was used for the association of
76 The Filipino Cardiopulmonary Resuscitation Outcomes Study PJC Vol. 43 No.2
PHA CPR Council et al July 2015 to December 2015
variables (significant P-value set at <0.05). Stratified analysis and logistic regression analysis were
done for confounders. Odds ratios (OR), 95% CI, and p-values were reported.
Results Out of 156 patients, 101 (65%) survived up to ROSC and hospital discharge. The investigators
found significant differences in patient survival under the following factors: CPR medications
(OR=16.729; 95% CI 1.979, 141.375; P=0.0097); and post-arrest therapeutic hypothermia
(OR=8.745; CI 1.826, 41.882; P=0.0067). There were some decision errors on the use of
defibrillation for shockable rhythms (26%) and non-shockable rhythms (14.5%). Eighteen rhythms
were not properly identified.
Conclusion Decreased interval time from arrest to intubation and from arrest to call for code, CPR medications,
and post-arrest therapeutic hypothermia were the most significant contributors to CPR survival
in this study. Future training should address the errors in cardiac rhythm interpretation and
defibrillator use.
amplitude, or a pulseless rate of <10 beats per 12. Presumed cause of arrest: an arrest is presumed
minute. to be of cardiac etiology unless it is known or
3. Bystander CPR: CPR performed by a person likely to have been caused by any of the following:
who is not from an organized emergency response trauma, submersion, drug overdose, asphyxia,
system or team (physicians, nurses and paramedics exsanguination, or any other non-cardiac cause as
may be considered bystanders when they are off best determined by the rescuers.
duty from the emergency response system). 13. ROSC: return of spontaneous circulation; the post-
4. CPR: any effort to restore effective ventilation, CPR presence of a palpable pulse or measurable
oxygenation and circulation to any patient by the blood pressure sustained for at least 20 consecutive
use of any or all of the following methods: chest minutes without need for chest compressions;
compressions, airway interventions, defibrillation, other signs of ROSC are breathing, movements,
cardiac pacing or intravenous medications. and coughing.
5. Code team/CPR team: the hospital team that 14. Shockable/non-shockable rhythm: a monitored
conducts the emergency response system for CPR; rhythm that is interpreted by either a rescuer or
usually composed of team captain, defibrillation an automated external defibrillator (AED) to be
operator, IV nurse, data recorder, compressor, and treatable by electric defibrillation. In general,
intubator. shockable cardiac arrest rhythms are divided into
6. Dead on arrival (DOA): patients who arrive at VF and pulseless VT. Nonshockable cardiac arrest
the emergency room (ER) without any signs of rhythms can be categorized as either asystole or
spontaneous circulation as pronounced by an ER pulseless electrical activity (PEA).
physician. 15. Survival at hospital discharge: when the patient
7. Do not attempt resuscitation (DNR) status: is alive upon discharge from the hospital, regardless
when CPR is not attempted in a patient who of neurological status, outcome, or destination.
had given prior written or verbal instruction – 16. Survival to ROSC: ROSC for >20 minutes after
either personally or through a legally authorized the first code, regardless of succeeding arrests or
representative (LAR)7 – that no mechanical, eventual death of the patient.
electrical, or pharmacological intervention shall be 17. Time from arrest to arrival of the code team:
given in case of arrest. interval in minutes between the time of cardiac
8. First code: the initial recognition of a patient’s arrest and the time of arrival of the complete code
cardiac arrest that precipitates a call for the code team.
team and subsequent CPR intervention. 18. Time from arrest to call for code: interval in
9. First monitored rhythm: the first cardiac rhythm minutes between the time of cardiac arrest and the
present when a cardiac monitor, with or without a time the first code was called.
defibrillator, is attached to the patient after arrest. 19. Time from arrest to first defibrillation: interval
10. Therapeutic hypothermia: also called targeted in minutes between the time of cardiac arrest and
temperature management, refers to the deliberate the time of first defibrillation.
reduction of the core body temperature – 20. Time from arrest to intubation: interval in
typically to a range of about 32°C to 34°C minutes between the time of cardiac arrest and the
(89.6°F to 93.2°F) – in patients who do not time of successful intubation.
regain consciousness after return of spontaneous 21. Time from arrest to first monitored rhythm:
circulation (ROSC) following a cardiac arrest. interval in minutes between the time of cardiac
11. Pre-hospital CPR: the administration of chest arrest and the time the patient was hooked to a
compressions, with or without ventilation, on cardiac monitor.
a patient who had an arrest at any time before 22. Time from arrest to start of CPR: interval in
arriving at a hospital for further management. minutes between the time of cardiac arrest and
78 The Filipino Cardiopulmonary Resuscitation Outcomes Study PJC Vol. 43 No.2
PHA CPR Council et al July 2015 to December 2015
the time when CPR with chest compressions Whenever code was called, one of the co-investigators
commenced. explained to the patient’s LAR7 the necessity of CPR for the
23. Total code duration: interval in minutes from the patient. They also explained that an observational study
time the code was sounded to the time that code was being conducted on all CPR patients and requested the
ended. LAR to sign the informed consent form within 24 hours
24. Witnessed arrest: cardiac arrest that was seen or from the time of code. All cardiac arrest adult patients (≥18
heard by any person as it occurred. years old) whose LAR consented to the study and signed the
informed consent form were included in the study. Patients
Methods who were DOA or with definite clinical signs of death (eg,
This was a prospective analytical study conducted in lividity, rigor mortis) when they came into the emergency
selected tertiary hospitals where the PHA had previously room, as well as those who had signed DNR directives, were
conducted CPR training among the medical personnel. excluded.
The total population of the areas serviced by these hospitals CPR was performed following AHA 2010 CPR
approximated 26,532,071 based on the Philippines’ guidelines. All pertinent data were recorded by a co-
National Statistics Circular Bulletin of 2010. Using the investigator in the PHA hospital cardiac arrest collection
Raosoft sample size calculator, we determined that the form within 24 hours from CPR. Patient records were
minimum sample size required for a 95% confidence level collected from all wards, intensive care units and emergency
(CI), 3% margin of error, and 50% response distribution is care units. For patients who had undergone multiple
1,068 patients. However, in November 2015, we abruptly codes, only the first code was used in this study. All forms
ended data collection after the AHA 2015 CPR guidelines were checked for completeness before submission to the
became publicly available. At that point, we had been able PHA CPR study team for encoding, data analysis and
to collect and analyze a total of 156 patient data forms. interpretation.
Before we started the study, technical review and The following hospitals contributed to the study
certification was given by the Research Committee of the data: Cardinal Santos Medical Center, Chinese General
PHA, while ethical review and certification was given by Hospital, Chong Hua Hospital, Manila Doctors Hospital,
the Ethics Review Committee of each participating hospital. Mary Mediatrix Medical Center–Lipa, Perpetual Succour
The informed consent form was made available in English, Cebu Hospital, University of Perpetual Help Dalta Medical
Filipino and Cebuano. The study was ethically approved Center–Las Pinas, The Medical City, the University of
based on the guidelines of the Philippine Health Research Santo Tomas Hospital and Zamboanga General Hospital.
Ethics Board 20117 and the World Medical Association This study used Epi Info 6 and STATA 11 statistical
Declaration of Helsinki–Ethical Principles for Medical software to generate descriptive and inferential statistics.
Research Involving Human Subjects 2013.8 The data The frequency distribution, mean and standard deviation
of the cases were recorded in the PHA hospital cardiac of the demographic, clinical and outcome variables of the
arrest collection form, which was adapted from the AHA- study were generated. Chi-square test with a significant
validated and AHA-recommended Utstein template for P-value of .05 was used to determine the association of
resuscitation registries.9 Each participating hospital was variables. Risk factors confounding the association between
given an institution code for reporting. The results of all CPR and outcome variables were controlled using stratified
hospitals were reported as a whole. analysis and logistic regression analysis. The OR, 95% CI,
Prior to data gathering, an orientation was given to and P-values were reported.
all study facilitators, research assistants, investigators,
and CPR practitioners to ensure uniformity of both the Results
comprehension and the execution of the study. Emphasis Table 1 shows the demographic, pre-hospital and
was given on accuracy of documentation in the PHA CPR characteristics of the patients. There was a near-equal
hospital cardiac arrest collection form, especially in terms of distribution of males and females. The youngest patient was
durations, intervals and patient status at discharge. 18 years old; the oldest was 93. The mean age of patients
PJC Vol. 43 No.2 The Filipino Cardiopulmonary Resuscitation Outcomes Study 79
July 2015 to December 2015 PHA CPR Council et al
Table 1. Frequency Distribution of Patient Demographics, Pre- The code teams were usually composed of medical
hospital and CPR/Code Characteristics.
residents, cardiac fellows, and nurses. Eighty-seven percent
Characteristics
of those who were reported to have done cardiac compression
(N=156) n %
Sex
had undergone BLS/ACLS certification training. Data entry
Female 73 46.8 regarding code team members was inconsistent. Although
Male 83 53.2 most of the reports listed down all the personnel involved,
Age, ya some only recorded the code team members who were
≤40 32 20.5
41–60 45 28.8 doctors and nurses. No orderlies were ever reported as code
61–80 63 40.4 team members.
≥81 16 10.3 Most of the arrests were presumed to be due to
CPR
cardiac- and respiratory-related diseases, followed closely by
Pre-hospital 10 6.4
In-hospital 146 93.6 neurologic disorders.
Type of witnessb Table 2 describes the variables related to the conduction
None 8 5.1 of a code. Eighty-one percent of the patients had airway
Lay person 11 7.1
Health care provider 137 87.8
intubation; the rest did not. This study highlights the use
Previous arrest countc of epinephrine, atropine and amiodarone more than other
0 144 92.3 medications because these three drugs are recommended
1 11 7.1 in all algorithms of the ACLS guidelines and are readily
2 1 0.6
available in all tertiary hospitals of the Philippines.
Code team composition
Consultant 8 5.1 Management of rhythm abnormalities are described
Nurse 34 21.8 in Table 3. Upon initial cardiac monitoring, 99 of 156
Fellow 35 22.4
Intern 3 1.9
patients (63.5%) were in asystole, 14 (9.0%) had VF, 14
Medical resident 125 80.1 (9.0%) had VT, and 11 (7.1%) had PEA. Eighteen rhythms
Orderly 0 0.0 were not properly recorded. Included in the PHA BLS/
Presumed cause of arrest ACLS training is the recognition of shockable (VF/VT)
Cardiac 58 37.2
Respiratory 48 30.8 and non-shockable (asystole/PEA) rhythms. This study
Neurologic 23 14.7 revealed errors in interpretation and management by the
Trauma 2 1.3 code members: defibrillation was given to 16 episodes of
Surgery 0 0.0
Oncology 1 0.6 non-shockable rhythms (15%) and not given to 12 episodes
Others 24 15.4 of shockable rhythms (26%).
CPR=cardiopulmonary resuscitation. Only 13 (8.3%) of the patients had post-arrest
a
Mean age±standard deviation = 59 years old; range = 18 to 93.
therapeutic hypothermia (Table 4). In this study, the use
b
Pre-hospital or in-hospital.
c
Excluding current admission. of therapeutic hypothermia for post-arrest patients was
consistently a significant factor for survival when associated
was 59 years. Most arrests occurred in the 61–80 age with successful ROSC after first CPR and with survival
range, corresponding with the World Health Organization upon discharge from hospital (OR=8.7; 95% CI 1.83,
(WHO) 2013 life expectancy census for Filipinos, which 41.88; P=0.0067). The use of epinephrine, atropine, and
shows 65.4 years in males and 72.5 years in females.10 amiodarone as CPR medications also significantly improved
Ten of the 156 arrests (6%) were reported to have had initial survival to ROSC and up to hospital discharge
pre-hospital CPR. Most of the CPR cases that we were (OR=16.73; 95% CI 1.98, 141.38; P=0.0097). Pre-hospital
able to record and describe were in-hospital cases and were CPR had a statistically significant association with harm in
witnessed by health personnel. In most of our recorded our study (OR=0.13; 95% CI 0.02, 1.00; P=0.0495).
cases, it was the first time that CPR was performed on the Code time intervals for survival to ROSC are reported
patient, with the exception of 12 patients who had had in Table 5.
previous episodes as attested by informants. After eliminating the data of patients who died after
80 The Filipino Cardiopulmonary Resuscitation Outcomes Study PJC Vol. 43 No.2
PHA CPR Council et al July 2015 to December 2015
Table 2. Frequency Distribution of CPR/Code Variables of the 1,068 patients needed to attain conclusive results because
Patients in the Study.
the gathering of data was abruptly stopped in November
n %
2015, when the new AHA 2015 CPR guidelines6 were
Intubation
No intubation 20 12.0
made public. As small as the observed population may be, it
Airway tube 126 80.8 revealed several facts that are consistent with international
Laryngeal mask airway 5 3.2 data. The survival rate to ROSC of 65% was comparable to
Nasopharyngeal 5 3.2
the 54% result of the study by Peters et al.3 On the other
Initial rhythm
Shockable hand, the 100% survival rate up to hospital discharge that
Ventricular fibrillation 14 we found in our study was far from the 5%–24% reported
Ventricular tachycardia 14 9.0 by previous studies.
Non-shockable 9.0
Asystole 99 Only 10 of the 156 cardiac arrests in this study were
Pulseless electrical activity 11 63.5 reported to have pre-hospital CPR. This low figure was
Not specified/others 18 7.1
expected because most Filipinos do not have the proper
11.5
Cardiac compression performer
training for performing BLS. We believe that more arrests
BLS-trained 21 13.5 occur pre-hospital, but either they were not reported or
BLS/ACLS-trained 135 86.5 the patients never reached the hospitals for definitive
Medication/procedures medical management. Pre-hospital CPR had a statistically
Epinephrine 124 79.5
Atropine 14 9.0 significant negative effect on survival in our study; this may
Amiodarone 10 6.4 have been caused by the small sample size, since the opposite
Therapeutic hypothermia 13 8.3
has already been established in previous studies.
CVP line insertion 1 0.6
Tracheostomy 1 0.6 This study revealed multiple errors in interpretation
ACLS=advanced cardiac life support; BLS=basic life support; and management by the code members: Defibrillation was
CPR=cardiopulmonary resuscitation; CVP=central venous pressure. given to 15% of non-shockable rhythms and withheld in
26% of shockable rhythms. These figures may be higher
first CPR, we were left with 101 surviving patients (65%). since we were unable to interpret whether the 18 unrecorded
All these patients survived to hospital discharge, thereby rhythms necessitated defibrillation or not. While these
cancelling the need for further computations. results may be explained by erroneous data entry, these
should also encourage the PHA CPR training team to put
Discussion more emphasis on accurate interpretation and management
The subjects for this study are limited to 156 (15%) of of cardiac rhythms in forthcoming training workshops.
Table 4. Association of Demographic, Pre-hospital and CPR/Code Characteristics with CPR Survival Up to ROSC.
95 % CI
Characteristics OR Lower Upper P-value*
Age 1.0084 0.9890 1.0281 0.4004
Sex 0.8009 0.3857 1.663 0.5515
CPR attempt pre-hospital 0.1395 0.0195 0.9953 0.0495
Resuscitation 5.5404 0.4771 64.337 0.1712
Arrest cause 0.9161 0.7609 1.1029 0.3547
Medications 16.7258 1.9788 141.3752 0.0097
Therapeutic hypothermia 8.7452 1.8261 41.8822 0.0067
Total code duration 0.9928 0.9785 1.0073 0.3295
*P<0.05 is considered statistically significant.
Table 5. Multiple Logistic Regression Analysis of Code Time Intervals from Arrest to ROSC.
Consistency of data entry regarding code team members to maintain such survival figures as we register more CPR
was an issue during the study. While some of the reports outcomes in succeeding studies.
listed all the personnel involved, others only recorded the
doctors and the nurses. None of the reports listed orderlies Conclusion
as code team members; this is unusual because orderlies are This study, which is based on the AHA 2010 CPR
considered vital alternative staff for doing chest compressions guidelines, is a landmark prospective study on CPR. It
in most Filipino hospitals. validates post-arrest therapeutic hypothermia and CPR
One consistent significant factor that we found in this medications as variables that can improve survival. It also
study was the use of therapeutic hypothermia for post- demonstrates that further training in cardiac rhythm
arrest patients. Numerous studies have already validated recognition and the correct use of the defibrillator must be
improvement of survival with this procedure, but it is not emphasized in future CPR training courses.
yet widely available in Filipino hospitals.
The variables for apparent CPR success in our study were Acknowledgements
multifactorial. The positive outcomes may be attributed to This research was fully supported by the Philippine
the quality of ACLS/BLS training given by the PHA, but Heart Association. We wish to thank the following PHA
they may be also be due to the innate competency of the code advisers for their valuable inputs to this project: Alisa
teams and the availability of CPR instruments and ACLS Bernan, MD (Chair, PHA Research Committee); Marcellus
medications in the various institutions. We recommend the Francis Ramirez, MD; Jude Erric Cinco, MD; Myrna Dela
installation of post-arrest therapeutic hypothermia units in Cruz; and the PHA secretariat.
all tertiary hospitals in the country. It will be a challenge
82 The Filipino Cardiopulmonary Resuscitation Outcomes Study PJC Vol. 43 No.2
PHA CPR Council et al July 2015 to December 2015