1. Introduction

The American College of Cardiology/American Heart Association (ACC/AHA) 2020 revised guidelines separated the chain of survival for in-hospital cardiac arrest (IHCA) and out-of-hospital cardiac arrest (OHCA) [1]. This separation was necessary because the circumstances surrounding IHCA and OHCA are extremely different [2]. OHCA survival depends on early recognition of the event, early initiation of high-quality cardiopulmonary resuscitation (CPR) by the by-stander, appropriate early defibrillation, effective advanced cardiovascular resuscitation procedures, such as drug administration and advanced airway maintenance, and integrated post-cardiac arrest care [3, 4, 5]. For IHCA, the chain of survival is supposedly well established; hence, a high survival rate and good prognosis can be expected. However, survival-to-discharge rates and prognosis of IHCA are different from those of OHCA. The survival-to-discharge rate for OHCA and IHCA are estimated to be about 10% and 12-25%, respectively [6, 7, 8, 9]. Many IHCA cases involve patients suffering from severe underlying diseases with a low resuscitation potential; 80% of the patients are usually physiologically and clinically unstable [10, 11].

In Korea, the Korea Centers for Disease Control and Prevention research has piloted studies to improve the survival rate of patients with acute cardiac arrest since 2008, and there is a report on survival analysis and prognosis after acute cardiac arrest [12]. However, most of the studies are limited to the analysis of OHCA and cardiac arrest at emergency medical center levels [13].

At each general tertiary hospital, for the safety and quality management of patients and the evaluation of certification by medical institutions, the medical staff’s ability to perform CPR and response systems are monitored. CPR start time after cardiac arrest, time to CPR team activation, communication system in case of cardiac arrest, defibrillator management, CPR survival rate, and cerebral performance categories (CPC) score are reported as performance indicators. However, information sharing and reporting of results at the academic level are completely absent.

Therefore, our study aimed to assess the prognosis of patients by analyzing the CPR team activation time and CPR time after cardiac arrest in IHCA to determine factors affecting survival to discharge, which reflect the neurological prognosis.

2. Methods

This study protocol was reviewed and approved by the Institutional Review Board of our hospital (approval no. 05-2020-163).

3. Results

3.1 Baseline characteristics

In total, there were 209,955 (average of 39,990 per year) hospitalized patients during the period, and 4,227 patients died during hospitalization, with an average of 805 deaths per year. During the study period, a total of 767 patients aged $\ge$ 18 years received CPR for cardiac arrest during hospitalization. Of these, 14 patients developed cardiac arrest in the operating room, while 20 patients developed cardiac arrest during other procedures, including coronary angiography. There were 9 cases of cardiopulmonary resuscitation in situations where a DNR order had been previously written, and 51 cases were discontinued after the creation of a DNR order at the request of a guardian within 20 minutes during CPR. Further, 68 cases of cardiac arrest occurred within 24 hours. The study was conducted on a total of 605 patients, excluding these 162 cases (Fig. 1).

Fig. 1.Patients’ flow diagram. ROSC*: Return of Spontaneous Circulation.

Of the 605 patients, 69.7% (n = 422) had their spontaneous circulation maintained for 20 minutes or more, and 36 of them maintained circulation with extracorporeal membrane oxygenation (ECMO).

The 24-hour, survival-to-discharge, and 1-month survival rates were 69.7% (422/605), 21.8% (132/605), and 20.3% (120/590), respectively. Good neurological prognoses (CPC 1, 2) were achieved in 14% (85/605) of all patients with ROSC and 64.4% (85/132) of survivors who were eventually discharged.

4. Discussion

Despite active management and advanced resuscitation, IHCA is considered to have a poor survival rate and neurological outcome. The incidence of IHCA varies among countries and regions, ranging from 0-45%; however, it is estimated to be approximately 20%. In the United States, the quality of CPR has improved; indicators are continuously monitored, and CPR training is continuously conducted owing to the revision of CPR guidelines every 5 years based on scientific evidence [15].

According to the Get With The Guidelines-Resuscitation (GWTG-R) registry, the incidence of IHCA in the United States is estimated to be 292,000 cases per year, approximately 9-10 cases per 1000 people. Survival-to-discharge rates vary widely from 0 to 42%; however, it is roughly 20%, and it was estimated as 25% in 2017 [16, 17]. Of these, 85% were discharged with good neurological outcomes. Since the introduction of the GWTG-R system and the In-hospital Utstein style, the United States has been continuously managing the quality of IHCA patients, and the survival rate has improved over the past 20 years [15, 18, 19].

In this study, patient-related factors significantly associated with survival to discharge included cause of cardiac arrest, underlying disease, and medical department at the time of admission. Survival rates were higher when the underlying cause of hospitalization was cardiac, especially for myocardial infarction or heart failure. This association may be due to the rapid action of ECMO insertion and defibrillation in cardiac arrest since patients with heart disease require close observation, including ECG monitoring. In the United States, according to Lars W. et al. and the GWTG-R registry in 2019, about 50-60% of cardiac arrest cases were due to cardiac causes [15, 16]. In Korea, GR Chon et al. reported that non-cardiogenic causes were responsible for 66% of IHCA cases [20]. The 21% of IHCA cases caused by cardiac-related factors reported in this study is lower than that in other studies, which may be due to the exclusion of cardiac arrest cases that occurred during surgery or procedures, such as coronary angiography.

Although many patients suffer from internal diseases, and surgery is often required, cardiac causes are the most important factor causing cardiac arrest; however, other diseases often simultaneously induce cardiac arrest. Supposedly, most patients admitted to the hospital have various complex medical conditions, and IHCA may not be due to one disease. Therefore, in IHCA, classification and application of reversible causes of cardiac arrest [(5H-hypoxia, hypovolemia, hydrogen ion, hypo/hyperkalemia, hypothermia) and (5T-toxins, tension pneumothorax, cardiac tamponade, coronary or pulmonary thrombosis, trauma)], as in OHCA, may help restore spontaneous circulation and improve survival and neurological outcomes [21].

The survival-to-discharge rate was significantly lower in patients with underlying diseases such as sepsis, pneumonia, and malignant tumors and when vasopressors were required. Several studies have reported that malignant tumors, sepsis, brain function status, pneumonia, hypotension, and underlying diseases, such as kidney and liver disease, can predict poor prognosis [6, 22, 23]. Another study reported that factors such as increasing age, male sex, active malignancy, and chronic kidney disease were associated with reduced survival rates [7]. Therefore, patients with these conditions should be identified as having a higher risk of cardiac arrest and closely observed.

In this study, approximately 60% of cardiac arrest cases occurred in the general wards. Although not statistically significant, the survival-to-discharge rate was high when cardiac arrest occurred in the intensive care unit. This finding is different from that observed for the GWTG-R system and others, where large-scale multicenter research was conducted. In these studies, cardiac arrest occurred more frequently in the intensive care unit in about 40% of total cases. In one study, the survival rate was highest in the telemetry unit in the general ward, followed by the intensive care unit and the non-telemetry unit in the general ward [24]. When a patient is hospitalized for severe disease or the disease condition is expected to worsen, patients are closely monitored. About 70% of patients with cardiac arrest in the hospital were monitored for the abovementioned reasons. The survival-to-discharge rate was higher with continuous ECG monitoring. This implies that there is a high survival-to-discharge rate even if cardiac arrest occurs when the patient’s condition and disease progression are predicted in advance, and measures are taken to prevent cardiac arrest.

IHCA is expected in many cases because of existing diseases. Approximately 80% of patients reported that their condition had worsened 6 to 8 hours before cardiac arrest, and 70% of patients experienced worsening of symptoms because of respiratory problems 8 hours before cardiac arrest [25, 26]. Approximately 10% of high-risk patients who could not be detected in the ward were reported to experience an unexpected cardiac arrest, accounting for 4% of all hospitalized patients [27]. It is possible to prevent more than 30% of unpredictable cardiac arrests by identifying high-risk patients in advance and continuously monitoring them [28].

Early recognition of cardiac arrest and activation of the CPR team, high-quality CPR, rapid defibrillation, appropriate advanced resuscitation, and post-cardiac arrest care are linked to the chain of survival and are currently considered essential elements of CPR. Several studies have reported “witnessed cardiac arrest” and cardiac arrest in about 80-85% and 50-70% of patients already on an ECG monitor, respectively [15, 20, 29, 30]. These measures are directly related to the patient’s survival to discharge and prognosis. In this study, no statistical difference in CPR activation time was observed between the surviving and non-surviving groups. As described earlier, to improve the quality of CPR in the hospital where this study was conducted, our policy is to regulate and manage BLS and ACLS activation time within 3 minutes from the initial recognition of cardiac arrest. In addition, we intend to regularly monitor these indicators and conduct continuous quality control through CPR training. Since our code blue activation system is well managed within an average of 1 minute, there seems to be no difference between the surviving and non-surviving groups, and the response time to code blue activation in general wards and intensive care units is considered to be similar.

According to some studies, only about 20% of the initial rhythms in cardiac arrest were shockable rhythm and required defibrillation; rapid defibrillation could lead to a 2-5 times or more increase in the survival rate [6, 31]. In our study, the survival-to-discharge rate was 2.48 times higher (95% CI: 1.36-4.53, P$<$ 0.01) when defibrillation was performed at least once, regardless of the initial rhythm, than when it was not performed. This highlights the importance of quickly identifying ventricular fibrillation or pulseless ventricular tachycardia, which are shockable rhythms, and providing defibrillation as soon as possible. Similar to the findings of an OHCA study [32], the survival rate increased in cases of witnessed arrest and initial shockable rhythm in our IHCA study.

In this study, the 64.4% of patients discharged with good neurological outcomes was significantly lower than the 85% reported in the US GWTG-R in 2017, which implies that the effects of integrated treatment, including targeted temperature management, cannot be overlooked [33, 34]. Younger age, shorter length of stay, witnessed cardiac arrest, and brain performance status with CPC scores of 1 and 2 were associated with good neurological prognosis. In another study, the survival-to-discharge rate increased when cardiac arrest occurred during daytime hours [6]; however, no statistical difference was observed in our study.

This study had several limitations. First, the study focused on a university hospital, which may not represent cardiac arrests in a general hospital and may introduce bias. Second, owing to its retrospective nature, missing and inaccurate records may have prevented the evaluation of other variables of interest. For example, the initial ECG records were often missing. In addition, the presence and monitoring of certain diseases were related to survival and prognosis. However, objective indicators (SOFA score, APACHE2 score) of disease severity at the time of hospitalization and cardiac arrest were not separately mentioned; thus, further investigation is necessary.

Finally, the 2020 AHA guidelines highlight the role of the rapid response team or the medical emergency team in cardiac arrest. In this study, the area of emphasis on survival to discharge and good neurological prognosis in patients with cardiac arrest was “continuous monitoring”. However, in this study, the rapid response team mentioned earlier did not screen patients separately for early prevention and treatment of cardiac arrest; thus, a difference in the survival rate may be witnessed in other hospitals. In the future, multi-center research is required through sharing data and experiences between hospitals.

5. Conclusions

The compulsory availability of the code blue activation system is not sufficient. Patients at high risk of cardiac arrest should be identified, and a monitoring system should be established to identify abnormal signs and symptoms that can predict cardiac arrest.

Author contributions

Ji Ho Ryu established the concept and planned the formal analysis and methodology, and reviewed and edited the writing. Byung Kwan Bae analyzed the data. Mun Ki Min, Dae Sup Lee, and Young Mo Cho validated the data. Soon Chang Park planned the methodology. Min Jee Lee wrote the original draft. Seok Ran Yeom and Mun Ki Min reviewed and edited the writing.

Ethics approval and consent to participate

This study protocol was reviewed and approved by the Institutional Review Board of our hospital (approval no. 05-2020-163).

Acknowledgment

This research was supported by Pusan National University Research Grant, 2020. We would like to thank Editage (www.editage.co.kr) for English language editing.

Conflict of interest

The authors declare that there is no conflict of interest regarding the publication of this article.

Data Availability

The data used to support the findings of this study are available from the corresponding author upon request.