When should on-scene CPR be terminated in OHCA patients?: Prognostic Models for Outcomes Based on Duration of On-Scene CPR in OHCA

doi:10.21203/rs.3.rs-4883915/v1

Download PDF

Research Article

When should on-scene CPR be terminated in OHCA patients?: Prognostic Models for Outcomes Based on Duration of On-Scene CPR in OHCA

https://doi.org/10.21203/rs.3.rs-4883915/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Out-of-Hospital Cardiac Arrest (OHCA) is a leading cause of mortality worldwide, with approximately 30,000 cases managed annually by Emergency Medical Services (EMS) in South Korea. Prolonged on-scene Advanced Life Support (ALS) in OHCA patients has demonstrated potential for prehospital return of spontaneous circulation (ROSC) and neurological improvement. However, the optimal timing for terminating on-scene CPR in patients who do not achieve ROSC remains challenging. This study aims to develop and validate a predictive model for patient outcomes based on the duration of on-scene CPR in OHCA patients using data from the Smart ALS (SALS) protocol in South Korea.

Methods

A multi-regional observational study was conducted from August 2015 to December 2022, involving 19 fire stations and nine academic tertiary hospitals across seven provinces. Data were sourced from the SALS database, including EMS prehospital care reports, SALS intervention logs, and hospital patient records. The study focused on non-traumatic OHCA patients who underwent SALS, excluding those with obvious signs of death, those under 18 years old, those who refused on-scene CPR, or those with a DNR status. Statistical analyses were performed using R software, employing logistic regression models to predict prehospital ROSC, survival to discharge, and favorable neurological outcomes.

Results

Out of 98,569 OHCA patients evaluated, 34,989 were eligible for SALS, and 16,052 received SALS. Significant predictors of prehospital ROSC included younger age, male gender, arrest occurring in public places, witnessed arrest, bystander CPR, and initial shockable rhythm. Logistic regression models for patients who did not achieve prehospital ROSC showed that longer on-scene CPR duration negatively impacted the probability of ROSC, survival to discharge, and neurological outcomes. The predictive model for ROSC had an AUC of 0.730, for survival to discharge AUC of 0.838, and for favorable neurological outcome AUC of 0.917.

Conclusions

This study emphasizes the critical role of prehospital ROSC in improving survival and neurological outcomes in OHCA patients. The predictive models can aid in making informed decisions about the cessation of on-scene CPR. Further research is needed to validate these models and explore their application in different EMS settings.

Trial registration

Retrospectively registered.

Out-of-Hospital Cardiac Arrest (OHCA)

Advanced Life Support (ALS)

Return of Spontaneous Circulation (ROSC)

On-Scene CPR

Cardiac arrest remains a major global public health issue, accounting for approximately 15–20% of total deaths worldwide.[1, 2] Out-of-Hospital Cardiac Arrest (OHCA) continues to be a leading cause of mortality globally, with approximately 30,000 cases annually managed by Emergency Medical Services (EMS) in South Korea.[3–9] Cardiac arrest is a critical medical emergency where timely and effective cardiopulmonary resuscitation (CPR) and advanced life support measures can significantly improve survival rates.[10] Early recognition of cardiac arrest, prompt provision of high-quality CPR, effective defibrillation strategies, and post-resuscitation care are key factors in improving survival rates.[11–13] In this context, early intervention and the provision of high-quality CPR by emergency services are crucial.[14, 15] Most countries implement CPR equivalent to Advanced Cardiac Life Support (ACLS) in the pre-hospital setting. However, Traditionally, the approach in Korea has involved providing two to three rounds of Basic Life Support (BLS) at the scene before transporting patients to hospitals. This practice has been primarily driven by the rapid hospital access within most urban areas, typically within five minutes.[16] However, with the global trend moving towards prolonged prehospital ACLS, regions in Korea with limited hospital access have started adopting a method known as "smart ALS (SALS)."

SALS involves extended CPR duration at the scene, real-time video-assisted CPR quality coaching by physicians, and the use of advanced pharmacological agents such as epinephrine and amiodarone. Preliminary studies have indicated that compared to conventional CPR, SALS significantly enhances prehospital Return of Spontaneous Circulation (ROSC) and improves neurological outcomes.[17] As SALS becomes more widespread, a new debate has emerged regarding how long on-scene CPR should be continued in the absence of ROSC.

Several studies have shown that survival rates decline rapidly when the duration of CPR exceeds 10 minutes and even more rapidly after 30 minutes.[18, 19] Some studies consider refractory cardiac arrest present if ROSC is not achieved after more than 30 minutes of appropriate CPR.[20] However, the 2010 and 2015 AHA Guidelines for CPR and Emergency Cardiovascular Care did not specify the appropriate duration of CPR before terminating out-of-hospital resuscitation efforts.[21, 22] Moreover, initial shockable rhythm has been shown to be a crucial prehospital variable for predicting favorable neurological outcomes after OHCA.[23, 24]

Given this background, this study aims to develop and validate a predictive model for patient outcomes based on the duration of on-scene CPR in OHCA patients. Since 2015, South Korea has been conducting a pilot project in specific regions where EMS personnel receive guidance from emergency medicine specialists via video calls during CPR. This initiative not only aims to enhance the quality of on-scene CPR but also involves the strategic use of ALS medications such as epinephrine and amiodarone. It is crucial to make the optimal decision on when to discontinue CPR at the scene and proceed with hospital transport if ROSC is not achieved in the on-scene.

While some studies have investigated the effectiveness of prolonged on-site CPR, research based on national data is sparse. This study aims to investigate the effectiveness of prolonged on-site CPR using national-level data, focusing on patients who experienced OHCA and participated in the SALS project from 2015 to 2022. We focused on patients who did not achieve ROSC at the scene initially and analyzed the impact of the duration of CPR performed in the on-scene on the likelihood of achieving final ROSC, survival to discharge, and favorable neurological outcomes.

Study design and SALS trial

A retrospective multi-regional observational study was conducted, and data were collected from August 2015 to December 2022. A total of 19 fire stations and nine academic tertiary hospitals from seven prefectures participated in this study. The total study area initially comprised seven regions starting from August 2015 and expanded over time, encompassing nine regions with an approximate population of 16 million by the year 2022. (Fig. 1) This manuscript follows STROBE reporting guidelines for observational studies in epidemiology. The institutional Review Board approved this study and waived the requirement for informed consent. (IRB file no.2024-05-001)

Patients

Non-traumatic OHCA patients who underwent SALS by EMS were collected. Patients were excluded if they had obvious signs of death, were under 18 years old, refused on-scene CPR, or had a do-not-resuscitate (DNR) status. Additionally, patients were excluded if hospital transport was determined by a clinician, if the EMS dispatch area was not a SALS target area, or if records were incomplete. For cardiac arrest patients who were candidates for SALS, SALS was not performed if cardiac arrest occurred during transport, if connection attempts with the supervising physician failed, if the guardians refused SALS, or if the supervising physician made the decision not to perform SALS. Ultimately, only patients who experienced cardiac arrest and underwent SALS were enrolled in the study.

SALS protocol

To perform SALS, EMS instituted procedural modifications. [25] When a cardiac arrest was identified, dispatchers were instructed to send two ambulances to the location. A smartphone app alerted the supervising physician about the OHCA incident through text messages. Upon the emergency medical technicians' (EMTs) arrival, they established a video link with the physician, who then provided guidance on the nuances of performing CPR, such as compression quality, rhythm, depth, allowing for full chest recoil, appropriate ventilation pacing, minimizing hands-off duration, and monitoring of end-tidal CO2 levels.

The overseeing physician was responsible for making critical decisions regarding the administration of epinephrine, the placement of advanced airway management devices like I-gel® and endotracheal tubes, and when to initiate defibrillation, drawing on the ECG readings. EMTs had the authorization to administer intravenous drugs, including epinephrine and amiodarone, and to carry out manual defibrillation following the physician's directives. Additionally, the physician used the app to communicate with the destined hospital, providing detailed patient information to prepare for post-cardiac arrest interventions.

Data sources and variables

Data for the study were sourced from the SALS database, which encompasses EMS prehospital care reports, SALS intervention logs, physician's digital records, and hospital patient files. All information was accessed through an online database. Each regional committee appointed a data manager to ensure data accuracy and to monitor patient outcomes in alignment with the Utstein Style guidelines. Physician entries were recorded in real-time using specialized video call software designed for medical guidance during CPR. The assessment of patient neurological outcomes was based on their cerebral performance category (CPC) scores.

Statistical analysis

All statistical analyses were performed using the R software version 4.0.3 for Windows (R Foundation for Statistical Computing, Vienna, Austria). Categorical variables are denoted by counts and proportions, while continuous ones that don’t follow a normal distribution are indicated by median values and their interquartile ranges (IQR). Kolmogorov-Smirnov test was used to evaluate the distribution of continuous variables. To contrast these variables, the study used the Chi-square test and the Wilcoxon rank-sum test, adopting a significance criterion of p < 0.05. Data division for the study was predetermined by setting a seed to ensure reproducibility of the results. The dataset for SALS performed patients was split into a training set and a test set in an 8:2 ratio to ensure the models’ robustness and prevent overfitting. The primary analysis involved the construction of logistic regression models to predict prehospital ROSC, survival to hospital discharge, and good neurologic outcome. To evaluate the predictive power of each variable, statistically significant predictors were identified, and logistic regression models were built using the training set. The model’s performance was evaluated using confusion matrices to calculate accuracy and Receiver Operating Characteristics (ROC) curves, along with Area Under the Curve (AUC) metrics.

To demonstrate the utility of logistic regression in our dataset, a Generalized Linear Model (GLM) was trained, specifying a binomial family to predict the likelihood of ROSC, survival to discharge, good neurologic outcome. Following model estimation, coefficients were extracted and printed to interpret the contribution of each predictor in the model. To facilitate clinical interpretability, we transformed these coefficients into a scoring system by multiplying each by a factor of ten and then rounding to the nearest integer. The resulting scores were printed to provide a straightforward, quantifiable measure of each variable's impact on the outcome.

From August 2015 to December 2022, we evaluated a cohort of 98,569 patients who experienced OHCA. Initial stratification excluded 22,318 cases due to non-medical causes of arrest. The remaining 76,251 medical cause cardiac arrest patients were further assessed for eligibility for SALS. Patients under the age of 18 years (N = 1,055), those with DNR orders (N = 1,356), refusal of CPR at the scene (N = 12,662), obvious death presentations (N = 18,492), incomplete records (N = 3,616), hospital transport as per physician’s decision (N = 4,133), and cardiac arrests outside the SALS region (N = 128) were subsequently excluded. This process resulted in a group of 34,989 SALS-eligible patients. Of these, 18,937 did not receive SALS, leaving a final cohort of 16,052 patients who received SALS. (Fig. 2)

Table 1 shows a demographic comparison between the groups that achieved prehospital ROSC and those that did not among patients who received SALS. Significant differences were observed between the group that did not achieve prehospital ROSC (N = 12,516) and the group that achieved prehospital ROSC (N = 3,536). The median age was 74 (60–82) in the group that did not achieve ROSC, compared to 63 (52–76) in the ROSC group, indicating a lower median age in the ROSC group (P < 0.01). The proportion of males was also higher in the ROSC group at 70.8%, compared to 63.1% in the group that did not achieve ROSC (P < 0.01).

Table 1

Outcomes and characteristics of patients with and without prehospital ROSC
	No prehospital ROSC (N = 12,516)	Prehospital ROSC (N = 3,536)	P
Age (median, IQR)	74 (60–82)	63 (52–76)	< 0.001
Sex, male (%)	7893 (63.1)	2505 (70.8)	< 0.001
Arrest place, home	9087 (73.8)	2246 (64.1)	< 0.001
Witnessed arrest	5417 (43.3)	2309 (65.3)	< 0.001
Bystander CPR	7775 (62.1)	2355 (66.6)	< 0.001
Shockable rhythm	1462 (11.7)	1453 (40.6)	< 0.001
Response time interval	8 (6–10)	7 (6–10)	< 0.001
Scene time interval	23 (18–28)	21 (16–26)	< 0.001
Total transport time	6 (4–9)	7 (4–11)	< 0.001
Hospital procedure
PCI (%)	148 (1.2)	669 (18.9)	< 0.001
ECMO (%)	149 (1.2)	111 (3.1)	< 0.001
TTM (%)	193 (1.5)	675 (19.1)	< 0.001
Comorbidities
Hypertension	4,381 (35.0)	1,016 (28.7)	< 0.001
Diabetes	3,203 (25.6)	683 (19.3)	< 0.001
Heart disease	2,238 (17.9)	674 (19.1)	0.113
Stroke	1,153 (9.2)	226 (6.4)	< 0.001
Cancer	1,296 (10.4)	233 (6.6)	< 0.001
Survival to admission	870 (7.0)	2,315 (65.5)	< 0.001
Survival to discharge	231 (1.8)	1,386 (39.2)	< 0.001
Good neurologic outcome	42 (0.3)	904 (25.6)	< 0.001
ROSC, return of spontaneous circulation; CPR, cardiopulmonary resuscitation; PCI, percutaneous coronary intervention; ECMO, extracorporeal membrane oxygenation; TTM, target temperature management.

When the location of cardiac arrest was at home, it was 73.8% in the group that did not achieve ROSC, compared to 64.1% in the ROSC group, indicating that prehospital ROSC occurred more frequently in public places (P < 0.01). The rate of witnessed cardiac arrests was also higher in the ROSC group at 65.3% compared to 43.3% in the group that did not achieve ROSC (P < 0.01). The rate of bystander CPR was 66.6% in the ROSC group, compared to 62.1% in the group that did not achieve ROSC, showing that bystander CPR was more frequently performed in the ROSC group (P < 0.01). The proportion of patients with an initial shockable rhythm was significantly higher in the ROSC group at 40.6%, compared to 11.7% in the group that did not achieve ROSC (P < 0.01).

Response time interval (RTI), scene time interval (STI), and transport time interval (TTI) were all shorter in the ROSC group, and these differences were statistically significant (P < 0.01). Among the treatment procedures performed in the hospital, percutaneous coronary intervention (PCI), extracorporeal membrane oxygenation (ECMO), and target temperature management (TTM) were all conducted at higher rates in the ROSC group (P < 0.01).

Survival to admission, survival to discharge, and good neurologic outcomes were all higher in the prehospital ROSC group, emphasizing the importance of prehospital ROSC. In the group that achieved prehospital ROSC, 65.5% of the patients admitted to the hospital survived, 39.2% of those discharged survived, and 25.6% had a good neurologic outcome (P < 0.01).

Based on the differences in demographics, a multivariate logistic model was created to predict prehospital ROSC, and the results are shown in Fig. 3. The AUC value was 0.730, and based on this, the probability of prehospital ROSC could be predicted according to STI. According to the prediction model, the probability of prehospital ROSC decreased as the on-scene CPR time increased. Notably, when the on-scene CPR time exceeded 20 minutes, the probability dropped to less than 10%.

Table 2 uses multivariate logistic regression to predict final ROSC, survival to discharge, and good neurologic outcomes among patients who did not achieve prehospital ROSC. The multivariate logistic regression split the total group into an 80:20 ratio, using 80% to create the scoring system and applying it to the remaining 20% to evaluate the scoring system's validity. In the multivariate logistic regression for predicting ROSC, age, witnessed arrest, shockable rhythm, RTI, STI, and TTI were prognostic indicators. In the multivariate logistic model for survival to discharge, age, witnessed arrest, shockable rhythm, RTI, STI, and TTI were independent prognostic indicators. In the multivariate logistic regression for a good CPC, age, gender, witnessed arrest, shockable rhythm, RTI, STI, TTI, and heart disease were independent prognostic indicators.

Table 2

Multivariate logistic analysis of factors associated with ROSC, survival to discharge, and good neurologic outcome in OHCA patients
All ROSC	Odd ratio (95% Confidence interval)	P
Age	0.981 (0.978–0.984)	< 0.001
Sex	0.932 (0.853–1.018)	0.116
Witnessed	2.092 (1.926–2.272)	< 0.001
Bystander CPR	0.927 (0.852–1.009)	0.078
Shockable rhythm	2.881 (2.596–3.196)	< 0.001
RTI	0.938 (0.927–0.950)	< 0.001
STI	0.963 (0.958–0.968)	< 0.001
TTI	1.041 (1.033–1.048)	< 0.001
Diabetes	0.982 (0.891–1.082)	0.716
Heart disease	1.086 (0.977–1.207)	0.125
Survival to discharge
Age	0.970 (0.966–0.974)	< 0.001
Sex	1.132 (0.968–1.326)	0.121
Witnessed	2.131 (1.847–2.462)	< 0.001
Bystander CPR	1.092 (0.945–1.263)	0.236
Shockable rhythm	6.816 (5.909–7.868)	< 0.001
RTI	0.914 (0.894–0.933)	< 0.001
STI	0.906 (0.897–0.915)	< 0.001
TTI	1.033 (1.023–1.043)	< 0.001
Diabetes	0.842 (0.706–1.002)	0.056
Heart disease	1.235 (1.085–1.518)	0.004
Good neurologic outcome
Age	0.956 (0.950–0.962)	< 0.001
Sex	1.591 (1.234–2.058)	< 0.001
Witnessed	2.192 (1.772–2.722)	< 0.001
Bystander CPR	1.180 (0.957–1.459)	0.123
Shockable rhythm	17.516 (13.996–22.093)	< 0.001
RTI	0.900 (0.873–0.927)	< 0.001
STI	0.871 (0.859–0.884)	< 0.001
TTI	1.035 (1.021–1.048)	< 0.001
Diabetes	0.705 (0.535–0.922)	0.012
Heart disease	1.559 (1.237–1.960)	< 0.001
ROSC, return of spontaneous circulation; CPR, cardiopulmonary resuscitation; RTI, response time interval; STI, scene time interval; TTI, transport time interval.

The results of the multivariate logistic model for predicting final ROSC are shown in Fig. 4. The AUC value is 0.705. When creating the ROSC scoring system based on each variable, it was found that a shockable rhythm had a score of 11 and witnessed arrest had a score of 7, positively impacting ROSC. However, it was also found that a longer RTI, being female, and bystander CPR were associated with lower ROSC rates.

The results of the multivariate logistic model for predicting survival to discharge are shown in Fig. 4. The AUC value is 0.838. When creating the scoring system based on each variable, it was found that shockable rhythm had a contribution score of 19, witnessed arrest had a score of 8, heart disease had a score of 3, being male had a score of 1, and bystander CPR had a score of 1, all positively impacting survival to discharge. This suggests that patients with heart disease may respond better to CPR, thus having a positive impact. Conversely, longer STI and RTI negatively impacted survival, decreasing by 1 point per minute. Additionally, having diabetes also negatively impacted survival, decreasing the score by 1 point.

The results of the multivariate logistic model for predicting a good CPC are shown in Fig. 4. The AUC value is high at 0.917. When creating the scoring system based on each variable, the contribution of a shockable rhythm was very high at 31, followed by 7 for witnessed arrest, 4 for heart disease, and 4 for being male, all positively impacting the outcome. Conversely, longer STI and RTI negatively impacted the outcome, decreasing by 1 point per minute, and having diabetes also negatively impacted the outcome, decreasing the score by 3 points.

The results of this study emphasize that prehospital ROSC has a positive impact on the prognosis of cardiac arrest patients. However, the probability of achieving prehospital ROSC decreased as the duration of on-scene CPR increased. In patients who did not achieve prehospital ROSC, when predicting survival to discharge and good neurologic outcomes, it was found that the probability of both outcomes decreased as the duration of on-scene CPR increased.

A similar study conducted in Beijing on OHCA patients who underwent resuscitation by EMS has been reported. According to Chen et al., when EMS attempted resuscitation, patients who achieved ROSC at any time had approximately six times higher discharge survival rate compared to those who did not achieve ROSC. [26] Additionally, with every 1% increase in the survived event, the survival rate increased directly up to 53.5%. Although this range is broader than the increase in survival rates observed in our study for prehospital ROSC, it shares the same implication that ROSC events positively impact prognosis.

Several previous studies have found that the probability of achieving prehospital ROSC decreases as the duration of on-scene CPR increases.[27–29] Similar to our study, the research by de Graaf et al. found that OHCA patients with shorter on-scene CPR durations were more likely to survive after 30 days (average 20 minutes vs. 26 minutes). Additionally, their study showed that patients with an initial shockable rhythm had higher survival rates, which is consistent with our findings.[29] Similarly, Matsuyama et al. also indicated that the probability of good neurological outcomes decreased as the on-scene CPR duration increased in OHCA patients.[30] The probability was highest within the first 20 minutes and then declined sharply over time. Furthermore, they suggested that in cases with a witnessed arrest or an initially recorded shockable rhythm, a longer duration of CPR could potentially lead to favorable outcomes, which aligns with our study's results. Therefore, these studies suggest that the duration of on-scene CPR can significantly impact patient survival rates. Moreover, in cardiac arrest situations, rapid decision-making and efficient CPR performance are crucial, especially for patients with witnessed cardiac arrest or shockable rhythms. Based on these results, it is expected that the appropriate duration of on-scene CPR can be determined, but additional research will be necessary.

In this study, the group of patients who achieved prehospital ROSC was younger, had a higher proportion of men, and experienced more cardiac arrests occurring in public places. This suggests that cardiac arrest is more likely to be witnessed at a young age and in public locations, and prompt emergency response can increase the survival rate. Additionally, the rate of witnessed cardiac arrest and the implementation of bystander CPR increase the probability of ROSC occurring, emphasizing the importance of public CPR education.[31] It can be observed that patients with ROSC have a higher rate of shockable rhythms, which provides an opportunity to normalize the heart rhythm through early intervention during cardiac arrest.[29] In the group of prehospital ROSC patients, treatment procedures such as PCI, ECMO, and TTM were performed at a higher rate. This can be interpreted as prehospital ROSC increasing the possibility of performing advanced treatment procedures in the hospital, ultimately having a positive effect on patient survival rate and neurological outcomes. Hypertension and diabetes appeared at a higher rate in the group that did not develop ROSC. This indicates that certain underlying diseases can affect the survival rate after cardiac arrest and suggests that management of underlying diseases is important in the prognosis of cardiac arrest patients.

Our study provides a differentiated approach from the existing literature. What distinguishes the SALS protocol from existing CPR methodologies is that it focuses on strengthening CPR continuity in the on-scene through real-time video support and the use of advanced medications. This approach presents a new methodology for managing cardiac arrest in the field and opens up the possibility of a wide range of practical applications.

The prognosis prediction model developed through this study confirmed that shockable rhythm and witnessed arrest are important prognostic indicators for ROSC, good neurologic outcome, and survival to discharge. This model can be an important tool in supporting decision-making about the management of OHCA patients in actual medical settings, and in the future, a personalized treatment approach based on this model can be developed.

Similarly to our study, Ji et al. developed and verified a prediction model for predicting the survival outcomes of permanent OHCA patients and a model for predicting ROSC. Important predictive factors included the patient's age, gender, etiology, whether bystander CPR was performed, causative disease, and initial rhythm. The survival model showed better classification ability and accuracy than the ROSC model, with AUC values of 0.86 and 0.67, respectively. [32] Furthermore, our study added STI, RTI, and TTI to further specify the field situation at the time, concluding that, in fact, the longer the CPR time in the on-scene, the lower the probability of ROSC.

Another study by Martinell et al. examined early indicators to predict long-term outcomes after survival. This was a retrospective analysis derived from patients' medical history and variables available upon admission to the intensive care unit. Predictors of poor outcome included older age, cardiac arrest at home, initial rhythm other than ventricular fibrillation/tachycardia, length of no-flow period, length of low-flow period, adrenaline administration, absence of bilateral corneal and pupillary reflexes, Glasgow Coma Scale score of 1, low pH upon admission, and arterial blood carbon dioxide partial pressure of less than 4.5 kPa upon admission. Information obtained from the patient's medical history was not included in our study, but the same factor was used.[33]

The strength of our study is that it is based on large-scale national data on OHCA patients, providing a broad analysis of various factors that affect the effectiveness and prognosis of CPR. One of the key strengths of this study is that it quantitatively evaluated the importance of each factor through complex multivariate logistic regression analysis and developed a practical model that can predict the patient's prognosis based on these factors. This approach allows medical professionals to easily obtain information about a patient's prognosis and make more efficient treatment decisions. Particularly, the model predicting good neurologic outcome showed high performance (0.917 of AUC).

This study also found that the longer the CPR time, the worse the patient's prognosis, emphasizing the importance of rapid decision-making and efficient staffing at the scene of cardiac arrest. This insight will contribute to improving the quality of emergency medical services by reducing unnecessary waste of manpower in emergency situations and suggesting ways to optimize patient survival rates and neurological outcomes.

Nevertheless, our study has several limitations. First, the use of multivariable logistic regression to analyze the factors influencing the prognosis of OHCA patients may have resulted in an overemphasis of certain variables over others. Specifically, variables such as shockable rhythm and bystander CPR demonstrated a significant impact on the prognosis post-ROSC and CPR. However, the strong influence of these variables may have overshadowed the relative contribution of other important factors, potentially leading to an overestimation of their influence and inadequate reflection of other significant variables.

Second, the absence of precise classification criteria for the application of the SALS protocol may have introduced selection bias in the inclusion of patients. Such biases could limit the generalizability of our findings and lead to inaccuracies in evaluating the effectiveness of the SALS protocol. Despite efforts to minimize potential bias through various statistical methods and a thorough data cleaning process to remove incomplete or inappropriate data, this remains a concern.

Third, the simultaneous implementation of several improved interventions during the study complicates the assessment of the independent impact of each intervention on patient outcomes. These interventions included CPR instruction via video call, modifications to ACLS, increased on-scene delay time, and the administration of epinephrine. Future research should employ study designs that can independently evaluate the effects of these individual interventions.

Fourth, the study was conducted across various regions, and regional differences, along with the Hawthorne effect (the phenomenon where behavior improves merely due to awareness of being observed), may have influenced the results. More detailed regional analyses are necessary to control for these factors.

Despite these limitations, our study is significant in identifying crucial factors for predicting the prognosis of OHCA patients and developing a prognostic prediction model based on these factors. Our findings offer valuable insights for making decisions regarding on-scene CPR termination. Future studies should aim to validate these results, ideally through prospective research designs.

Out-of-Hospital Cardiac Arrest (OHCA)

Emergency Medical Services (EMS)

Advanced Life Support (ALS)

return of spontaneous circulation (ROSC)

cardiopulmonary resuscitation(CPR)

Smart ALS (SALS)

Advanced Cardiac Life Support (ACLS)

do-not-resuscitate (DNR)

emergency medical technicians' (EMTs)

cerebral performance category (CPC)

interquartile ranges (IQR)

Receiver Operating Characteristics (ROC)

Area Under the Curve (AUC)

Generalized Linear Model (GLM)

Response time interval (RTI)

scene time interval (STI)

transport time interval (TTI)

percutaneous coronary intervention (PCI)

extracorporeal membrane oxygenation (ECMO)

target temperature management (TTM)

Ethics approval and consent to participate

This study received approval from The institutional Review Board and waived the requirement for informed consent. (IRB file no.2024-05-001)

Consent for publication

All authors have given their consent for the publication of this manuscript.

Availability of data and materials

The data and materials used in this study are available from the corresponding author upon reasonable request.

Competing interests

The authors declare that they have no competing interests.

Funding

This work was supported by Soonchunhyang University resarch fund.

Authors' contributions

Soh Yeon Chun, M.D.

Contribution: Data collection. Analysis and interpretation of data and and drafting the article and revising it critically.

Han Bit Kim, M.D.

Contribution: The conception and design of the study. Analysis and interpretation of data and drafting the article and revising it critically for important intellectual content.

Gi Woon Kim, Ph.D., M.D.

Contribution: Evaluation and modification of the manuscript. Addition of intellectual content

Acknowledgement

This work was supported by the Soonchunhyang University research fund. We thank for thier contributions and support to this resarch.

Hayashi M, Shimizu W, Albert CM. The spectrum of epidemiology underlying sudden cardiac death. Circul Res. 2015;116:1887–906.
Berdowski J, Berg RA, Tijssen JGP, Koster RW. Global incidences of out-of-hospital cardiac arrest and survival rates: Systematic review of 67 prospective studies. Resuscitation 2010;81:1479-87.
Fernando SM, Vaillancourt C, Morrow S, Stiell IG. Analysis of bystander CPR quality during out-of-hospital cardiac arrest using data derived from automated external defibrillators.Resuscitation 2018;128:138 – 43.
Gavin DP, Samantha JB, Mike S, Giok O, Simon G. Out-of-hospital cardiac arrest:recent advances in resuscitation and effects on outcome. Heart 2012;98:529.
Roh S-Y, Choi J-I, Park SH, et al. The 10-year trend of out-of-hospital cardiac arrests: a Korean nationwide population-based study. Korean Circulation Journal 2021;51:866.
Eisenberg M. The quest to reverse sudden death. A history of cardiopulmonary resuscitation.The cardiac arrest. The science and practice of resuscitation medicine by Paradis,Halperin, Nowak 1996.
Pantridge JF, Geddes JS. A mobile intensive-care unit in the management of myocardial infarction. Lancet 1967;2:271-3.
Stiell IG, Wells GA, Field B, et al. Advanced cardiac life support in out-of-hospital cardiac arrest. N Engl J Med 2004;351:647 – 56.
Zoch TW, Desbiens NA, DeStefano F, Stueland DT, Layde PM. Short-and long-term survival after cardiopulmonary resuscitation. Archives of internal medicine 2000;160:1969-73.
Lund-Kordahl I, Olasveengen TM, Lorem T, Samdal M, Wik L, Sunde K. Improving outcome after out-of-hospital cardiac arrest by strengthening weak links of the local Chain of Survival; quality of advanced life support and post-resuscitation care. Resuscitation 2010;81:422-6.
Moretti MA, Cesar LAM, Nusbacher A, Kern KB, Timerman S, Ramires JAF. Advanced cardiac life support training improves long-term survival from in-hospital cardiac arrest. Resuscitation 2007;72:458 – 65.
Cummins RO, Ornato JP, Thies WH, Pepe PE. Improving survival from sudden cardiac arrest: the" chain of survival" concept. A statement for health professionals from the Advanced Cardiac Life Support Subcommittee and the Emergency Cardiac Care Committee,American Heart Association. Circulation 1991;83:1832-47.
Noel L, Jaeger D, Baert V, et al. Effect of bystander CPR initiated by a dispatch centre following out-of-hospital cardiac arrest on 30-day survival: Adjusted results from the French National Cardiac Arrest Registry. Resuscitation 2019;144:91 – 8.
Smith GB, Prytherch DR, Meredith P, Schmidt PE, Featherstone PI. The ability of the National Early Warning Score (NEWS) to discriminate patients at risk of early cardiac arrest, unanticipated intensive care unit admission, and death. Resuscitation 2013;84:465 – 70.
Hasselqvist-Ax I, Riva G, Herlitz J, et al. Early cardiopulmonary resuscitation in out-of-hospital cardiac arrest. New England Journal of Medicine 2015;372:2307-15.
Kim GW, Moon HJ, Lim H, et al. Effects of Smart Advanced Life Support protocol implementation including CPR coaching during out-of-hospital cardiac arrest. The American Journal of Emergency Medicine 2022;56:211-7.
Chen Y-S, Lin J-W, Yu H-Y, et al. Cardiopulmonary resuscitation with assisted extracorporeal life-support versus conventional cardiopulmonary resuscitation in adults with in-hospital cardiac arrest: an observational study and propensity analysis. The Lancet 2008;372:554 – 61.
Shih C-L, Lu T-C, Jerng J-S, et al. A web-based Utstein style registry system of in-hospital cardiopulmonary resuscitation in Taiwan. Resuscitation 2007;72:394–403.
Le Guen M, Nicolas-Robin A, Carreira S, et al. Extracorporeal life support following out-of-hospital refractory cardiac arrest. Critical care 2011;15:1–9.
Morrison LJ, Kierzek G, Diekema DS, et al. Part 3: ethics: 2010 American Heart Association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 2010;122:S665-S75.
Mancini ME, Diekema DS, Hoadley TA, et al. Part 3: ethical issues: 2015 American Heart Association guidelines update for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 2015;132:S383-S96.
Goto Y, Maeda T, Goto YN. Termination-of-resuscitation rule for emergency department physicians treating out-of-hospital cardiac arrest patients: an observational cohort study. Critical Care 2013;17:1–9.
Goto Y, Maeda T, Nakatsu-Goto Y. Neurological outcomes in patients transported to hospital without a prehospital return of spontaneous circulation after cardiac arrest. Critical Care 2013;17:1–10.
Kim GW, Moon HJ, Lim H, et al. Effects of Smart Advanced Life Support protocol implementation including CPR coaching during out-of-hospital cardiac arrest. Am J Emerg Med 2022;56:211-7.
Chen Y, Yue P, Wu Y, et al. Trend in survival after out-of-hospital cardiac arrest and its relationship with bystander cardiopulmonary resuscitation: a six-year prospective observational study in Beijing. BMC Cardiovascular Disorders 2021;21:625.
Reynolds JC, Frisch A, Rittenberger JC, Callaway CW. Duration of resuscitation efforts and functional outcome after out-of-hospital cardiac arrest: when should we change to novel therapies? Circulation 2013;128:2488-94.
Adnet F, Triba MN, Borron SW, et al. Cardiopulmonary resuscitation duration and survival in out-of-hospital cardiac arrest patients. Resuscitation 2017;111:74–81.
de Graaf C, Beesems SG, Koster RW. Time of on-scene resuscitation in out of-hospital cardiac arrest patients transported without return of spontaneous circulation. Resuscitation 2019;138:235 – 42.
Matsuyama T, Ohta B, Kiyohara K, Kitamura T. Cardiopulmonary resuscitation duration and favorable neurological outcome after out-of-hospital cardiac arrest: a nationwide multicenter observational study in Japan (the JAAM-OHCA registry). Critical Care 2022;26:120.
Girotra S, van Diepen S, Nallamothu BK, et al. Regional Variation in Out-of-Hospital Cardiac Arrest Survival in the United States. Circulation 2016;133:2159-68.
Ji C, Brown TP, Booth SJ, et al. Risk prediction models for out-of-hospital cardiac arrest outcomes in England. Eur Heart J Qual Care Clin Outcomes 2021;7:198–207.
Martinell L, Nielsen N, Herlitz J, et al. Early predictors of poor outcome after out-of-hospital cardiac arrest. Critical Care 2017;21:96.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

When should on-scene CPR be terminated in OHCA patients?: Prognostic Models for Outcomes Based on Duration of On-Scene CPR in OHCA

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Trial registration

Figures

Background

Methods

Study design and SALS trial

Patients

SALS protocol

Data sources and variables

Statistical analysis

Results

Discussion

Conclusions

Abbreviations

Declarations

References

Additional Declarations

Status:

Version 1