Compliance with cardiopulmonary resuscitation guidelines in witnessed in-hospital cardiac arrest events and patient outcome on monitored versus non-monitored wards

Background : Adherence to cardiopulmonary resuscitation (CPR) guidelines in treatment of in-hospital cardiac arrest (IHCA) have been associated with favourable patient outcome. The aim of this study was to evaluate if compliance with initial CPR guidelines and patient outcome of witnessed IHCA events were associated with the place of arrest defined as monitored versus non-monitored ward. Methods : A total of 956 witnessed IHCA events in adult patients at six hospitals during 2018 to 2019, were extracted from the Swedish Registry of Cardiopulmonary Resuscitation. Initial CPR guidelines were: (cid:1) 1 min from collapse to alert of the rapid response team, (cid:1) 1 min from collapse to start of CPR, (cid:1) 3 min from collapse to defibrillation of shockable rhythm. Results : The odds of compliance with guidelines was higher on monitored wards vs non-monitored wards, even after adjustment for factors that could affect staffing and resources. The place of arrest was not a significant factor for sustained return of spontaneous circulation, survival at 30 days, or neurological status at discharge, when adjusting for clinically relevant confounders. Compliance with initial CPR guidelines remained a significant factor for survival to 30 days and favourable neurological outcome at discharge regardless of other confounders. Conclusion : Compliance with initial CPR guidelines was higher in witnessed IHCA events on monitored wards than on non-monitored wards, which indicates that healthcare professionals in monitored wards are quicker to recognize a cardiac arrest and initiate treatment. When initial CPR guidelines are followed, the place of arrest does not influence patient outcome.


Background
The Swedish guidelines for cardiopulmonary resuscitation (CPR) of in-hospital cardiac arrest (IHCA) follow the European Resuscitation Council guidelines and the chain of survival concept. 1,2Time to call a rapid response team (RRT), 3 time to initiation of CPR, and time to defibrillation are highly associated with IHCA survival. 4,5Since 1989, advanced CPR training is provided to doctors and nurses on monitored wards in Swedish hospitals.In 2006, CPR with an automated external defibrillator (CPR AED) was introduced on non-monitored wards and other non-monitored areas to shorten time to defibrillation. 6Over time, an overall increase in the proportion of patients defibrillated before the arrival of the RRT has been shown. 7However, the time from cardiac arrest to calling the RRT, time to initiation of CPR, and time to defibrillation are relatively unchanged. 8A slight overall increase in adherence to initial CPR guidelines has been seen over time.Adherence to initial CPR guidelines has been shown to be higher in patients with non-shockable rhythms and in monitored areas of hospitals and is associated with increased 30-day survival https://doi.org/10.1016/j.resuscitation.2024.1101250300-9572/Ó 2024 The Author(s).Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/ licenses/by/4.0/).regardless of cardiac rhythm. 9Differences regarding patient-to-nurse ratio 10 and differences regarding the availability of hospital resources and competence during on-call hours and holiday periods could explain lower adherence in non-monitored areas. 11,12atients treated for IHCA events in non-monitored hospital areas have an overall lower rate of survival to 30 days than patients treated in monitored areas: 23% versus 43%. 7These patients are often older, have a non-cardiac aetiology with a non-shockable rhythm, have comorbidity associated with decreased survival, 7 and more often have a non-witnessed IHCA event compared with patients in monitored areas. 13revious studies have found associations between advanced CPR and increased rate of return of spontaneous circulation (ROSC) 14,15 and survival to discharge. 16Though the evidence of the effect of advanced CPR treatment is of low certainty. 3,17This suggests that other factors attained through advanced CPR training, such as leadership and team-building skills, which improve team performance, 18 may affect patient outcome.
The aim of this study was to evaluate if compliance with initial CPR guidelines and patient outcome of witnessed IHCA events were associated with the place of arrest defined as monitored versus nonmonitored ward.

Design and setting
This retrospective observation study was performed at six hospitals through purposeful sampling from January 1, 2018, to December 31, 2019.The hospitals had a variation regarding geographic location, hospital size, hospital resources, and provided levels of care.The hospitals had a total of 2,403 in-patient beds (min 111 to max 783 beds) with 341,584 hospital admissions (20,667-84,803 admissions) during the study period.All hospitals provided 24/7 care, had RRTs that responded to cardiac arrests calls, and had intensive care units (ICUs).All hospitals but one had a percutaneous coronary intervention unit.Three hospitals had a medical emergency team (MET) that responded to patients at risk of deterioration on general wards to prevent further deterioration.

Data collection
Data were extracted from the Swedish Registry of Cardiopulmonary Resuscitation (SRCR).The extracted data were validated against inhospital medical records in a previous study 35 (registered at Clini-calTrials, ID NCT05184972) to attain complete case ascertainment.All adult patients (18 years and older) who were treated for an IHCA were included.Children and adult patients with an IHCA that was not treated were excluded.

Definition of initial CPR guidelines
The Swedish resuscitation guidelines refer to the European resuscitation guidelines.The initial guidelines of CPR AED and advanced CPR are 1 minute from collapse to call to RRT, 1 minute from collapse to initiation of CPR, and 3 minutes from collapse to defibrillation of a shockable rhythm.Time from collapse to defibrillation is evaluated among patients found with ventricular fibrillation (VF) or pulseless ventricular tachycardia (VT) as initial rhythm.The guidelines for reporting IHCAs to the SRCR state that if members of the RRT are present at the location of the arrest (as in ICUs and operating theatres), the event should be reported as 'No call to the RRT', which means that no time of call is reported.This falsely lowers the proportion of IHCA events treated in accordance with initial guidelines.Therefore, in this study, we did not include the 'alert to the RRT' variable in ICUs and operating theatres; only time to CPR and time to defibrillation of a shockable rhythm were evaluated at these locations.

Definition of variables
An IHCA was defined as a cardiac arrest that occurred inside the hospital perimeter in a patient who entered the hospital alive, and who received chest compressions or defibrillation due to absent or insufficient circulation.Monitored wards included all hospital wards or units with continuous monitoring equipment (ICUs, intermediate wards, operating theatres, coronary care units, coronary catheterization laboratories, and emergency departments).A witnessed IHCA event meant recognition by hospital healthcare professionals or surveillance equipment at the time of collapse.The aetiology was divided into cardiac or non-cardiac causes, where cardiac causes were ischemia/infarction, primary arrythmia, cardiomyopathy, postcardiac surgery arrythmia, and cardiac tamponade.All other causes, including unknown, were considered as non-cardiac.The initial rhythm was defined as shockable (VT/VF) or non-shockable (asystole or pulseless electrical activity).Comorbidity was previous history of diabetes, heart failure, stroke, cancer, chronic kidney disease, or respiratory insufficiency.Chronic kidney disease was considered as an estimated glomerular filtration rate <60 ml/min/1.73m 2 .This was estimated based on sex, age, and serum creatinine using the Lund Malmo formula LMR18. 19Sustained ROSC meant spontaneous circulation for 20 consecutive minutes.Favourable neurological outcome at discharge was considered as a Cerebral Performance Classification (CPC) score of 1-2, which means normal neurological status to moderate disability.

Statistical analysis
All variables were presented with appropriate measures for type and distribution.Potential differences in categorical variables were analysed with chi-squared tests and in continuous variables with Mann Whitney's U tests.Binary logistic regression was used in the analysis regarding possible association of the dependent variable, compliance with initial CPR guidelines, and the place of arrest (monitored vs non-monitored ward).The model was adjusted for clinically relevant factors such as initial rhythm, on-call hours, holiday period, and day of the week.Binary logistic regression was also used in the analyses of the dependent variables (sustained ROSC, survival to 30 days, and CPC score 1-2 at discharge) and possible associations with the place of arrest (monitored vs non-monitored ward).These regression models were adjusted for clinically relevant factors such as compliance with initial CPR guidelines, initial rhythm, age, sex, aetiology, and comorbidity.The total proportion of missing items of the variables included in the regression analyses were 2.7%, but 35% for compliance with initial CPR guidelines.Missingness was considered as missing at random.Complete case analysis yielded the same standard errors as with an imputed dataset, and the complete case analysis was therefore considered to give unbiased results.Some patients had more than one IHCA event, therefore, all regression models were analysed with robust standard errors.The goodness of fit was evaluated with the Wald test and Hosmer-Lemeshow's test.All analyses were performed with IBM SPSS Statistics, version 28, and R: A language and environment for statistical computing, using RStudio: Integrated Development for R.
Methods and results are reported using the Utstein template 20 and the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist for cross-sectional studies. 21

Ethical considerations
The study obtained ethical approval from the Swedish Ethics Review Authority in Stockholm, dnr 2021-04102.Informed consent from living study participants is retrieved by the SRCR.The study was performed in accordance with the Declaration of Helsinki. 22

Results
A total of 956 witnessed IHCA events in patients aged 18 years were extracted from the SRCR.The proportion of IHCA events was 69% (n = 666) on monitored wards versus 31% (n = 290) on non-monitored wards.Of all extracted events, 619 had complete information of initial CPR guidelines, time from collapse to alert of RRT, time from collapse to initiation of CPR, and time from collapse to defibrillation of a shockable rhythm.The proportion of IHCA events with compliance with initial CPR guidelines was higher on monitored wards than on non-monitored wards (86% versus 67%).Also, the proportions of sustained ROSC, survival to 30 days, and CPC score 1-2 at discharge were higher on monitored wards than on nonmonitored wards (Table 1).
The odds of compliance with initial CPR guidelines among witnessed IHCA events was higher on monitored wards vs nonmonitored wards (OR 3.07; 95% CI 2.01-4.67),even after adjustment for factors that could affect staffing and resources.The place of arrest (monitored vs non-monitored ward) was not a significant factor regarding sustained ROSC, survival at 30 days, and CPC score at discharge, when clinically relevant confounders were adjusted for.Compliance with initial CPR guidelines remained a significant confounder of survival to 30 days and CPC score at discharge regardless of other confounders (Table 2).

Discussion
This study showed that the odds of compliance with initial CPR guidelines in witnessed IHCA events were higher on monitored wards than on non-monitored wards, even after adjustment for factors that could affect staffing and available resources.However, the place of arrest was not a significant factor for patient outcome when adjusting for clinically relevant confounders, including compliance with CPR guidelines, which remained a significant confounder of survival to 30 days and CPC score at discharge.
relations have been shown to independently be associated with patient survival. 10Also, monitoring of patients with a high risk of a cardiac arrest, 23 especially in non-monitored areas, could increase the chance of a witnessed IHCA event.In this study, the confounders that could potentially affect staffing and available resources (on-call hours, holiday period, and day of the week), which in turn could affect compliance with initial CPR guidelines of witnessed IHCA events, did not seem to be important predictors in the regression model.This suggests that other factors affected compliance with CPR guidelines in witnessed IHCA events.Unnecessary actions, lack of knowledge, lack of skills regarding treatment algorithm, communication, role allocation, leadership, and shared knowledge have been shown to be associated with delays of the first links in the chain of survival concept 24 and with errors during performance of resuscitation. 25Delays of resuscitation have been associated with an increased risk of death prior to hospital discharge. 26A delay in initiating CPR has been found to be associated with decreased survival regardless of time to defibrillation or administration of epinephrine. 27Stress is common in cardiac arrest situations 28 and higher odds of perceived stress in CPR situations have been associated with greater likelihood of working on a non-monitored ward as compared with a monitored ward. 29High stress can delay initiation of CPR. 30 On monitored wards, most physicians and nurses attend advanced CPR training that emphasizes leadership skills and explicit task distribution, which have been found to be important factors for compliance with initial CPR guidelines in simulated scenarios. 31Adherence to advanced life support guidelines 14,15,32 and knowledge of advanced life support 33 have previously been associated with sustained ROSC.Also, use of continuous monitoring equipment on monitored wards could contribute to faster recognition of a cardiac arrest and faster initiation of treatment of witnessed arrests than on non-monitored wards.Furthermore, it is reasonable to assume that the incidence of IHCA is higher in monitored areas making the healthcare professionals in these areas more experienced and prepared for the event of an IHCA.Repeated in-situ CPR training, especially in non-monitored areas, is therefore important to strengthen initial treatment in accordance with guidelines.
In this study, there was no association with the place of arrest and patient outcome when adjusting for compliance with CPR guidelines and other clinically relevant confounders.This implies that there do not seem to be specific factors regarding the treatment provided in monitored areas that influence patient outcome.All patients, regardless of location, were treated with advanced CPR either at the time of collapse or at the arrival of the RRT.The median arrival time for the RRT was three minutes, but with greater variability in non-monitored areas.However, the proportion of missing items for the variable time from collapse to arrival of RRT was 47%, making arrival times difficult to interpret.It is important to note that the initial CPR guidelines evaluated in this study should be performed within the first three minutes, before the arrival of the RRT in most events.These results underline the importance of the early links in the chain of survival concept -the most important factors for patient survival being early recognition and call for help, immediate treatment with high-quality CPR, and rapid defibrillation of patients with a shockable rhythm. 4,5

Clinical implications
Repeated in-situ CPR training according to Swedish resuscitation guidelines, every six months or at least once a year, can strengthen healthcare professionals' ability to quickly recognize a cardiac arrest situation and call the RRT, initiate high-quality CPR, and defibrillate a patient with a shockable rhythm, especially in non-monitored wards.Any possible weaknesses in or obstacles to the links in the chain of survival that could delay the time to initial treatment need to be detected and handled.

Limitations
In this study we only included witnessed IHCA for the possibility to evaluate compliance with initial CPR guidelines.The sample was too small for reliable subgroup analysis of initial rhythm, hence creating a risk of unstable estimates.Also, multiple outcome measures were tested, which may induce a risk of type I error.However, the results of the outcome measures in our study were similar as in a previous study of compliance with CPR guidelines in the Swedish setting. 9Information on time was often missing in relation to initial CPR guidelines, which is a challenge in most IHCA registries.There was a larger proportion of missing items regarding time measures in monitored wards compared to non-monitored wards (41% versus 21%).This is surprising considering that there is equipment that aids to correctly report time measures.Missing information regarding compliance with CPR guidelines is a potential risk of bias.It is also a challenge to report correct time measures in stressful situations, perhaps more difficult in areas without monitoring equipment, which induces a risk of time bias.Post-cardiac arrest care regarding temperature target control and estimation of neurological prognosis has been found to be similar in ICUs throughout Sweden, 34 though there may be differences in hospital resources and care provided that can affect patient outcome.There could also have been unknown factors involved at resuscitation that affected the outcome in this study.

Conclusions
Compliance with initial CPR guidelines was higher on monitored wards than on non-monitored wards for witnessed arrests, which indicates that healthcare professionals on monitored wards are quicker to recognize a cardiac arrest and initiate treatment.When initial guidelines are followed, the place of arrest does not affect the patient outcome.

Declaration of competing interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: 'J.S, M-L.S.K., B.A ¨., A.S. have no affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers' bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript.'.

Table 1 -
Characteristics of witnessed IHCA events on monitored versus non-monitored wards.

Table 2 -
Compliance with initial CPR guidelines and patient outcome of witnessed IHCA events and associations with the place of arrest (monitored versus non-monitored ward).