In-hospital cardiac arrest (IHCA): survival status and its determinants in Malaysian public healthcare

Setting and study design

This retrospective cross-sectional study utilized an internally built IHCA registry database from a tertiary public hospital in Peninsular Malaysia for the years 2018 to 2019. The facility was selected due to the availability of an advanced resuscitation team and as it also functions as the teaching and referral center for the northern region of Malaysia. This hospital is equipped with 1,163 beds and 29 operational general ICU beds. The IHCA database encompasses patient admission details, medical records, clinical data, resuscitation procedures, event timelines, and the patient’s CPC status at discharge (Table S3).

Population and sample

Based on proportions and effect sizes from similar studies, the maximum required sample size was estimated to be 1,230 using OpenEpi software, with a significance level of 0.05 and a statistical power of 0.8 (Chen et al., 2016; Fernando et al., 2019). Since the dataset included only 934 patients, universal sampling was employed. A post-hoc analysis using G-Power confirmed that a sample size of 934 was sufficient for this study.

Inclusion and exclusion criteria

All IHCA patients aged 18 years and above, reported in the IHCA registry between February 1, 2018, and January 31, 2019, were included in this study. This analysis excluded OHCA patients who were successfully resuscitated before admission, those who underwent CPR concurrently with hospital admission, and those who did not receive resuscitation attempts during the study period. All individuals who met the inclusion and exclusion criteria, and experienced cardiac arrest at the study site and receiving care from the nearest healthcare provider (staff) were included.

Data collection for IHCA registry

The primary treatment team initiated resuscitation efforts once the “Code-blue” calls, with ICU staff assisting when needed and all procedure adheres to guidelines for resuscitation (Ministry of Health Malaysia, 2016). After resuscitation, the team documented patient details, resuscitation methods, and cardiac arrest outcomes. Survivors were scheduled for follow-up at their outpatient clinics as standard practice. Data was collected from case report forms (CRFs) completed by the primary medical team at the study site. For missing data, a clinical research associate (CRA) reviewed CRFs, retrieved relevant information from patient case notes, and cross-validated it with the medical team before reporting to the IHCA registry. An automated validation program checked for data discrepancies, including invalid or out-of-range values. The process flow is illustrated in Fig. 1.

Data management and validation

A data collection team was assigned by the core research team to monitor and ensure the quality of the data collection process from the registry department. The list of the selected samples was distributed among the enumerators and data was captured using a standardized data collection form. Variables used in this study were divided into three data domains-patient demographic details, clinical features, and post-arrest outcomes. The data domains and codebook for related data elements are shown in Table S1. There was no direct interaction between the investigators and the subjects as the data was provided by the resuscitating and primary teams.

Statistical analyses

The data collected were entered into Microsoft Excel 365, cleaned, and then transferred to STATA 17.0 for analysis. The incidence rate of IHCA was calculated based on new IHCA cases per 100,000 hospital admissions. The dataset was examined for centrality, dispersion, normality, and missing data. Variables with < 5% missing data were included. Data visualization included histograms and quantile-quantile (QQ) plots for continuous variables and bar plots for categorical variables. Continuous data were reported as mean and standard deviation (SD), while categorical variables were presented as frequency and percentage. The chi-square or Fisher’s exact test was used to assess differences between categorical variables and outcomes.

Logistic regression was performed to identify survival determinants among IHCA patients. Survivors were coded as “1” and deaths as “0.” Explanatory variables included admission reason, admission time, IHCA location, initial cardiac rhythm, witnessed arrests, rescuer qualifications, total shocks, ROSC, adrenaline CPR ratio, total adrenaline administered, and CPR duration. These variables were coded as “1” (yes) and “0” (no). The model building followed the Hosmer-Lemeshow model development strategy. Variables with a p-value < 0.25 in univariate analysis were included in the full logistic regression model. Variables with p-values > 0.05 were removed using backward elimination, guided by the likelihood-ratio test. Effect modification of first-order interactions was assessed. The final model was evaluated for heteroscedasticity, linearity, goodness of fit, AUROC, and multicollinearity. A p-value < 0.05 was considered statistically significant.

Kaplan-Meier survival analysis was also applied to estimate the survival projections at key time points. This study examined survival rates from survival to discharge or 30 days following Utstein-reporting guidelines (Nolan et al., 2019). Survival time (T) was defined as the interval from resuscitation initiation to discharge or death. Survival status was coded as “0” (censored, non-death) and “1” (death, event). The Kaplan-Meier survival curve was plotted using Jamovi 2.3 for visualization. In this study, the data elements, outcomes of cardiac arrest, and cardiopulmonary resuscitation were referenced according to the Utstein resuscitation registry template (Table S1) for in-hospital cardiac arrest (Nolan et al., 2019).

Ethics statement

This study was conducted according to the guidelines of the Declaration of Helsinki. This study was registered under the National Medical Research Register (NMRR), Ministry of Health Malaysia (NMRR ID-22-02528-FV7). Informed consent was not required, as the data were obtained anonymously from the registry without patient identifiers.

In-hospital cardiac arrest survival status

The Kaplan-Meier survival plot (Fig. 3) and Table 2 show a rapid decline in IHCA survival, dropping by 84.6%, leaving only 15.4% (95% CI [13.28–17.96]) of survivors within the first 24 h post-resuscitation. The survival rate declines to 8.4% by day 7, 3.6% by day 30, and 2.7% by day 60, with an overall median survival time of 0.80 h (95% CI [0.7–0.8]).

Factors associated with survival of IHCA

Nine variables with p < 0.25 from the univariate analysis (Table 3) were included in the final logistic regression model. These variables were the location of the IHCA event, first documented cardiac rhythm, arrest witnesses, CPR duration: adrenaline dosage ratio, age, ROSC, resuscitation duration (minutes), rescuer ALS training, and rescuer designation. Among these, the final logistic regression analysis (Table 4) identified three significant factors (p < 0.05) positively associated with IHCA survival: IHCA occurring in critical care areas (ICU, HDU, HDW, and ED), a shockable first documented cardiac rhythm, and a CPR duration: adrenaline dosage ratio of ≤ 3 min per 1 mg of adrenaline administered. Arrest witnesses remained a key factor for model fit and stability. The model showed a good fit, with multicollinearity excluded for all variables, including arrest witnesses, as shown in Table 4.

Determinants of IHCA outcomes

In this study, the shockability of the initial cardiac rhythm emerged as a key independent factor influencing survival among IHCA cases. These findings were supported by studies that observed the positive relationship between the survival rates and the presence of shockable rhythms such as ventricular tachycardia (VT) and ventricular fibrillation (VF) (Fennessy et al., 2016; Aziz et al., 2018). Timely defibrillation significantly improves survival outcomes in these cases (Andersen et al., 2019a). However, a study from the USA reaffirms that survival is not solely dependent on rhythm shockability but also on the effectiveness of acute resuscitation efforts and post-resuscitation care (Girotra et al., 2012). This underscores the need for a comprehensive approach to IHCA management that extends beyond initial rhythm assessment.

A high proportion of non-shockable rhythms (91.8%) was observed in this study, similar to findings from Abu Dhabi (91.1%) (Aziz et al., 2018) and Turkey (87.9%) (Yilmaz & Omurlu, 2019) but higher than reports from other countries. According to Denmark’s IHCA registry, non-shockable rhythms are often linked to non-cardiovascular conditions, particularly sepsis (Stankovic et al., 2021). Given the high proportion of sepsis in our study cohort (Fig. 2), this could explain the predominance of non-shockable rhythms. These findings emphasize the importance of early sepsis detection and management in preventing the deterioration of IHCA (Yakar et al., 2022). Studies indicate that public awareness of sepsis remains low (Parsons Leigh et al., 2022), yet targeted sepsis education has been shown to improve recognition and early intervention (Fiest et al., 2022). In resource-limited settings, healthcare providers could leverage opportunistic education strategies during routine consultations to inform high-risk patients about sepsis symptoms and timely medical intervention.

Another critical observation in this study was the lower-than-recommended adrenaline administration interval, with infusion occurring in less than 3 min, despite protocol recommendations of three to 5 min (American Heart Association, 2000). While shorter total resuscitation durations (<20 min) and adequate adrenaline dosages were observed, a high proportion of unfavorable neurological outcomes were also noted upon discharge. This raises concerns about the potential neurological impact of early adrenaline administration. Observational data suggest that while earlier adrenaline use may improve ROSC, administration within 2 min of the first defibrillation shock may be detrimental (Gough & Nolan, 2018). A large trial funded by the National Institute for Health and Care Research (NIHR) found that although adrenaline increased 30-day survival, it was associated with a higher risk of severe brain damage (Perkins et al., 2018). Similarly, other studies indicate that adrenaline does not necessarily correlate with long-term favorable neurological outcomes (Ludwin et al., 2021). Given these uncertainties, regular audits and follow-up studies on adrenaline and other clinical parameters used during resuscitation are necessary to refine their role in IHCA management. Based on the study findings, IHCA recognition and survival may be improved through effective early warning systems, continuous monitoring, and rapid response activation. High-quality CPR and early defibrillation are critical, supported by training and real-time feedback. A well-coordinated Code Blue team with clear roles enhances response efficiency. Post-resuscitation care, including temperature management and hemodynamic support, aids patient recovery.

Policy implication

Our findings underscore the significant association between the location of cardiac arrest and IHCA survival rates. Patients who experienced cardiac arrest in monitored areas, such as the ICU, Cardiac Care Unit (CCU), and emergency care units, had notably higher survival rates than those in general wards (Sandroni et al., 2004). This suggests that continuous monitoring and rapid intervention play a protective role, even in critically ill patients. Similarly, our study highlights the importance of the location of IHCA events, as areas with advanced monitoring such as the ICU, CCU, and emergency units showed higher IHCA survival rates. This is due to the readily available monitoring system, rapid access to resuscitation equipment, trained staff, and heightened vigilance, enabling faster intervention and better critical care management (Perman et al., 2016; Vo et al., 2024). To improve survival in general wards and facilities where ICU beds are limited, a systematic approach to patient bed allocation is necessary (Sandroni et al., 2004; Pangan et al., 2021). Identifying high-risk individuals and ensuring they receive enhanced monitoring and immediate resuscitation support can mitigate delays in advanced life support (ALS). However, the absence of local risk stratification guidelines complicates this process (Starks et al., 2018). Developing a comprehensive IHCA registry tailored to the Malaysian healthcare system would help identify at-risk patients and inform evidence-based preventive strategies (Chan et al., 2023; Vo et al., 2024).

Beyond initial resuscitation efforts, post-resuscitation care plays a crucial role in determining survival outcomes. Our study found that IHCA location, witnessed arrests, shockable rhythms, and CPR adrenaline duration independently influenced patient survival. This suggests that factors beyond rapid defibrillation, such as targeted temperature management, arterial line monitoring, perfusion maintenance, and early cardiac catheterization, are essential components of IHCA management (Nolan et al., 2015). These interventions are more likely to be available in ICU settings than in general wards, reinforcing the critical role of intensive care services in improving IHCA outcomes. Therefore, strengthening post-resuscitation protocols and ensuring their accessibility beyond ICU settings should be a priority.

The disparity between high ROSC rates and low survival-to-discharge rates in our study highlights the urgent need to expand intensive care capacity. A 2015 report indicated that Malaysian public hospitals had a median ICU bed occupancy rate of 91.6%, with approximately 32% of patients denied ICU admission due to bed shortages (Ling et al., 2015). Our study hospital reported ICU occupancy rates ranging from 89.0% to 94.0%, further emphasizing the resource strain. To improve IHCA survival, expanding ICU infrastructure and optimizing resource allocation are critical. In the short term, deploying additional critical care personnel, advanced monitoring systems, and resuscitation equipment to high-risk areas such as general wards can help bridge the gap in post-resuscitation care.

Study limitations

A key limitation of this pioneering study, despite its substantial sample size of 934 cases, is its single-center design. This restricts the ability to generalize findings across the broader Malaysian healthcare system. While the study provides valuable insights into IHCA incidence, survival rates, and associated factors, it does not account for several critical determinants that could influence patient outcomes. Notably, comorbidities, post-resuscitation vital signs, delays before the first rhythm assessment, certain laboratory parameters, and medication usage were not analyzed due to registry data limitations. These factors are essential in understanding the full spectrum of IHCA predictors and interventions, and their exclusion may limit the study’s ability to offer a comprehensive assessment of survival determinants. Thus, it is recommended that the relevant specialized department improve the IHCA registry, ensuring compliance with the Utstein resuscitation registry template for in-hospital cardiac arrest reporting guidelines (Table S1). Future research incorporating multicenter data and a more extensive range of clinical variables would be beneficial in refining IHCA management and policy strategies.