In-hospital cardiac arrest (IHCA) remains a major cause of morbidity and mortality in acute care hospitals. Recent international literature suggests that survival to discharge after IHCA has increased in the last two decades, however remains between 15% to 25%, depending also upon patient population and arrest characteristics and rhythm.¹-³ Australian data demonstrates similar outcomes, with mortality exceeding 74%, again dependent upon patient and hospital demographics and arrest characteristics.⁴ Despite technological and systems improvements in rapid response and resuscitation science, outcomes remain particularly poor among elderly patients and those with significant comorbidity.³,⁵

Advance care planning (ACP) and formal resuscitation planning are critical components of high-quality, patient-centered care. The Australian Commission on Safety and Quality in Health Care emphasises that goals-of-care discussions should occur early and be clearly documented to guide clinical decision-making during deterioration.^6,7 Similarly, NSW Health Policy Directive PD2014_030 outlines that resuscitation plans should be discussed and formalised for patients at risk of clinical deterioration or where CPR may be inappropriate.⁸  Patients with progressive advanced life limiting illnesses or poor 6-12 month survivability prognosis should have resuscitation planning discussion and formalisation during their hospital admission. ⁸

Despite these policy frameworks, resuscitation decision-making in acute hospitals frequently remains reactive rather than anticipatory. Goals-of-care discussions are often deferred until significant clinical deterioration occurs, including during rapid response activation or at the time of cardiac arrest itself^.9–11 Decision-making under such crisis conditions may limit meaningful patient participation, increase emotional burden on families, and reduce the likelihood that care aligns with previously expressed values.¹² Furthermore, in the absence of clearly documented resuscitation plans, patients who may have been clinically inappropriate candidates for CPR can undergo invasive and potentially non-beneficial interventions, including chest compressions, endotracheal intubation, and escalation to intensive care. Beyond the risk of patient harm, inappropriate or unwanted resuscitation attempts may contribute to moral distress among healthcare professionals, conflict with families, and utilisation of substantial healthcare resources in circumstances where the probability of meaningful recovery is extremely low.^{13, 14}

Evaluating the prevalence of formalised resuscitation planning prior to IHCA provides an important measure of whether advance care planning processes are functioning as intended. If patients experiencing cardiac arrest lack documented resuscitation plans despite identifiable high-risk features during admission, this may reflect missed opportunities for earlier goals-of-care discussions and formalisation. Accordingly, this study aimed to evaluate the prevalence of documented resuscitation plans prior to ward-based IHCA in a metropolitan Australian hospital and to identify potential systems gaps in advance care planning.

Methods

Study Design and Setting

A retrospective observational study was conducted at a 433-bed metropolitan referral hospital in Sydney, Australia, serving a culturally and linguistically diverse population.

Inclusion Criteria

• Patients aged ≥18 years
• Ward-based IHCA between 1 January 2022 and 31 December 2024

Exclusion Criteria

• Arrests occurring in ICU
• Arrests in emergency department
• Operating theatres or cardiac catheter laboratory
• Outpatient settings

Data Collection

Electronic medical records were reviewed. Extracted variables included:

• Age and gender
• Patient medical record number (MRN)
• Admitting specialty and ward
• Time from admission to IHCA
• Presence of formal resuscitation plan at time of arrest
• Documentation of goals-of-care discussions during admission
• Prior MET activation during admission
• ICU step-down status
• Immediate survival (ROSC)
• Survival to discharge

Statistical Analysis

A standardized date collection tool was developed in Microsoft Excel to ensure consistent extraction of data using the above extracted variables. A separate independent investigator extracted data for all in-hospital deaths from 2022 to 2024. The medical record number (MRN) of each patient who was included in the retrospective study was searched in both data sets to ensure accuracy of collected data.

The data underwent descriptive analysis, with descriptive statistics calculated including frequencies, percentages, means, medians, and ranges.

Subgroup analyses were performed to compare rates of resuscitation plan formalisation amongst patients who had previously required a medical emergency team (MET) call during their admission, or those who had been stepped down from ICU compared to those who had not.

Additionally, comparisons were made between the rates of documentation of discussions between medical teams and patients regarding resuscitation planning and the formalisation of plans.

Figure 1: Inclusion of cases and survivability data

Results

Patient Characteristics

A total of 64 patients were included in the analysis. The cohort was predominantly male, comprising 64% (n = 41) of patients, with females accounting for 36% (n = 23).

The mean age was 76.2 years, with a median age of 78.0 years (range 45–97 years), reflecting an elderly population with a broad age distribution.

Most patients were admitted under medical specialties. Geriatrics was the most common admitting team, accounting for 30% (n = 19) of cases, followed by Cardiology at 22% (n = 14) and Respiratory Medicine at 19% (n = 12). Renal Medicine accounted for 9% (n = 6) of admissions, and General Surgery for 5% (n = 3). Other admitting specialties collectively represented 16% of cases and included Medical Oncology (5%, n = 3), Colorectal Surgery (3%, n = 2), and single cases (2% each) under Neurology, Gastroenterology, Rehabilitation Medicine, Upper Gastrointestinal Surgery, and Ear, Nose and Throat Surgery.

More than half of all in-hospital cardiac arrests occurred on three medical wards: 2G (19%, n = 12), 3F (17%, n = 11), and 3H (16%, n = 10), together accounting for 52% of events. Additional arrests occurred on 2J (11%, n = 7) and 3B (9%, n = 6). The remaining cases were distributed across several other wards, including 3J (8%, n = 5), 3A (5%, n = 3), 2D (5%, n = 3), 2A (2%, n = 1), and 2B (2%, n = 1).

Only 3 patients (4.7%) had a completed resuscitation plan at the time of IHCA. One was documented as “for CPR” and two as “not for CPR”.

Thirteen patients (20%) had documentation of goals-of-care  discussions during admission. Five Patients were identified as ‘for CPR’. Of the eight patients identified in discussion as not appropriate for CPR, only two had formalised “not for CPR” orders.

Clinical Deterioration Triggers

Ten patients (15.6%) had a prior MET call during admission. Only two had resuscitation planning discussions following MET review. In one case, ICU recommended limitation of resuscitation; however, this was not formalised.

Eight of the sixty-four included patients had been stepped down from ICU prior to arrest. Only one had a completed resuscitation plan at the time of ward transfer.

Outcomes

Return of spontaneous circulation occurred in 37.5% of patients. Survival to discharge was 17.2%. Overall in-hospital mortality was 82.8%

Figure 1: Documentation of goals of care discussions and formalisation of resuscitation plans

Figure 2: Resuscitation planning following MET Calls

Figure 3: Resuscitation planning for ICU-Stepdown patients