Introduction

There are approximately 292,000 in-hospital cardiac arrests (IHCA) in the United States every year, with one in four survivors [1]. The factors associated with survival and favorable neurological outcomes following IHCA are poorly understood. Girotra et al showed improvements in IHCA survival to discharge from 13.7% in 2000, up to 22.3% in 2009. Among those survivors, 85% were discharged from the hospital with a favorable neurological outcome [2]. Survival following IHCA decreases with increasing age, especially among elderly patients (age > 70) [3]. Kagawa et al found IHCA patients were often older and sick with underlying co-morbid conditions [4].

Altered metabolic parameters, inadequate organ perfusion, decompensation of comorbid conditions and medical interventions are some of the variables that can precipitate a cardiopulmonary arrest in the ICU. Moskowitz et al found that delay in responding to impending respiratory failure, incorrect diagnosis, and delay in responding to clinical deterioration were some of the preventable targets for cardiac arrest in ICU [5]. Approximately 59% of all IHCA occur in the ICU, with only 14% survival during the hospitalization [6].

Variations in the reported incidence of cardiac arrest, the relevance of these findings to the specific healthcare system, and the potential for under-reporting plague our understanding of IHCA and CPR outcomes in the ICU [7]. In our retrospective observational study, we attempt to identify patient characteristics of IHCA in the ICU and determine the impact of duration of cardiopulmonary resuscitation (CPR) on survival outcomes. We hypothesize that prolonged CPR in critically ill patients admitted to the ICU is associated with higher mortality among those who survive the initial resuscitation.

Materials and methods

Results

The baseline characteristics of our study population and characteristics of IHCA based on the duration of resuscitation are noted in Table 1 and Table 2 respectively. The mean age was 61 years (IQR, 51–73) of which 55% were men. The mean duration of resuscitation was 11.5 minutes (IQR, 4–17). Return of spontaneous circulation (ROSC) was obtained in 224 patients (61%) of which 82 survived to ICU discharge (23.4%) and 60 patients (16.3%) survived to hospital discharge. Prolonged CPR (>15 minutes) was associated with poor outcomes (N = 6; 5.5%).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Patient characteristics.

https://doi.org/10.1371/journal.pone.0340658.t001

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Characteristics of cardiac arrests in the ICU.

https://doi.org/10.1371/journal.pone.0340658.t002

Survival based on duration of CPR

After adjusting for age, APACHE III score, and initial rhythm; every one-minute increase in CPR duration was associated with 1.161 (95% CI 1.119–1.204; p < 0.0001) odds of death during acute resuscitation and 1.154 (95% CI 1.059–1.258; p < 0.0001) odds of death at the time of ICU discharge.

The Kaplan-Meier plot illustrating survival probability relative to resuscitation duration is presented in Fig 1. Additionally, Fig 2 displays patient outcomes stratified by 5-minute increments of CPR duration.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Kaplan-Meier survival curve illustrating the relationship between CPR duration (minutes) and survival probability.

https://doi.org/10.1371/journal.pone.0340658.g001

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Rates of acute resuscitation survival, ICU discharge survival, and hospital discharge survival according to CPR duration.

Survival rates across all three clinical endpoints demonstrate a steady decline as the duration of resuscitation increases.

https://doi.org/10.1371/journal.pone.0340658.g002

Hospital survival

Sixty patients survived to hospital discharge (16.3%). Hospital survivors were marginally younger (mean age of 59; SD 18) of which 55% were men. The APACHE III score (Mean = 88.3, SD 29.8, IQR 66–106) and acute physiology score (Mean = 75, SD 30, IQR 56–94) were lower compared to non-survivors. Hospital survivors were more likely than non-survivors to have a shockable rhythm at the time of arrest (20% vs 7.5%), shorter average CPR duration (5.4 minutes versus 12.8 minutes), a longer length of ICU stay (14 days versus 1.8 days) and longer length of hospital stay (25 days versus 6.1 days). Characteristic features of hospital survivors and non-survivors are listed in Table 3.

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Characteristics of hospital survivors compared to non-survivors.

https://doi.org/10.1371/journal.pone.0340658.t003

ROSC Achievement (Table 4): Out of 367 total patients, 224 (61%) successfully achieved Return of Spontaneous Circulation (ROSC). The duration of CPR was drastically shorter for patients who obtained ROSC compared to those who died without achieving it (Median: 5.0 minutes vs. 18.0 minutes; p < 0.0001).

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Outcome at ROSC.

https://doi.org/10.1371/journal.pone.0340658.t004

ICU Survival Post-ROSC (Table 5): Among the 224 patients who initially achieved ROSC, 82 (36.6%) survived to ICU discharge. Patients who survived the ICU had significantly shorter resuscitation times than those who achieved ROSC but subsequently died in the ICU (Median: 4.0 minutes vs. 7.0 minutes; p = 0.0001).

Download:

• PNG
  larger image
• TIFF
  original image

Table 5. ICU survival for patients with ROSC.

https://doi.org/10.1371/journal.pone.0340658.t005

Discussion

In our study, we report ICU-IHCA survival of 16.3% at hospital discharge in comparison to previously reported 14% survival to discharge after IHCA in ICU patients over the years 2003–2010 [6]. Improved survival rate in our study population is likely related to continued advancements in resuscitation methods and clinical care over the last decade [8,9].

In our analysis, patients with CPR lasting longer than 15 minutes were relatively younger with a mean age of 59 (SD 15.9), compared to patients with CPR lasting less than 5 minutes with a mean age of 64 (SD 14.9). Khan et al reported that younger age (18–40 vs > 65 years; odds ratio [OR] 1.81; 95% CI 1.69 to 1.95; P < 0.001) and female sex (OR 1.05; 95% CI 1.02 to 1.09; P = 0.005) were associated with longer duration of resuscitation in patients who did not obtain ROSC after cardiac arrest [10]. It is assumed that this observed difference could be related to a perception among health care workers that younger patients have a higher probability of survival and merit a longer duration of resuscitation [9]. In contrast to our study results, a nationwide Swiss data analysis by Amacher et al reported women were less frequently admitted to ICUs, received fewer advanced treatments, and had a higher risk of ICU mortality when compared to men [11].

Patients with chronic illnesses such as CHF, CKD, COPD, malignancy, and cirrhosis who had IHCA were found to have significantly higher mortality in our analysis. Stapleton et al reported increased risk for death after IHCA among older adults with severe COPD (Hazard ratio 1.39; 95%CI 1.32–1.48; p < 0.001), severe CHF (HR 1.81; 95% CI 1.75–1.87; p < 0.001), severe CKD (HR 1.58; 95% CI 1.52–1.64; p < 0.001), severe cirrhosis (HR 1.69; 95% CI 1.40–2.05; p < 0.001) and malignancy (HR 1.70; 95% CI 1.58–1.84; p < 0.001) when compared to patients without those co-morbid conditions [12]. Our patients were acutely ill with a mean APACHE III score of 102.3. In our study population, 17.4% were on chronic dialysis, 32.5% had diabetes, 17.4% had severe COPD, 11.7% had cirrhosis and 30.8% were immunosuppressed (Table 1).

Furthermore, 27% of our IHCA patients had sepsis with a mortality rate of 87% at hospital discharge. This is similar to IHCA mortality in patients with severe sepsis reported by Koivikko et al [13]. Champigneulle et al reported a six-month survival rate of 14% among cancer patients admitted to the Intensive care unit (ICU) after a cardiac arrest [14]. In our cohort, 25% had an underlying malignant condition (leukemia, lymphoma, or solid tumor with metastasis) with a mortality rate of 93% at hospital discharge.

In our study, we noticed an increase in mortality proportional to the duration of resuscitation. Fernando et al in their meta-analysis of 23 cohort studies reported lower odds of survival (OR 0.12; 95% CI 0.07 to 0.19; p < 0.001) in patients with CPR longer than 15 minutes [15]. Radeschi et al studied the outcomes of IHCA in Italy and reported the median duration of CPR for survivors was 8 minutes and 26 minutes for non-survivors [16]. Kantamineni et al and Rafati et al also reported better outcomes when CPR lasted less than 10 minutes [17,18]. In our study, the mean duration of CPR was 5.2 minutes (IQR 2–6) for patients who survived to ICU discharge; and 16.4 minutes (IQR 7–24) for patients who died during acute resuscitation.

Prolonged resuscitation is associated with various traumatic and non-traumatic complications [19]. There is no clear definition for prolonged CPR as the ideal duration of resuscitation is unknown. Loisa et al reported that approximately 14% of IHCA are terminated early due to presumed futility [20]. American Heart Association’s cardiac arrest guidelines do not have any specific recommendations on termination of CPR.

Lately, clinical decision rules have been developed to assist physicians in terminating potentially futile resuscitation attempts [21']. Val Walraven et al proposed a model (the UN10 rule) for ‘futile CPR’ incorporating three intra-arrest variables i) unwitnessed arrest ii) non-shockable rhythm and iii) ROSC not obtained in 10 minutes [22,23]. Petek et al reported that among patients who meet the UN10 criteria for futile resuscitation, only 6.3% survived to discharge and 4.8% survived with favorable neurological function [23]. Our cohort had witnessed cardiac arrest in a monitored hospitalized setting, however, among non-survivors, 93% had a non-shockable rhythm and an average CPR duration of 12.8 minutes.

Conclusion

Based on our study findings, the mortality after IHCA increases with the increase in the duration of CPR regardless of age, initial rhythm, and severity of underlying illness. Further studies using clinical decision rules like GO-FAR score might assist the healthcare team in selecting appropriate patients who might benefit from prolonged resuscitation.