Introduction

Cardiac arrest is defined as the ‘sudden cessation of heart beat and the abrupt loss of cardiac mechanical activity in a person who may or may not have been diagnosed with heart disease, confirmed by the absence of a detectable pulse and unresponsiveness of apnea’ [1–3]. Depending on the scene, cardiac arrest can be categorized as either out-of-hospital cardiac arrest (OHCA) or in-hospital cardiac arrest (IHCA) [4, 5]. There are approximately 350,000 and 275,000 adult OHCA patients annually in the United States and Europe, respectively [6–8]. OHCA remains a public health concern.

The chain of survival provides a standard protocol for resuscitation and treatment. A strong chain of survival has been shown to improve the chances of survival and recovery for victims. In 2010, the chain of survival for adult OHCA was extended to five links: early recognition, early cardiopulmonary resuscitation (CPR), rapid defibrillation, emergency medical services (EMS), advanced life support and in-hospital post-cardiac arrest care. Rapid transport to the hospital for advanced care is determined by the classification of diseases and clinical characteristics [9], and post-arrest care has been demonstrated to significantly impact the survival of OHCA patients.

Approximately 60% of OHCA events are treated by EMS [10, 11]. The pooled rate of survival to admission was 23.4% [12]. The reported rate of survival to hospital discharge ranges from 0.2% to 37.8% [13–16]; however, a meta-analysis yielded a pooled rate of 7.6% [12]. Even when EMS was activated promptly, the rate of survival to discharge among EMS-treated OHCA patients was approximately 10% [17].

Studies have documented a variety of factors associated with prognosis after OHCA, including patient variables (e.g., age, gender, and comorbidities) [18, 19], event variables (e.g., initial cardiac rhythm, time to CPR and defibrillation) [7, 18] and hospital variables (e.g., level and post-cardiac arrest care) [20].

The concept of how post-cardiac arrest care influences the survival of OHCA patients admitted to a hospital has been highlighted in recent years. Yet, to the best of our knowledge, only a few hospital-based studies have investigated this issue in Taiwan. To bridge this gap, we analyzed the health insurance database to examine the survival outcome of hospitalized OHCA patients. Next, we identified factors associated with improved survival in terms of survival to discharge.

Materials and methods

Results

A total of 11,000 cases of hospitalized OHCA from 2007 to 2013 were collected. In all, 2,499 patients (22.7%) survived to hospital discharge. As shown in Table 1, the mean age of subjects who survived to hospital discharge and those who did not was 66.7±16.7 and 71.7±15.2 years, respectively (P<0.001). The percentage of males was 59.0% in the discharge group and 58.2% in the non-discharge group. Respiratory failure was the most common problem among the clinical characteristics. Heart, renal, respiratory, and metabolic failure were less frequent in the discharge group; inversely, neurological failure was more likely in the discharge group (discharge: 24.5%; non-discharge: 17.2%; P<0.001). There were higher proportions of hypertension, CAD, valvular heart disease, cardiomyopathy and atrial flutter/fibrillation in patients who survived to discharge. However, malignancy was less frequent in the discharge group (discharge: 4.6%; non-discharge: 8.1%; P = 0.001).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Demographic and clinical characteristics of hospitalized OHCA patients.

https://doi.org/10.1371/journal.pone.0191954.t001

Survivors were more likely to have been admitted to a higher hospital level with a higher volume. The geographic area of the admission hospital was associated with the rate of survival to discharge. In the discharge group, 89.7% of the hospitals were teaching hospitals, while 84.8% were teaching hospitals in the non-discharge group (P<0.001). Details of the hospital characteristics are given in Table 2.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Hospital characteristics of hospitalized OHCA patients.

https://doi.org/10.1371/journal.pone.0191954.t002

As shown in Table 3, the percentage of patients diagnosed with VF on admission was 26.5% in the discharge group. In all, 7.4% subjects were transferred to another hospital; the percentage was higher in the discharge group (13.9%) than in the non-discharge group (5.5%) (P<0.001). Length of hospital stay was positively associated with prognosis. There was no significant difference in the distribution of patients who survived to discharge from 2007 to 2010 and from 2011 to 2013.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Event characteristics of hospitalized OHCA patients.

https://doi.org/10.1371/journal.pone.0191954.t003

As shown in Table 4, multivariate analysis demonstrated that each yearly increment in age decreased the probability of survival to hospital discharge by 2% [adjusted odds ratio (aOR): 0.98; 95% confidence interval (CI): 0.98–0.99]. The number of failing organs was negatively related to survival to hospital discharge (aOR: 0.75; 95% CI: 0.71–0.80). Cardiac comorbidities were independent predictors of improved survival. In terms of the hospital level, patients admitted to a medical center had a better prognosis than those admitted to a regional hospital (aOR: 0.76; 95% CI: 0.67–0.87). Cardiac rhythm was a strong predictive factor associated with survival to discharge (VF vs. non-VF: aOR: 3.51; 95% CI: 3.06–4.01). Patients transferred to another hospital were 59% more likely to survive to discharge (aOR: 1.59; 95% CI: 1.30–1.93). Meanwhile, the aOR of survival to discharge was 1.01 for each daily increment in hospital stay (aOR: 1.01; 95% CI: 1.01–1.02). The rate of survival to hospital discharge was not significantly elevated by calendar year (P for trend = 0.074) (Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Temporal trend of survival to hospital discharge among hospitalized out-of-hospital cardiac arrest patients.

https://doi.org/10.1371/journal.pone.0191954.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Survival factors for hospitalized OHCA patients.

https://doi.org/10.1371/journal.pone.0191954.t004

Discussion

In our study, 2,499 subjects survived to hospital discharge, with an overall rate of 22.7%, from 2007 to 2013 in Taiwan; however, the temporal trend of survival to hospital discharge was static over the study period. Cardiac rhythm coded VF on admission was a confirmed dominant independent variable.

Previous results have shown survival to vary widely from 8.3% to 38.0% [23, 26, 27]. Although we extracted subjects using ICD-9-CM codes, the inclusion/exclusion criteria made little difference due to the database characteristics. In addition, studies have recruited different types, such as those resuscitated by EMS, treated by teams in the hospital ED, or surviving to hospital admission. To elucidate the effect of post-cardiac arrest care, we only analyzed the survival of OHCA patients admitted to a hospital; the do-not-resuscitate (DNR) policy is not widely selected in Taiwan, which could result in discrepancies with the results of other studies. However, aside from this issue, post-cardiac arrest care indirectly influences survival in various ways.

Previous results have indicated that the rate of OHCA survival has not dramatically changed for decades [12]. In our study, the temporal change in the rate of survival to hospital discharge showed no significant improvement, similar the results of a study performed in Canada [27]. Nevertheless, data from Sweden indicated that out-of-hospital death decreased by 4% each year [28], and the in-hospital mortality rate of patients admitted to a hospital for cardiac arrest decreased by 11.8% in the United State [25]. After implementation of the chain of survival concept for post-resuscitation care, one-month survival with a favourable neurological outcome in OHCA patients increased by 7.8-fold [6]. There are potential explanations for the discrepancies in these temporal trends. First, most studies have focused on the survival improvement of OHCA patients suffering from conditions with a cardiac aetiology, and it has been shown that a condition having a cardiac origin was a positive predictor for outcome. Second, eligible patients were more likely to include survivors transported from an out-of-hospital location to an ED. In our study and in a former study [27], only subjects who survived to hospital admission were tracked. Finally, such non-identical outcomes seemed to have led to different conclusions. In our study, the rate of survival to hospital discharge was applied to demonstrate the effectiveness of post-cardiac arrest care.

Recognizing different factors may help to improve post-cardiac arrest care. As found by previous studies, age was a predictor of outcome [25, 29]. We found that the mean age of the discharge group was 5 years younger than that of the non-discharge group. Our results were also consistent with those of several studies in that the risk of in-hospital mortality increased with the number of failing organs and comorbidities [23, 25]. However, the incidence of neurological failure was higher in the discharge group. This result is probably because patients who survived to discharge tended to have symptoms of neurological impairment recorded during their hospitalization. In other words, non-survivors might expire before being evaluated for neurological function. Consistently, the results of a previous study indicated a higher percentage of survival to discharge among OHCA patients with mild/moderate neurological and performance disability [16]. Because there were no data regarding cerebral performance category score [30], we could only track neurological failure by ICD-9-CM code, which does not indicate the severity of brain damage. Cardiac comorbidities were likely associated with VF or ventricular tachycardia (VT); thus, they were rationally positive predictors of survival [31]. Our findings were similar to those of Dudas et al., who reported that malignancy was a risk factor for survival [28]. Since 2000, when the Hospice Palliative Care Act was passed in Taiwan, the DNR policy has prevented unnecessary CPR. Therefore, the CPR success rate was lower in cancer patients than in non-cancer patients in Taiwan, and cancer seemingly had a negative effect on survival [32].

In larger hospitals, patients benefit from higher intensity service, more rapid action, and new emerging technology. The present study found that a higher volume of ICU beds was strongly associated with survival to hospital discharge, while other hospital characteristics have been linked to in-hospital mortality and discharge among OHCA patients [5, 27, 33, 34]. Therefore, variations in the delivery and quality of healthcare usually result in different OHCA survival outcomes. In accordance with previous observations, our analysis revealed that the rate of discharge increased with ICU bed number. However, the variables hospital level, bed number, and ICU bed number might have redundant effects on the primary outcome. Our analysis showed that ICU bed number is likely to be a much better independent predictor in post-cardiac arrest care.

International findings revealed that the initial cardiac rhythm of patients, VF/VT, has an enormous impact on OHCA survival [7, 29]. In this context, cardiac rhythm on admission was also a profound factor correlated with survival to discharge. Hospitalized patients with VF were more than three times as likely to survive to discharge than those with a non-VF rhythm. In our study, 11.3% of OHCA patients who survived to hospital admission presented with VF. However, studies have illustrated higher percentages of VF/VT. Data from New Zealand showed that close to half of cardiac arrest patients presented with VF [15]. Approximately one quarter of OHCA patients in Sweden presented with a VF/VT rhythm [35, 36]. These differences might be because we retrieved cases using the primary diagnostic code assigned upon hospital admission, and information regarding the initial cardiac rhythm occurring before admission might be lost. Furthermore, in OHCA, a secondary rhythm is likely to be present on admission. Additionally, the studies mentioned above limited their subjects to those with OHCA of a cardiac origin and a higher proportion of VF/VT rhythms. In summary, our finding that cardiac rhythm on admission influenced survival is in agreement with previous reports. Two other variables, transfer to another hospital and length of hospital stay, were positively associated with survival. In clinical practice, resuscitated OHCA patients with the return of spontaneous circulation are sent to the ICU for advanced nursing care. If patients are diagnosed with this event in a small hospital with little infrastructure, they would be transferred to a well-equipped and experienced care center or a larger hospital. In addition, stable subjects who survived were more likely to have been transferred to another hospital closer to their home than were non-survivors in Taiwan. Thus, survivors, who recovered better, tended to stay longer in a hospital than non-survivors [23,24,34], which might be why referral to another hospital and the length of hospital stay contributed to a positive outcome. In the future, more studies of patient movement between hospitals should investigate this issue further.

Conclusions

In conclusion, we utilized a structural framework to investigate OHCA survival. No differences in survival were found between the examined calendar year periods. The results show that factors involved in cardiac comorbidities, hospital level, ICU bed number, cardiac rhythm on admission, transfer to another hospital and length of hospital stay have a significant positive association with survival to hospital discharge. Our findings also suggest that age, number of failing organs, and malignancy are negatively related to survival to discharge.

Acknowledgments

This study was based in part on data from the NHIRD provided by the NHIA, Ministry of Health and Welfare, managed by the NHRI. The interpretation and conclusions contained herein do not represent those of the NHIA or NHRI.