Introduction

A cardiopulmonary arrest can be defined as the cessation of cardiac mechanical activity as confirmed by the absence of signs of circulation [1]. When this condition occurs in a hospital setting, it is referred to as an in-hospital cardiac arrest (IHCA). An out-of-hospital cardiac arrest (OHCA) takes place in prehospital environments.

Cardiac arrests are more frequently seen in people with advanced age. A study in Japan reported that patients aged 65 years and older accounted for more than 80% of all patients with OHCA [2]. In view of the rapid, worldwide expansion of this older section of the population, a higher rate of cardiac arrests can be expected [3].

Cardiopulmonary resuscitation (CPR) can be lifesaving for persons with cardiopulmonary arrest. However, patients who survive the initial resuscitation are at risk of anoxic brain damage [4, 5]. In addition, significant emotional disorders such as anxiety, depression and post-traumatic stress disorder (PTSD) are frequent in those who survive [6]. Therefore, in some cases, the use of CPR may lead to prolongation of the dying phase and suffering without adding to quality of life.

As such, it is important that resuscitation attempts occur in accordance with the preferences of the well-informed patient [7,8,9]. The driving force for their decision regarding resuscitation is quality of life post-resuscitation. Many patients are unrealistically optimistic about the results of CPR and thus make uninformed choices. Relevant and correct information, once provided, influences the final decision considerably [10]. It is up to the healthcare professional to explain the success rate and possible harmful effects of CPR to the patient and their family [11].

To provide accurate information and make the optimal decision for each individual, it is important for healthcare professionals to identify those subgroups with good chances of a satisfactory post-arrest quality of life [12, 13].

In this scoping review, we aimed to collect all available and recent evidence on outcome regarding survival and quality of life after CPR following both IHCA and OHCA in the older population, including indicators of poor prognosis.

Methodology

Search strategy

A search was conducted to find the most recent systematic reviews concerning survival and quality of life following IHCA and OHCA in patients 70 years or older. Subsequently, a search was conducted to find all primary studies published after these systematic reviews to acquire the most recent available evidence. This was done separately for IHCA and OHCA, using analogous search entries.

Review selection The MEDLINE, EMBASE, PubMed and Google Scholar databases were electronically searched in October 2018 to find reviews concerning CPR outcomes in the older population. The search strategy cross-referenced ‘elderly’, ‘resuscitation’ and ‘functional outcome’ using appropriate synonyms, medical subject headings (MeSH), keywords and filters (Appendix 1 in ESM). The references from relevant reviews were also scanned for additional studies. The remaining titles were scanned to exclude other study types (such as randomized controlled trials, observational studies, editorials, case reports, case-series, and narrative reviews), duplicates and articles in a language different from Dutch or English or of which no free full text could be obtained.

Other reasons for exclusion were reviews that lacked relevant outcomes, focused on the effect of a specific treatment or specifically investigated the association with a pathology (such as malignancies or pneumonia). Reviews with irrelevant study populations (too young) or conditions (such as victims of avalanches) were also excluded (Fig. 1). Indistinct cases were discussed and resolved by consensus with RP, NVDN and PD.

Fig. 1
figure 1

Search and study selection: systematic reviews

Two systematic reviews were found using the search strategy mentioned above, concerning IHCA and OHCA, respectively [14, 15].

Primary study selection To find complementary primary studies published after these systematic reviews, a search was conducted by cross-referencing ‘elderly’, ‘resuscitation’, ‘functional outcome’ and ‘in-hospital’ or ‘out-of-hospital’, using appropriate medical subject headings (MeSH), keywords and filters in MEDLINE, EMBASE and PubMed to find observational studies and RCTs (Appendix 2 in ESM). The search period for the primary studies was determined by the end of the search period of the two included reviews: November 2012 to October 2018 concerning IHCA, and from May 2011 until October 2018 concerning OHCA. Older publication dates, unrelated articles, duplicates and articles in a language different from Dutch or English were excluded. Other reasons for exclusion were articles that focused on irrelevant outcomes, populations or contexts; comparisons of scores, guidelines, measurement tools or models, and the effect of a specific pathology or treatment on outcome (Figs. 2, 3). In total, 11 articles were included concerning IHCA [16,17,18,19,20,21,22,23,24,25,26] and 19 articles concerning OHCA [27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45].

Fig. 2
figure 2

Search and study selection: primary studies IHCA

Fig. 3
figure 3

Search and study selection: primary studies OHCA

Quality assessment

For quality assessment of the reviews, the AMSTAR measurement tool was applied. This checklist consists of 11 items. A review was considered to be of low, moderate or high quality when, respectively, ≤ 5, 6–8, or ≥ 9 of the requirements were fulfilled. Both reviews proved to be of moderate quality (Appendix 3 in ESM).

The STROBE checklist was used to assess the quality of the primary studies. A review was considered to be of low, moderate or high quality when, respectively, ≤ 10, 11–16, or ≥ 17 of the requirements were fulfilled. For the primary articles concerning IHCA and OHCA, all included studies proved to be of high quality after assessment (Appendix 4 in ESM). In 7 IHCA articles and 14 OHCA articles, Utstein-style guidelines were used to report the data.

Data extraction

Data extraction was performed by RZ and subsequently verified by RP, NVDN and PD. These data concerned: author, place and research period, study design, sample size, sample characteristics, study characteristics and outcomes of interest. We collected both unadjusted and adjusted data. Outcomes were chosen a priori, and consisted of: return of spontaneous circulation (ROSC), survival until hospital discharge, long-term survival (≥ 6 months following the event), discharge location and other quality of life measurements. These outcomes were structured and presented separately for the IHCA review (Table 1), the OHCA review (Table 2), the primary IHCA articles (Table 3) and the primary OHCA articles (Table 4), and additionally according to age (Table 5).

Table 1 Review in-hospital
Table 2 Primary studies in-hospital
Table 3 Review out-of-hospital
Table 4 Primary studies out-of-hospital
Table 5 Results categorized by age

Secondary variables concerning the condition and location of the arrest such as initial cardiac rhythm, type of ward, witnessed event, bystander-initiated CPR and emergency medical services (EMS) response time were reported when registered specifically for the older population. Risk factors positively or negatively associated with survival or quality of life outcomes were also documented.

Post-resuscitation neurological and functional state were mostly analyzed using the Cerebral Performance Category scale (CPC scale) [46]. This measurement tool uses a score of 1–5 for the categories: 1: “good cerebral performance”; 2: “moderate cerebral disability”; 3: “severe cerebral disability”; 4: “coma or vegetative state” and 5: “dead or brain death” [47]. In this review, scores of 1 or 2 are considered to be good neurological outcomes.

Results

Survival rates

The systematic review concerning IHCA consisted of 29 studies. An overall ROSC was found in 38.6% of the resuscitated patients. Overall survival rates were lowest in the category of 90 years and older (11.6%), followed by octogenarians (15.4%), and highest in patients aged 70–79 years (18.7%). Long-term survival (6 months to 1 year) varied between 5.7 and 20.9% concerning patients aged 70 years and older.

In the complementary IHCA studies, overall survival until hospital discharge was systematically higher and varied from 16 to 28.5% [28, 29]. Five studies registered an overall 1-month survival ranging from 13.5 to 28% [16, 18, 23, 24, 26]. Three studies found an overall 1-year survival between 9.4 and 25% [23, 24, 26].

The systematic review concerning OHCAs consisted of 23 studies. Four of these studies featured patients of 70 years and up exclusively, while five had a mean age of at least 70 years and the remaining studies provided various age categories. A meta-analysis for survival was performed on 14 studies. In patients aged 70 years or older, the overall survival until discharge was 4.1% (range 0–9%). The probability of survival significantly decreased as age increased, both in univariate and multivariate analyses.

Survival until hospital admission registered in five primary studies was 15.3–30.5% [31, 32, 34, 44, 45]. The survival until hospital discharge was registered in nine additional studies was also slightly higher than the previously published review and varied from 6.9 to 11.1% [32, 35,36,37, 42, 44, 45, 47, 50]. Twelve studies registered the 1-month survival rate, ranging from 3.8 to 7.8% of all patients [28, 29, 31,32,33,34,35,36, 38, 41, 43, 44]. Of those who survived until hospital discharge, 1-year survival reached 88% [45]. Two studies found a 3-year survival rate ranging from 6.8 to 52.8% [31, 36]. Survival up to 5 years was also registered in two studies and ranged from 2 to 32.1% [36, 39].

Neurological assessment

The results of both the IHCA and OHCA reviews on quality of life following CPR were scarce and contradictory [14, 15].

In the primary IHCA studies exploring CPC as an outcome, a favorable outcome (CPC 1–2) was found in a mean of 20.1–23.6% of all resuscitated patients [16, 23]. This proved to be 82.4–93.1% of the patients surviving until hospital discharge [16, 23, 26]. In the primary OHCA studies, a CPC 1–2 at discharge or at 1 month was registered in only 1–4.8% of all resuscitated patients [28, 29, 32, 34, 35]. On the other hand, one study saw that of all survivors until hospital discharge aged 90 years or older, a favorable outcome at 1 month (CPC 1 or 2) was found in all patients (100%) [43].

Discharge location

In the IHCA review, the discharge location was registered in only one of the included studies. Out of 50 patients who survived until hospital discharge, 38% were discharged to their homes, 24% were admitted to a rehabilitation or psychiatric facility, 18% to a nursing home and 20% were transferred to a chronic care hospital with ventilator capabilities [14]. The proportion of patients who returned home after hospital discharge was registered in three primary IHCA studies and three OHCA studies, ranging from 4.8 to 40% and 52.6–77.9%, respectively [17, 19, 26, 27, 31, 39]. Concerning one OHCA study, 88% returned to the same type of facility from which they originated [40].

Functional recovery

Furthermore, one study concerning IHCA found that although 74.9% of the survivors were functionally independent at hospital admission only 20.1% remained so after the event. In total, 63.4% of survivors were less functional upon hospital discharge compared to their state at the time of admission [17].

With regard to the OHCA cases, two studies used additional QoL measures such as the Overall Performance Category, SF-12 score, GOSE score, and the EQ-5D score [27, 39].The mean SF-12 mental component summary was 56.5 (SD 6.5). In contrast, the mean SF-12 physical component summary was 44.8 (SD 11.2), and this decreased with increasing age (p < 0.001). They found an upper and lower severe disability in 18.4% of the survivors, with the highest proportion in the age category 85 years and older (44.3%) [27].

Age categories

See Table 5.

Risk factors

Nineteen studies analyzed risk factors associated with short-term and long-term survivals following IHCA and OHCA [17,18,19,20,21,22,23, 26, 30,31,32, 36, 38,39,40,41, 43,44,45].

Patient-related factors associated with worse survival rates were increasing age, a previous history of heart failure or chronic obstructive pulmonary disease, malignancy, renal dysfunction, and multiorgan failure and/or septicemia after CPR. Pertaining OHCA, several studies found that increasing age was associated with worse survival outcomes, but a high Charlson Comorbidity Index (CCI ≥ 4) was not significantly associated with survival.

Arrest characteristics repeatedly associated with increased survival included: initial shockable rhythm (ventricular fibrillation and pulseless ventricular tachycardia), witnessed or ECG monitored arrests, cardiac etiology of arrest, monitored hospital location and time of arrest (during working hours on weekdays). The same risk factors were found concerning OHCA. In addition, arrests at public places were associated with increased survival rates, while arrests that occurred in nursing homes and multiple CPR events resulted in lower chances of survival.

Finally, rescue characteristics such as bystander CPR/defibrillation and shorter time to EMS response, defibrillation and ROSC were also associated with better survival outcomes following both IHCA and OHCA.

Eight studies analyzed risk factors associated with neurological and functional outcome following IHCA or OHCA. Increasing age was associated with worse functional outcomes, whereas initial shockable rhythm, cardiac etiology, public location of the arrest or witnessed arrest, and resuscitation factors such as bystander CPR, early EMS response and early use of automated external defibrillator were associated with a favorable neurological outcome [27, 28, 33, 34, 42, 45]. In addition, being functionally dependent before hospital admission and an admission diagnosis of trauma were associated with less optimal functional outcomes [17, 18].

Discussion

This scoping review presents an overview of the survival and quality of life of older patients following IHCA or OHCA. Data regarding the last decennium indicate a slight improvement in the survival until hospital discharge rates at ages 70 years and older following IHCA. Van Gijn et al. (including articles from 1968 to 2012) found an overall survival until hospital discharge of 18.7% for patients between 70 and 79 years old, 15.4% for patients between 80 and 89 years old, and 11.6% for patients of 90 years and older, whereas in more recent primary studies (including articles from 2012 to 2018), these rates were consistently higher (19–28%, 11.3–19% and 11–15%, respectively). The same trend was noted for OHCA with an overall survival rate of 4.1% in the systematic review from 1980 to 2011 versus a survival rate of 4.3–12% in the primary studies from 2011 to 2018, for patients aged 70 years or older.

Neurological outcome of resuscitated patients following IHCA or OHCA was only evaluated in a limited number of studies. Considering all resuscitated patients, the proportion that survived until hospital discharge with a favorable neurological outcome never exceeded 25%. However, considering the subgroup of patients surviving until hospital discharge or 30 days, the proportion with a good neurological outcome was eminently high, namely 82–100% regardless of the age.

Considerations

Possibly partially responsible for these improved survival rates is the increased initiation of CPR and/or defibrillation by bystanders out-of-hospital. Better management of complications and post-resuscitation care in-hospital, a specific task of the geriatrician, could also be an important factor contributing to the improved outcomes [48]. However, favorable results are more often seen following cardiac arrests with shockable rhythms which are only present in a minority of the older population. Consequently, overall survival rates remain low.

The effect of age on outcome remains controversial. While the included systematic reviews found that old age was associated with worsened survival outcomes for both IHCA and OHCA, other previously published studies could not find any association [49,50,51]. These inconsistent results may partially be explained by differences in patient characteristics, circumstances of the cardiac arrest, post-resuscitation care, cultural practices, do-not-attempt resuscitation policies, and withdrawal of life-sustaining treatment practice [52]. Moreover, the older population is particularly heterogeneous. The prevalence of multi-morbidity is increasing from 35 to 65% in patients aged 60–69 years to 80–99% in octogenarians [53]. In addition, the reduction of the physiological reserve of multiple organ systems with aging, called frailty, results in an increased inability to maintain homeostasis when faced with disease or injury. As such, not chronological age, but the level of frailty, which results in the reduced ability to withstand acute stress situations such as cardiac arrests, may be an important contributing factor for worse survival outcomes [3]. This was confirmed by a recent study published after this review [54]. These results may facilitate clinical decision making regarding whether CPR may be considered futile.

In conclusion, age alone should not be used as the sole criterium to decide whether CPR is a medically appropriate treatment for the older patient. Patient-related factors (age, medical history and frailty), arrest characteristics (initial rhythm, witnessed of monitored, etiology, location and time) and rescue characteristics (bystander CPR/defibrillation and shorter time to EMS response, defibrillation and ROSC) should also be taken into account.

Strengths and limitations

Strengths of this study include giving a comprehensive overview of the most recent available data on outcomes concerning survival, neurological state and quality of life of both in-hospital and out-of-hospital CPR in older patients, by systematically including return of spontaneous circulation (ROSC), survival until hospital discharge, long-term survival (≥ 1 year following the event), discharge location and other quality of life measurements as outcome variables.

This review has several limitations, due to the heterogeneity of classification in the individual studies we were not able to compare outcomes of specific age categories. The age categories were randomly selected, making comparison of outcomes between studies impractical. As a result, no meta-analysis was performed. However, no tools were used to measure the heterogeneity in this review. Furthermore, the pooled odds ratios concerning the risk factors were also not calculated in this study.

In addition, as a result of the inclusion of studies from various regions, different views and regulations concerning CPR performance have been adopted. For example, EMS providers in Japan are not permitted to terminate resuscitation in the field (except in case of decapitation or dissolution), and Do-Not-Resuscitate (DNR) orders are not generally accepted. Therefore, the generalisability of these study results is restricted.

Moreover, the term ‘frailty’ was only mentioned in one study concerning IHCAs [20]. Four OHCA studies mentioned frailty, but did not include this variable in their analysis. This would, however, allow healthcare providers to better differentiate older patients with poor prognosis from those with a good chance of survival. A multidimensional interdisciplinary diagnostic process is required to determine an older person’s medical, psychological and functional capability, such as a comprehensive geriatric assessment. The Clinical Frailty Scale is also a simple bedside assessment that can provide invaluable information when considering treatment escalation plans [54].

Finally, it is also possible that some eligible studies were missed because they did not specifically refer to older patients in the title or abstract. However, by performing an elaborate search and thorough cross-referencing, this risk was reduced to a minimum.

Implication for practice

The review provides an overview of the recent evidence on outcome regarding survival and quality of life after CPR following both IHCA and OHCA in the older population. CPR for patients of advanced age should be seen as a conditional therapy that may be worthwhile in some older patients, but may cause significant harm and suffering when applied in an undifferentiated way [55]. This knowledge can contribute to end-of-life conversations between health care practitioner and patient to make well-informed decisions and promote advance care planning (ACP). In hospitalized, frail or elderly patients, the etiology of the arrests is more likely to be complex due to other pathologies and complications [56]. These patients should be well informed about the overall low chance of survival and a significant possibility of neurological deficits, especially for nursing home residents. On the other hand, all patients with a limited risk profile should be informed of their chances of survival with a good neurological outcome following CPR [57].

Advance care planning in the hospital and in long-term-care facilities such as nursing homes can help avoid unwanted or ill-advised CPR. Results show that the implementation of ACP possibly decreases potentially inappropriate life-sustaining treatment, increases the use of hospice and palliative care and prevents hospitalization. It has also proven useful to include end-of-life choices such as Do-Not-Resuscitate decisions [55, 58].

When cardiac arrests happen outside the hospital, information about possible advance care planning is generally unavailable. Validated rules for prehospital termination of resuscitation (TOR) could be highly valuable in such cases to avoid futile transport and escalation of care. Until now, there is no international agreement on TOR rules neither on how to apply them. Verbeek et al. and Morrison LJ et al. propose the following criteria: arrest unwitnessed by EMS provider, no shock delivered, no prehospital ROSC, unwitnessed by bystander or no bystander CPR in an Advanced Life Support setting [59, 60]. Shibahashi et al. propose the following three criteria: non-shockable initial rhythm, unwitnessed by bystanders, and age ≥ 73 years; this TOR rule provided an excellent positive predictive value (> 99%) for unfavorable 1-month neurological outcome after OHCA [61]. These criteria were validated in North America by Grunau and Glober [62, 63].

Future research

None of the included studies was able to capture a good image of the patients’ quality of life broader than the neurological outcome. Most of the studies merely used the CPC scale which is insufficiently attuned to identifying all the impairments of cardiac arrest survivors [46]. In addition, this score excludes return to private homes, participation in society or other aspects of life greatly valued by this category of patients such as the desire to avoid being a burden to their family, suffering pain, the loss of speech, the loss of dignity and the incapacity to think clearly [64]. Only two studies incorporated multiple measurement tools to assess the quality of life, namely the Extended Glasgow Outcome Score (or GOSE), the Overall Performance Category (OPC), the SF-12 score, and the EQ-5D score [27, 29]. More extensive research into quality of life after resuscitation will allow patients to make better-informed decisions and facilitate decision making adopted to the patient’s clinical situation.

A core outcome set for cardiac arrest (COSCA) was recently developed for adults that identified survival, neurological function, and health-related quality of life as essential outcomes in cardiac arrest effectiveness trials. They stated that survival until hospital discharge, at 30 days, or both should be reported, accompanied by neurological state. Health-related quality of life should be measured with ≥ 1 tools from the Health Utilities Index-3 questionnaires, Short Form Health survey 36 version 2 or the five-level EQ-5D instruments at 90 days and at periodic intervals up to 1 year after cardiac arrest, if resources allow [65, 66]. Further research should implement this assessment method and determine if it is also optimal for evaluation of the older population. Prospective study designs are preferred, because many potentially relevant prearrest factors or outcome measures cannot be retrieved retrospectively.

Conclusion

Hospital survival rates following IHCA and OHCA in the older population improved in the recent decade, though do not exceed 28.5% and 11.1%, respectively. Several risk factors were identified, among which increasing age and nursing home residency. However, the effect of age on outcome remains controversial and age should not be used as the sole decision criterium whether to initiate CPR. Future research should analyze frailty as an independent variable regardless of age and include more extensive quality of life measures as outcome variables.