Can the Clinical Frailty Scale predict futility in out-of-hospital cardiac arrest?

Cardiopulmonary resuscitation (CPR) is a critical intervention in the management of cardiac arrest and early intervention is associated with improved survival. Multiple studies have demonstrated the critical importance of bystander CPR in extending the time frame during which defibrillation is likely to be effective in those presenting in shockable rhythms.

However, there is significant heterogeneity among individuals who experience a cardiac arrest, and the chance of survival depends upon a multitude of factors including the aetiology of the cardiac arrest, the initial presenting rhythm, initiation of bystander CPR and access to early defibrillation (Perkins et al, 2015). Furthermore, the duration of time in cardiac arrest, the patient's age and the presence of comorbidities are factors considered in prognostication (Wissenberg et al, 2015; Goto et al, 2016; Hirlekar et al, 2018). Despite this, in emergency medical services (EMS) systems around the world, CPR continues to be undertaken irrespective of prognosis (Druwé et al, 2020).

This would not be an issue if CPR were a benign procedure; however, the intervention is associated with physical injuries to the patient (Yamaguchi et al, 2017) and psychological trauma for bystanders (Meier et al, 2016; Cole et al, 2021; Cokljat et al, 2022). Even if a resuscitation attempt is deemed successful and a return of spontaneous circulation (ROSC) is achieved, the burden of potential disability can be huge and lead to significant physical disability requiring long-term care (Mongardon et al, 2011; Ebell et al, 2013; Yamamoto et al, 2023). A joint statement from the British Medical Association, the Resuscitation Council (UK) and the Royal College of Nursing (British Medical Association et al, 2016) as well as the General Medical Council (2023) acknowledge that the very nature of CPR could deprive an individual of a respectful and dignified death.

Defining a threshold for when a treatment (e.g. CPR) may be futile is difficult. Traditionally, this was when it was likely to be <1% successful (Schneiderman et al, 1990) but this has been criticised for not acknowledging the harms that treatment may incur (Ardagh, 2000; Rosoff and Schneiderman, 2017; Grunau et al, 2019; Druwé et al, 2020).

Given that CPR may be of fundamental importance to survival or cause significant harm, it is not surprising that clinical decision-making regarding starting of CPR is extremely challenging. Such decision-making involves clinical, legal and ethical aspects.

Various authors have argued that paramedics working in prehospital settings are exempt from making such decisions because of the paucity of information available (Georgiou and Georgiou, 2019). Algorithms have been developed to remove complex decision-making from paramedics; however, guidelines have frequently become dogmatic in their application despite the nuances of individual incidents. With 86% of paramedics having reluctantly performed CPR (Moffat et al, 2019), it can be argued that paramedics are not only sensitive to identifying patients who will not benefit but may experience moral injury if compelled to carry out CPR against their clinical judgement. This is now recognised in clinical guidelines (Association of Ambulance Chief Executives (AACE), 2019).

Current guidelines set out by the Joint Royal Colleges Ambulance Liaison Committee (JRCALC), state that although resuscitation efforts must be made whenever there is a chance of survival, CPR is not supported for patients who have no chance of survival (AACE, 2019). They note that patients in the final stages of an advanced and irreversible condition will not benefit from resuscitation. This allows far greater scope of practice to UK-based ambulance crews than in other countries, such as the United States, where many states require ambulance crews to attempt resuscitation in all cases where a do not attempt resuscitation document is not present until physician advice can be sought (Cole et al, 2021).

Despite this greater autonomy, estimates suggest that 11 000 patients receiving futile resuscitation are transported to hospital every year (Cokljat et al, 2022). Furthermore, since the majority of resuscitation attempts are terminated in the field, there may be as many as 40,000 additional patients receiving unethical care at the end of life (Cokljat et al, 2022). In addition to denying an individual a dignified death, this has an impact on the workload of emergency departments and has an economic impact on the wider healthcare system.

Several authors have investigated what the barriers are that prevent paramedics from implementing these guidelines. It has been suggested that UK paramedics lack training around end-of-life decision-making and the ethics of CPR, particularly as most of their training prompts an interventionist approach (Moffat et al, 2019). The authors of the present study have also found that clinicians would knowingly perform futile resuscitation to avoid conflict with families and because they feared disciplinary action if they did not. These trends are not isolated to the UK, and several authors have criticised training that omits prognostication and whose algorithmic approach does not allow for clinical nuance (Druwé et al, 2020).

Bashiri et al (2019) noted how clinical decisions in the prehospital setting can be complex and are often characterised by uncertainty. They found the use of clinical decision-making tools in such scenarios improved triage, the quality of care and diagnostic accuracy.

To help mitigate the barriers to appropriate clinical decision-making, the authors recommend the development of a tool that ambulance clinicians can apply in the out-of-hospital cardiac arrest (OHCA) setting which can provide a quantitative assessment of the patient's chance of survival.

First published in 2005, and updated in 2020, the Clinical Frailty Scale (CFS) is a nine-point linear tool based on clinical judgement that allows healthcare professionals to grade a person's level of frailty (Rockwood et al, 2005; Rockwood and Theou, 2020). Frailty is defined as an interconnection between age and accumulated health deficits that can be used to predict vulnerability to adverse health outcomes such as mortality.

From its inception, the CFS was designed to be easy to use in clinical settings and its simplicity and rapid applicability mean extended assessments, calculations or testing are avoided (Stille et al, 2020). The tool is inherently objective and has the additional benefits of not requiring access to electronic health records (Akpan et al, 2018). Clinicians have reported it is easy to use without specialist education and training, and note it is a viable alternative when more comprehensive testing is not readily available (Young and Smithard, 2020; Hamlyn et al, 2022).

Aim

The primary aims of this rapid review of the literature are to determine: whether the CFS could be used in the prognostication of cardiac arrest; and, secondarily, whether the tool could be adapted for use by UK paramedics in the context of OHCA.

Methods

This rapid review of the literature followed the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines (Page et al, 2021) and MEDLINE (PubMed), Embase (ScienceDirect) and the Cochrane Central Register of Controlled Trials (Cochrane Library) were searched. A date range of August 2005 to May 2023 was chosen as August 2005 was when the CFS was first published (Rockwood et al, 2005).

Because test searches gave limited and inconsistent results, the search terms were kept intentionally broad, consisting only of ‘frailty’ and ‘cardiac arrest’. Reference lists from studies assessed for eligibility were screened to identify any research missed in the database searches.

Inclusion criteria were kept broad to ensure the maximum scrutiny of available evidence. Only two factors were required for inclusion: did the sample experience a cardiac arrest, and was a CFS recorded? Articles were then excluded if they did not record mortality or were not available in the English language.

After removing duplicates, one author (MA) independently screened titles and abstracts for relevance before assessing the full text for eligibility. All remaining articles were included for analysis regardless of methodological limitations. The author was then able to extract data for analysis from the published articles and supplementary data, except in the case of one article (Ibitoye et al, 2021), where he contacted a researcher who was able to provide this information.

In total, 63 papers were identified. Following exclusions, two systematic reviews and meta-analysis (Mowbray et al, 2021; Hamlyn et al, 2022) and six primary research articles remained. Five of these (Wharton et al, 2019; Fernando et al, 2020; Ibitoye et al, 2021; Thomas et al, 2021; Hu et al, 2022) were included for final review; their characteristics are shown in Table 1. Because of heterogeneity issues with the data, the two articles with pooled data and one additional study identified (Xu et al, 2020) could not be included in the final analysis. The systematic reviews and meta-analysis were not included in this analysis; however they are included in the discussion nonetheless. The PRISMA literature selection process is shown in Figure 1.

Results

The five studies used the CFS as a measure of frailty (Table 1). One study defined frailty as CFS ≥4 (Ibitoye et al, 2021), Two studies defined frailty as CFS ≥5 (Fernando et al, 2020; Thomas et al, 2021), and one study defined frailty as CFS ≥6 (Wharton et al, 2019). Hu et al (2022) categorised CFS scores of 4, 5 and 6 as very mildly, mildly or moderately frail respectively. For pooling data, CFS>5 (mildly frail) was used to define frailty from the Hu et al (2022) study.

In the present study, the pooled random effects model indicated that frail patients had a higher in-hospital mortality than those who were not frail (RR 1.61; 95% CI (1.17–2.22); P<0.01) (Figure 2).

Discussion

Frailty is associated with poorer outcomes for in-hospital cardiac arrest

A multitude of factors are associated with worse outcomes in cardiac arrest, and these include but are not limited to aetiology, presenting rhythm, age, sex, duration of resuscitation attempt and the presence of comorbidities. This rapid review of the literature, however, has sought to ascertain whether a higher CFS result is associated with worse outcomes in the context of cardiac arrest, and it is apparent that frailty is strongly associated with increased mortality.

One of the largest systematic reviews and meta-analyses on the topic, which included a cohort of 1134 individuals who had experienced an in-hospital cardiac arrest, concluded that frail individuals had 37% lower odds of achieving a ROSC than non-frail individuals (Mowbray et al, 2021). Another meta-analysis comprising seven studies with a total patient sample of 1704 also concluded that frailty was strongly correlated with an increased likelihood of mortality following CPR (Hamlyn et al, 2022).

These findings are not unique; another study, published in 2019, assessed 179 in-hospital cardiac arrests for which CFS could be calculated and concluded that, while 32.1% of individuals with a CFS score of 6–9 achieved ROSC, only 1.8% (1/56) of these survived to hospital discharge (Wharton et al, 2019). Further studies have corroborated this, and a 2021 study analysed data for 90 patients and concluded that frailty was a strong predictor of in-hospital mortality. In this study, no patient with a CFS >4 survived (Ibitoye et al, 2021).

Although frailty is typically defined with a CFS ≥5, it is at ≥6 that the evidence shows a dramatic drop in survival. It is therefore the authors' prediction that this is where the threshold for futility will be found. Only one study contradicts this hypothesis, in which a 16% (8/50) survival rate in those with a CFS ≥7 was found (Hu et al, 2022). This was the only study that showed high survival rates among those with severe frailty and its authors acknowledged this disparity within the existing literature, suggesting it is a result of superior palliative care within their system.

It should also be noted that ROSC or even survival to hospital discharge are arbitrary endpoints and do not consider quality of life. Indeed, Hu et al's (2022) anomalous study recorded eight patients who survived to 30 days with a CFS ≥7, a marker characterised by a complete dependence on personal care, a factor many older adults cite as significantly reducing their quality of life (van Leeuwen et al, 2019). None of the studies included in this review gave any insight into patients' perspectives on their quality of life.

One study, which included a retrospective analysis of 113 patient records (out of which use of CFS was determined in 89), looked beyond hospital discharge and concluded that no patient with a CFS ≥6 survived to 1 year (Thomas et al, 2021).

When limited to the UK, research suggests the survival rate for patients who experience an in-hospital cardiac arrest (IHCA) with a CFS ≥6 to be 1% (1/97) (Wharton et al, 2019; Ibitoye et al, 2021; Thomas et al, 2021). With the average UK OHCA patient 61.8% less likely to survive to discharge than their IHCA counterpart (Perkins et al, 2021), the number of survivors is expected to be negligible. While survival with an acceptable quality of life will always remain a possibility in extremely rare cases, prehospital clinicians must decide what harms they are willing to inflict to achieve this.

Furthermore, as the life expectancy of an individual with a CFS of 6 is expected to exceed 3 years, these findings suggest the JRCALC (AACE, 2019) threshold of a patient needing to be in the final stages of an irreversible illness to be considered unsuitable for resuscitation attempts is not sensitive enough to identify patients who will not benefit.

Can paramedics use the Clinical Frailty Scale?

If the CFS is to be used to predict futility in OHCA, paramedics must be able to:

• Assign scores to patients accurately while attempting resuscitative efforts
• Understand how other factors may influence the relationship between frailty and outcomes
• Discuss their clinical decisions with family or caregivers in a sensitive manner. Whether this is within the capabilities of paramedics must be explored.

Because of the retrospective nature of the research reviewed, no studies could provide evidence on whether the CFS is suitable for paramedic use in the field. Although Fernando et al (2020), Thomas et al (2021), Ibitoye et al (2021) and Hu et al (2022) detailed their methods, they did not explicitly state what clinical grade or specialist training their assessors had. Only Wharton et al (2019) provided this information, noting CFS was assigned by a single senior geriatrician with an interest in frailty scoring.

Despite the lack of evidence validating CFS for paramedic practice, Hu et al (2022) provided details of an approach for assessment that simplifies the scoring process. They had two geriatricians in their clinical practice who were experienced in calculating CFS develop an algorithm (Figure 3) with feedback from one of the researchers who first developed and validated the CFS. They then confirmed validity of their algorithm's use among the research team against geriatrician-assigned CFS from in-person assessment, and the reliability of scores between researchers was calculated.

At the extremes of Hu et al's (2022) algorithm, knowledge of a patient's life expectancy, exercise tolerance and experience of symptoms is required. It may be unlikely that paramedics could gather these details from lay persons on scene. However, if the threshold for futility is found to be between a CFS of 5 and 6 as the authors predict, the paramedic will need to establish only if the patient is fully dependent on instrumental activities of daily living or partially dependent on basic activities of daily living (Figure 4). Simple questions to family or caregivers such as, ‘Does the patient have any carers?’, ‘Do they require assistance with washing or dressing?’, ‘Are they still able to get out of the house by themselves?’ may be enough to assign a binary score (whether the score is <6 or ≥6).

Whether bystanders at an OHCA would be able to provide this information and whether it can then be appropriately interpreted by paramedics will require further research. It was the view of Goldstein et al (2012), however, that the paramedics' unique position of being present in the patient's living conditions would enhance frailty assessment rather than hinder it.

As to whether paramedics are able to calculate, interpret, apply and communicate the CFS appropriately, the Health and Care Professions Council (2023) standards of proficiency for paramedics state that they should be able to use evidence-based practice to adapt and modify treatment approaches to meet individual patient needs and communicate this in a way to support informed clinical decision-making. If the CFS is validated for use, the authors' proposed use of it should fall within the UK paramedic scope of practice, similar to any other approved clinical decision tool.

Limitations

This rapid review of the literature has some limitations. There is a paucity of data relating to the prehospital setting and the studies included in this review all discuss in-hospital cardiac arrest. Extrapolating such conclusions to the context of OHCA has not been possible.

Although several studies were found that measured frailty in cases of OHCA (Søholm et al, 2014; Sulzgruber et al, 2017; Cokljat et al, 2022; Mowbray et al, 2023; Yamamoto et al, 2023), they either did not use the CFS or were not comparing CFS to mortality so were not included in this review.

Conclusion

This rapid review of the literature has summarised the existing evidence around how frailty impacts upon futility in cardiac arrest.

Although this evidence is limited, there is a consensus that as frailty increases, a patient's chance of survival decreases. Although the body of evidence is not sufficient to change practice, it has highlighted significant limitations in current guidelines. Further research that assesses OHCA outcomes against retrospectively calculated clinical frailty scores could be used as the basis for the development of a futile resuscitation clinical decision-making tool, which might include a component based upon the CFS.

If paramedics are able to use such a tool, this could lead to a reduction in relatives being exposed to traumatic interventions, a decrease in moral injury for paramedics and a subset of patients being spared inappropriate resuscitation attempts and afforded a more dignified death.

Key Points

• Although CPR is a crucial intervention in surviving OHCA, it is performed on many who would not benefit resulting in a traumatic, undignified, and sometimes prolonged death
• Lack of training, fear of conflict, and fear of disciplinary action prevent pre-hospital clinicians from exercising their clinical judgement. A clinical-decision-making tool may help overcome this
• Frailty as measured by the CFS appears to be a good predictor of mortality of IHCA and may be more sensitive than current JRCALC guidelines
• The authors predict a threshold for futility can be drawn at a CFS of ≥6
• A validated algorithm exists to simplify the CFS scoring process
• Future research must focus on whether these findings can be applied to the OHCA setting

CPD Reflection Questions

• What factors do you feel are most important in deciding when resuscitation would be appropriate?
• What do you think should be the measure of a successful resuscitation?
• Do you feel the results of this review will be similar in the out-of-hospital cardiac arrest setting?