Frailty Status Predicts Futility of Cardiopulmonary Resuscitation in Older Adults

Sarah E. Ibitoye; Sadie Rawlinson; Andrew Cavanagh; Victoria Phillips; David J.H. Shipway


Age Ageing. 2021;50(1):147-152. 

In This Article

Abstract and Introduction


Aim: To determine if frailty is associated with poor outcome following in-hospital cardiac arrest; to find if there is a "frailty threshold" beyond which cardiopulmonary resuscitation (CPR) becomes futile.

Methods: Retrospective review of patients aged over 60 years who received CPR between May 2017 and December 2018, in a tertiary referral hospital, which does not provide primary coronary revascularisation. Clinical Frailty Scale (CFS) and Charlson Comorbidity Index were retrospectively assigned.

Results: Data for 90 patients were analysed, the median age was 77 (IQR 70–83); 71% were male; 44% were frail (CFS > 4). Frailty was predictive of in-hospital mortality independent of age, comorbidity and cardiac arrest rhythm (OR 2.789 95% CI 1.145–6.795). No frail patients (CFS > 4) survived to hospital discharge, regardless of cardiac arrest rhythm, whilst 13 (26%) of the non-frail (CFS ≤ 4) patients survived to hospital discharge. Of the 13 survivors (Age 72; range 61–86), 12 were alive at 1 year and had a good neurological outcome, the outcome for the remaining patient was unknown.

Conclusion: Frail patients are unlikely to survive to hospital discharge following in-hospital cardiac arrest, these results may facilitate clinical decision making regarding whether CPR may be considered futile. The Clinical Frailty Scale is a simple bedside assessment that can provide invaluable information when considering treatment escalation plans, as it becomes more widespread, larger scale observations using prospective assessments of frailty may become feasible.


Survival to discharge following in-hospital cardiac arrest is reported to be as high as 17–20%.[1–3] However, this is unlikely to be representative of outcomes in older people, where myriad studies have demonstrated that patients of advanced age[3] or comorbidity[4,5] have significantly reduced survival following cardiac arrest. In recent months, frailty has also been linked to adverse outcomes following in-hospital cardiac arrest.[6,7]

Frailty is considered to be a syndrome of impaired physical function and reduced physiological reserve; it is associated with adverse outcomes including mortality, physical dependence and hospitalisation.[8,9] In the frail, major surgery is associated with higher rates of complications and prolonged length-of-stay;[10] furthermore, admission to intensive care is linked to increased mortality in patients with established frailty.[11] Various scales and tools exist to identify frailty, and there is no universal consensus on which tool is superior. However, the Clinical Frailty Scale (CFS) is increasingly used in UK clinical practice.[12] It is easy to apply at the bedside and considers an individual's pre-existing level of function and mobility in a 9-point visual and descriptive scale (Figure 1).

Figure 1.

The CFS.

Cardiopulmonary resuscitation (CPR) was originally developed to treat cardiac arrest after myocardial infarction;[13] it is now the default position for all in-patient cardiac arrest unless a 'Do Not Attempt CPR' (DNACPR) decision has been documented. Failure to discuss and document a DNACPR decision will, therefore, typically result in CPR being attempted on patients sustaining cardiac arrest, even if such patients are naturally dying from advanced or irreversible disease. In such cases, CPR is associated with low rates of success and the application of inappropriate and futile CPR treatment can deprive patients of a dignified and peaceful death at the end of their natural lives.[14] CPR, in this context, can therefore arguably be considered harmful. Furthermore, CPR is considered to be an aerosol-generating procedure and in the era of the 2020 global COVID-19 pandemic, the risks of CPR to healthcare staff are not insignificant.[15] It is therefore important that CPR should only be undertaken when there is a chance of successful outcome.

In the UK, the General Medical Council (GMC) is the independent regulator of doctors and publishes guidance on various subjects including CPR. On this topic, GMC guidance indicates that when patients are at foreseeable risk of cardiac arrest, a judgement about the likely benefits, burdens and risks of CPR should be made as early as possible.[16] Discussions about CPR should be had with patients, or if they lack capacity with their relatives or caregivers. However, the evidential basis for DNACPR is sometimes inadequately understood by clinicians who can struggle to determine when CPR may be considered futile, and therefore should not be offered to the patient.[17] Legal cases pertaining to DNACPR orders have achieved high-profile coverage in the mainstream media and have generated anxiety surrounding the issue of DNACPR.[18] As a result of these issues, some doctors feel increasingly dis-incentivised to address the challenging issue of DNACPR.[19] The evidence currently indicates that CPR is attempted in many patients in whom successful resuscitation is highly unlikely, and therefore possibly inappropriate.[20]

A number of scoring tools have been proposed to predict survival after in-hospital cardiac arrest. These include the pre-arrest morbidity score,[21] and subsequent modifications.[22,23] These scores include multiple clinical characteristics that are weighted according to importance to generate a final score. Although theoretically feasible in routine clinical practice, calculation at the bedside can be cumbersome. Furthermore, some components are rapidly dynamic and related to acute illness. Whilst it is clear that they are of adverse prognostic value in an acute illness, these parameters are easily reversible and therefore of limited value in making long-term resuscitation decisions. Although these scores have an acceptable level of accuracy in predicting outcomes of CPR, their complexity makes them difficult to use in real-life clinical practice. This is reflected in the paucity of their use in clinical decision-making despite being formulated >20 years ago.[24]

In clinical practice, we have observed that frailty is often cited as an indicator that resuscitation is unlikely to be successful and documented as a justification that resuscitation should not be attempted. However, at the time of commencing this study, there was no published research in UK populations describing the association between frailty and mortality following in-hospital cardiac arrest.

We aimed to determine whether frailty is associated with in-hospital mortality following cardiac arrest in older patients (aged >60 years), independently of age and comorbidity. To support clinical decision-making, we considered whether a 'frailty threshold' exists at which CPR is likely to be futile. We hypothesised that such a threshold might inform clinicians of when CPR is unlikely to change the outcome, and therefore at what point it should not be offered to patients.