Background
The population admitted to hospital following myocardial infarction is ageing. Advances in patient care have reduced age-specific mortality rates in developed countries, but this effect is offset by an expanding older population [
1,
2]. However, clinical trials in acute coronary syndrome have consistently failed to represent older adults, limiting the generalizability of findings to this age group [
3]. The Global Registry of Acute Coronary Events (GRACE) sought to provide a larger and more representative sample, demonstrating significant disparities in the management and in-hospital outcomes for the oldest patients with myocardial infarction [
4,
5]. These data have generated GRACE risk estimates which include age amongst other clinical parameters to predict outcomes following myocardial infarction [
6,
7].
However, it is recognized increasingly that frailty, as a metric of depleted physiological reserves, better reflects biological age in older adults [
8]. Frailty is three-fold more common in older people with cardiovascular disease [
9] and these individuals experience double the mortality risk of fitter people independent of age or comorbidity [
10]. For risk prediction, gait speed has been shown to add value to Framingham risk scores of patients with ST-segment elevation myocardial infarction [
11]. Similarly in patients with non-ST-segment elevation infarctions, the physical frailty phenotype independently predicted major cardiovascular events beyond the GRACE score in the TRILOGY ACS randomised control trial comparing antiplatelet strategies [
12]. Similar reports of the effect of frailty on GRACE risk estimates come from studies using physical measures that necessitate additional patient testing [
13].
The Clinical Frailty Scale (CFS) is a brief guided tool to assess frailty in hospital settings without specialist equipment [
14]. It has been widely used to identify older patients at risk of poorer outcomes [
15]. Although it does not require additional equipment or physical measures, the CFS is a precise frailty tool, using specific descriptors of patient symptoms and activity. Our objective was to test the performance of the CFS in an older population with myocardial infarction. We hypothesised that addition of this simple frailty measure would improve the prognostic properties of the GRACE score.
Discussion
We have studied the utility of a simple frailty measure for risk prediction in older patients after myocardial infarction and have made a number of important observations. First, the GRACE tool overestimated 12-month mortality risk in our study population. Second, a guided frailty tool could be incorporated into routine clinical care to identify patients at high risk of poor outcomes. Third, the Clinical Frailty Scale was an independent predictor of mortality in an older myocardial infarction population and added significant discrimation to the GRACE 12-month mortality estimate. Finally, frailty reclassified mortality risk in nearly half of this population, largely through identification of older but fitter individuals. This simple measure of frailty has potential to improve risk assessment in the large population of older patients recovering from myocardial infarction.
There are a number of strengths to our study. Frailty assessment was performed by clinical nursing staff using existing clinical data and without any specialist equipment. This approach would appear feasible for inclusion into the routine care of older cardiology patients based on our experience in a busy tertiary referral centre. In contrast to other studies [
22,
23], we have not focussed solely on short-term risk estimation and have directly assessed frailty against an existing and widely used clinical tool. Further, we have undertaken external validation suggesting wider applicability of these findings, although this would benefit from larger validation studies. With advances in cardiac care, the majority of even the oldest patients survive an acute infarct [
24]. The attention of risk prediction after myocardial infarction has therefore shifted from immediate survival to recovery, rehabilitation and future risk-stratification. Identification of vulnerability to poor outcomes could guide tailored cardiac rehabilitation, follow-up and future intervention strategies.
It may be surprising that most reclassification gains occurred by downgrading the estimated GRACE risk in robust patients. This challenges preconceptions that frailty assessment only adds value in those nearing death; resilience in fitter older patients may be equally informative. However, at the other end of this spectrum, it is striking that half of the patients identified as frail by CFS assessment had died within 12 months of myocardial infarction. Frailty assessment offers the potential for future care planning in this targeted population, which we have previously shown to be feasible and acceptable in our phase II study [
16]. Individualized decision-making including frailty could therefore increase clinician confidence in the management of patients across the range of vulnerability from low to high risk. Such an approach is the antithesis of ageism, and may assist in targeting increasingly complex interventions towards to those most likely to experience benefit [
25].
Our findings may reflect excessive reliance on age in current risk determination. At a population level there is no doubt that ageing increases the risk of almost all harmful outcomes, but chronological age fails to capture individual differences. At its core, frailty exposes the variation in ageing trajectories between individuals [
8]. In this study, age was not independently predictive of any outcome once frailty was included in modelling. It is however important to recognize that the GRACE estimation, which includes age as a key variable, peformed well in the discrimination of 12-month mortality in the study population, but worth acknowledging that frailty could improve this further and provide net reclassification benefit. Such improvements in the discrimination of existing risk scores are rarely achieved by new biomarkers [
26] and this finding merits further evaluation in larger cohorts, or perhaps in future iterations of the GRACE tool.
The 2015 European Society of Cardiology guidelines recommend use of the GRACE 2.0 calculator to assess patient risk after myocardial infarction, stating that the value of such tools is “undisputed”. However, the guidelines acknowledge that “the impact of risk score implementation on patient outcomes has not been adequately investigated” [
27]. The complexities of managing frail patients are recognized, particularly with regard to invasive strategies and adjusted dose regimes for antiplatelets, beta-blockers and ACE-inhibitors. In our study no frail patients attended cardiac rehabilitation, in keeping with referral patterns observed elsewhere [
28]. This is despite increasing evidence in favour of structured physical activity programmes amongst individuals living with frailty [
29]. The European Association of Preventative Cardiology have recently identified the pressing clinical need for further research into the area of frailty and cardiac rehabilitation [
30]. No specific frailty tool is recommended by these guidelines, which in part reflects a lack of consensus in the broader frailty literature. A recent systematic review identified reports of 67 different frailty tools, of which the CFS is one of the most highly cited [
15].
Other recent guidelines for the management for older patients have focussed on multimorbidity, such as from the National Institute for Health and Care Excellence (NICE) [
31]. However, the Charlson comorbidity index did not add to the GRACE estimation in our study, suggesting that simplistic counts of comorbidity may be less important than assessing the functional manifestations of these conditions. We have demonstrated that the Karnofsky performance scale was an independent predictor of mortality beyond GRACE, although discrimination was not as strong as with the CFS. It is possible that such functional and disability scales may add further objectivity to the classification of frailty.
Our study has some limitations. The study population was recruited for potential selection into a study of future care planning. An age cutoff was chosen to enrich for the frailty measure, but this limited the identification of younger patients with impairments. The study cohort was therefore likely to have been at higher risk than a general, older cardiology population. Despite this, only 20% of our study population were graded with a CFS score ≥ 5, which represents a realistic proportion for additional intervention such as future care planning. The CFS did not independently predict hospital readmission, but this may be related to competing mortality risk amongst frail patients. It is critical that an effective frailty measure does not saturate in the target population, as this would lack any clinical utility beyond age. The distribution of CFS scores across our study was significantly less frail than in the Canadian older community-dwelling population in which it was first described [
14]. This may be due to a younger but more medically comorbid sample hospitalized with myocardial infarction, but may also reflect our limited sample size. Although we undertook external validation this was also performed in a small cohort.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.