Frailty Screening in Critical Care at Scale

Abstract

FOR RELATED ARTICLE, SEE PAGE 1292The concept of frailty as a determinant of outcome from critical illness has resonated hugely with the critical care community since its description a decade ago.1McDermid R.C. Stelfox H.T. Bagshaw S.M. Frailty in the critically ill: a novel concept.Crit Care. 2011; 15: 301Crossref PubMed Scopus (121) Google Scholar Having long appreciated the importance of “physiologic” as opposed to “chronologic” age, frailty is now widely understood to represent a state of diminished physiologic reserve more prevalent with age, which is associated with increased vulnerability towards adverse outcomes, but which is distinct from comorbidity.2Fried L.P. Ferrucci L. Darer J. Williamson J.D. Anderson G. Untangling the concepts of disability, frailty, and comorbidity: implications for improved targeting and care.J Gerontol A Biol Sci Med Sci. 2004; 59: 255-263Crossref PubMed Google Scholar The Clinical Frailty Scale (CFS) lends itself to screening for frailty in the critical care setting; it has “face validity,” combines clinical judgment with objective measurement, can be applied readily without adaptation, and has demonstrable reliability3Pugh R.J. Battle C.E. Thorpe C. et al.Reliability of frailty assessment in the critically ill: a multicentre prospective observational study.Anaesthesia. 2019; 74: 758-764Crossref PubMed Scopus (33) Google Scholar and predictive validity.4Muscedere J. Waters B. Varambally A. et al.The impact of frailty on intensive care unit outcomes: a systematic review and meta-analysis.Intensive Care Med. 2017; 43: 1105-1122Crossref PubMed Scopus (351) Google Scholar Given global demographic trends and potential critical care resource requirements of an ageing population, the added value of frailty assessment to clinical discussions and decision-making and to risk-adjusted outcome reporting have been highlighted previously.5De Biasio J.C. Mittel A.M. Mueller A.L. Ferrante L.E. Kim D.H. Shaefi S. Frailty in critical care medicine: a review.Anesth Analg. 2020; 130: 1462-1473Crossref PubMed Scopus (31) Google Scholar However, until now there has been no published research regarding the feasibility and prognostic value of population-scale screening for frailty in critically ill patients.In this issue of CHEST, the study by Darvall et al6Darvall J.N. Bellomo R. Paul E. et al.Routine frailty screening in critical illness- a population-based cohort study in Australia and New Zealand.Chest. 2021; 160: 1292-1303Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar is therefore an important addition to the literature. In a well-designed study, they report successful implementation of routine frailty screening using the CFS across diverse ICUs in Australia and New Zealand. Frailty was common in this cohort, approaching one in five patients when assessed at the time of admission. Consistent with findings in other studies, increasing frailty was associated with higher hospital death. Furthermore, addition of CFS score to the APACHE III-j risk prediction model improved model performance as measured by discriminant function (area under the receiver operating characteristics curve).Importantly, the authors6Darvall J.N. Bellomo R. Paul E. et al.Routine frailty screening in critical illness- a population-based cohort study in Australia and New Zealand.Chest. 2021; 160: 1292-1303Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar looked beyond mortality rates and investigated whether frailty was associated with a number of additional person-centered outcomes. The presence of frailty increased length of ICU and hospital stay and increased the likelihood of discharge to a nursing home or chronic care facility. Furthermore, patients with frailty were more likely to experience delirium and pressure injuries within ICU. These findings provide a more complete picture of the likely consequences of ICU admission in the context of frailty to inform clinicians and patients.A significant strength of the study was the extremely large sample size (n = 234,568). The authors6Darvall J.N. Bellomo R. Paul E. et al.Routine frailty screening in critical illness- a population-based cohort study in Australia and New Zealand.Chest. 2021; 160: 1292-1303Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar were able to conduct analyses using the Australian and New Zealand Intensive Care Society Adult Patient Database, a bi-national audit database. This allowed associations to be reported with precision for each category of the CFS, rather than the dichotomization that other studies have needed to undertake. Furthermore, this larger sample size allowed subgroup analyses to be undertaken in a younger cohort of patients, which confirmed that the presence of frailty was associated with worse outcomes.An important limitation was the proportion of missing data relating to CFS recording. With one in three observations missing a value for the primary exposure, the potential for bias could have been substantial. Ideally, evaluating the impact of missing data requires an understanding of the mechanism by which missing data have arisen, along with sensitivity analyses to impute missing values if the mechanism of missingness is random.7Sterne J.A. White I.R. Carlin J.B. et al.Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls.BMJ. 2009; 338: b2393Crossref PubMed Scopus (3823) Google Scholar However, the authors undertook a number of additional analyses that were reassuring and that indicated that the likelihood of significant bias due to missing data was low. This included a comparison of baseline characteristics in those with and without missing data and subgroup analyses on ICUs with low levels of missing data.Training of assessors was not mandated, and the accuracy of CFS recording was not evaluated. Although the authors have suggested that lack of individualized training was not a significant barrier to implementation, this may have contributed to measurement error. As long as such error is not differential (not systematically biased towards over- or under-estimating CFS), this would tend to reduce the strength of association between exposure and outcome.8Hutcheon J.A. Chiolero A. Hanley J.A. Random measurement error and regression dilution bias.BMJ. 2010; 340: c2289Crossref PubMed Scopus (426) Google ScholarLast, the CFS originally was developed for a cohort of patients aged ≥65 years.9Rockwood K. Song X. MacKnight C. et al.A global clinical measure of fitness and frailty in elderly people.CMAJ. 2005; 173: 489-495Crossref PubMed Scopus (4075) Google Scholar The interrelationship between age, frailty, and comorbidity is complex, and identification of “frailty” may have different biologic and clinical implications across the age spectrum.10Hanlon P. Nicholl B.I. Jani B.D. Lee D. McQueenie R. Mair F.S. Frailty and pre-frailty in middle-aged and older adults and its association with multimorbidity and mortality: a prospective analysis of 493 737 UK Biobank participants.Lancet Public Health. 2018; 3: e323-e332Abstract Full Text Full Text PDF PubMed Scopus (310) Google Scholar,11Spiers G.F. Kunonga T.P. Hall A. et al.Measuring frailty in younger populations: a rapid review of evidence.BMJ Open. 2021; 11e047051Crossref PubMed Scopus (14) Google Scholar As such, this study provides much needed insight into the implications of frailty for younger people in critical care: 23% of the patients were aged <50 years, among whom frailty was identified in 6% (vs 23% in those ≥50 years). With the exceptions of chronic cardiovascular disease and metastatic cancer, the distribution of chronic illness followed a similar pattern according to frailty among younger and older patients. Variations in processes of care were evident, with an apparent decreasing tendency to undergo invasive ventilation with increasing frailty not observed in the younger cohort. However, the relationship between frailty and main outcome measures (hospital mortality, length of stay, and discharge disposition) appeared consistent between age cohorts. Interestingly, risk of readmission among frail patients recently was reported to decrease with age when frailty was identified with the use of administrative data12Hill A.D. Fowler R.A. Wunsch H. Pinto R. Scales D.C. Frailty and long-term outcomes following critical illness: a population-level cohort study.J Crit Care. 2021; 62: 94-100Crossref PubMed Scopus (1) Google Scholar; longer-term outcome data that might evidence rate and extent of recovery from critical illness according to CFS, age cohort, and comorbidity is therefore still greatly needed.As with all studies that use data collected in critical care audit databases, generalizing findings to other settings requires an understanding of critical care service organization, admission practices, and prevailing cultural attitudes in wider society to critical care admission. This means that the prevalence of frailty and its association with outcomes in Australia and New Zealand may not necessarily translate to other settings. Similarly, although a dose-response relationship was demonstrated between CFS and a range of outcomes, these should be interpreted as being associations rather than ascribing causal inference.To conclude, Darvall et al6Darvall J.N. Bellomo R. Paul E. et al.Routine frailty screening in critical illness- a population-based cohort study in Australia and New Zealand.Chest. 2021; 160: 1292-1303Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar have answered important questions as to the feasibility and prognostic value of large-scale frailty screening in critically ill patients. Crucially, they have been able to evaluate the implications of frailty assessment among younger patients and provide assurance of the validity of routine frailty assessment with the use of CFS among unselected patients who are admitted to adult critical care units. FOR RELATED ARTICLE, SEE PAGE 1292The concept of frailty as a determinant of outcome from critical illness has resonated hugely with the critical care community since its description a decade ago.1McDermid R.C. Stelfox H.T. Bagshaw S.M. Frailty in the critically ill: a novel concept.Crit Care. 2011; 15: 301Crossref PubMed Scopus (121) Google Scholar Having long appreciated the importance of “physiologic” as opposed to “chronologic” age, frailty is now widely understood to represent a state of diminished physiologic reserve more prevalent with age, which is associated with increased vulnerability towards adverse outcomes, but which is distinct from comorbidity.2Fried L.P. Ferrucci L. Darer J. Williamson J.D. Anderson G. Untangling the concepts of disability, frailty, and comorbidity: implications for improved targeting and care.J Gerontol A Biol Sci Med Sci. 2004; 59: 255-263Crossref PubMed Google Scholar The Clinical Frailty Scale (CFS) lends itself to screening for frailty in the critical care setting; it has “face validity,” combines clinical judgment with objective measurement, can be applied readily without adaptation, and has demonstrable reliability3Pugh R.J. Battle C.E. Thorpe C. et al.Reliability of frailty assessment in the critically ill: a multicentre prospective observational study.Anaesthesia. 2019; 74: 758-764Crossref PubMed Scopus (33) Google Scholar and predictive validity.4Muscedere J. Waters B. Varambally A. et al.The impact of frailty on intensive care unit outcomes: a systematic review and meta-analysis.Intensive Care Med. 2017; 43: 1105-1122Crossref PubMed Scopus (351) Google Scholar Given global demographic trends and potential critical care resource requirements of an ageing population, the added value of frailty assessment to clinical discussions and decision-making and to risk-adjusted outcome reporting have been highlighted previously.5De Biasio J.C. Mittel A.M. Mueller A.L. Ferrante L.E. Kim D.H. Shaefi S. Frailty in critical care medicine: a review.Anesth Analg. 2020; 130: 1462-1473Crossref PubMed Scopus (31) Google Scholar However, until now there has been no published research regarding the feasibility and prognostic value of population-scale screening for frailty in critically ill patients. FOR RELATED ARTICLE, SEE PAGE 1292 FOR RELATED ARTICLE, SEE PAGE 1292 In this issue of CHEST, the study by Darvall et al6Darvall J.N. Bellomo R. Paul E. et al.Routine frailty screening in critical illness- a population-based cohort study in Australia and New Zealand.Chest. 2021; 160: 1292-1303Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar is therefore an important addition to the literature. In a well-designed study, they report successful implementation of routine frailty screening using the CFS across diverse ICUs in Australia and New Zealand. Frailty was common in this cohort, approaching one in five patients when assessed at the time of admission. Consistent with findings in other studies, increasing frailty was associated with higher hospital death. Furthermore, addition of CFS score to the APACHE III-j risk prediction model improved model performance as measured by discriminant function (area under the receiver operating characteristics curve). Importantly, the authors6Darvall J.N. Bellomo R. Paul E. et al.Routine frailty screening in critical illness- a population-based cohort study in Australia and New Zealand.Chest. 2021; 160: 1292-1303Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar looked beyond mortality rates and investigated whether frailty was associated with a number of additional person-centered outcomes. The presence of frailty increased length of ICU and hospital stay and increased the likelihood of discharge to a nursing home or chronic care facility. Furthermore, patients with frailty were more likely to experience delirium and pressure injuries within ICU. These findings provide a more complete picture of the likely consequences of ICU admission in the context of frailty to inform clinicians and patients. A significant strength of the study was the extremely large sample size (n = 234,568). The authors6Darvall J.N. Bellomo R. Paul E. et al.Routine frailty screening in critical illness- a population-based cohort study in Australia and New Zealand.Chest. 2021; 160: 1292-1303Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar were able to conduct analyses using the Australian and New Zealand Intensive Care Society Adult Patient Database, a bi-national audit database. This allowed associations to be reported with precision for each category of the CFS, rather than the dichotomization that other studies have needed to undertake. Furthermore, this larger sample size allowed subgroup analyses to be undertaken in a younger cohort of patients, which confirmed that the presence of frailty was associated with worse outcomes. An important limitation was the proportion of missing data relating to CFS recording. With one in three observations missing a value for the primary exposure, the potential for bias could have been substantial. Ideally, evaluating the impact of missing data requires an understanding of the mechanism by which missing data have arisen, along with sensitivity analyses to impute missing values if the mechanism of missingness is random.7Sterne J.A. White I.R. Carlin J.B. et al.Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls.BMJ. 2009; 338: b2393Crossref PubMed Scopus (3823) Google Scholar However, the authors undertook a number of additional analyses that were reassuring and that indicated that the likelihood of significant bias due to missing data was low. This included a comparison of baseline characteristics in those with and without missing data and subgroup analyses on ICUs with low levels of missing data. Training of assessors was not mandated, and the accuracy of CFS recording was not evaluated. Although the authors have suggested that lack of individualized training was not a significant barrier to implementation, this may have contributed to measurement error. As long as such error is not differential (not systematically biased towards over- or under-estimating CFS), this would tend to reduce the strength of association between exposure and outcome.8Hutcheon J.A. Chiolero A. Hanley J.A. Random measurement error and regression dilution bias.BMJ. 2010; 340: c2289Crossref PubMed Scopus (426) Google Scholar Last, the CFS originally was developed for a cohort of patients aged ≥65 years.9Rockwood K. Song X. MacKnight C. et al.A global clinical measure of fitness and frailty in elderly people.CMAJ. 2005; 173: 489-495Crossref PubMed Scopus (4075) Google Scholar The interrelationship between age, frailty, and comorbidity is complex, and identification of “frailty” may have different biologic and clinical implications across the age spectrum.10Hanlon P. Nicholl B.I. Jani B.D. Lee D. McQueenie R. Mair F.S. Frailty and pre-frailty in middle-aged and older adults and its association with multimorbidity and mortality: a prospective analysis of 493 737 UK Biobank participants.Lancet Public Health. 2018; 3: e323-e332Abstract Full Text Full Text PDF PubMed Scopus (310) Google Scholar,11Spiers G.F. Kunonga T.P. Hall A. et al.Measuring frailty in younger populations: a rapid review of evidence.BMJ Open. 2021; 11e047051Crossref PubMed Scopus (14) Google Scholar As such, this study provides much needed insight into the implications of frailty for younger people in critical care: 23% of the patients were aged <50 years, among whom frailty was identified in 6% (vs 23% in those ≥50 years). With the exceptions of chronic cardiovascular disease and metastatic cancer, the distribution of chronic illness followed a similar pattern according to frailty among younger and older patients. Variations in processes of care were evident, with an apparent decreasing tendency to undergo invasive ventilation with increasing frailty not observed in the younger cohort. However, the relationship between frailty and main outcome measures (hospital mortality, length of stay, and discharge disposition) appeared consistent between age cohorts. Interestingly, risk of readmission among frail patients recently was reported to decrease with age when frailty was identified with the use of administrative data12Hill A.D. Fowler R.A. Wunsch H. Pinto R. Scales D.C. Frailty and long-term outcomes following critical illness: a population-level cohort study.J Crit Care. 2021; 62: 94-100Crossref PubMed Scopus (1) Google Scholar; longer-term outcome data that might evidence rate and extent of recovery from critical illness according to CFS, age cohort, and comorbidity is therefore still greatly needed. As with all studies that use data collected in critical care audit databases, generalizing findings to other settings requires an understanding of critical care service organization, admission practices, and prevailing cultural attitudes in wider society to critical care admission. This means that the prevalence of frailty and its association with outcomes in Australia and New Zealand may not necessarily translate to other settings. Similarly, although a dose-response relationship was demonstrated between CFS and a range of outcomes, these should be interpreted as being associations rather than ascribing causal inference. To conclude, Darvall et al6Darvall J.N. Bellomo R. Paul E. et al.Routine frailty screening in critical illness- a population-based cohort study in Australia and New Zealand.Chest. 2021; 160: 1292-1303Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar have answered important questions as to the feasibility and prognostic value of large-scale frailty screening in critically ill patients. Crucially, they have been able to evaluate the implications of frailty assessment among younger patients and provide assurance of the validity of routine frailty assessment with the use of CFS among unselected patients who are admitted to adult critical care units. Routine Frailty Screening in Critical Illness: A Population-Based Cohort Study in Australia and New ZealandCHESTVol. 160Issue 4PreviewLarge-scale population screening for frailty degree in critical illness was possible and prognostically important, with greater frailty (especially CFS score of ≥ 6) associated with worse outcomes, including among younger patients. Full-Text PDF