Clinical Frailty Scale in prediction of mortality, disability and quality of life for patients in need of intensive care

Clinical Frailty Scale (CFS) is an assessment tool used to describe the frailty of a patient. The scale originated in Canada and was initially developed to estimate the need for institutional care and to predict life expectancy. The CFS version currently used has 9 levels (CFS-9) where an assessment of frailty is made between 1, very fit, to 9, terminally ill.

Question

Is the Clinical Frailty Scale able to predict mortality, disability or quality of life in patients admitted to intensive care, either due to respiratory tract infection or due to other cause?

Summary

SBU Enquiry Service identified 23 primary studies considered within the scope of the review question after literature search and study selection. Eleven studies were critically appraised as low to moderate risk of bias regarding at least one outcome. In the included studies, the Clinical Frailty Scale (CFS) was used to estimate the frailty of individuals prior to admission to intensive care. The included studies were mainly conducted on elderly patients in need of intensive care, where the cause of need for intensive care varied. None of the studies examined patients who exclusively had respiratory tract infections. The studies were published from 2014 and onwards, and most studies during the last two years. Most studies originated from Europe and Canada. Several were multicentre studies, in which some had participation from Swedish clinics. The studies that were appraised as having a low to moderate risk of bias are briefly described in the text below, and all studies are presented in Table 1 and in Appendix 3 and 4.

SBU Enquiry Service identified:

  • Eight studies examined the validity and reliability of CFS in patients in intensive care (low to moderate risk of bias) [1–8]. Seven of these studies examined the prediction of mortality [1–5,7,8].
  • Two studies evaluated to which degree CFS could predict future disability for patients in intensive care (moderate risk of bias) [4,6].
  • One study examined to which degree CFS could predict future quality of life for patients in intensive care (moderate risk of bias) [4].
  • Four studies examined interrater reliability for CFS in patients in intensive care (low to moderate risk of bias) [2,3,9,10].

In studies with low to moderate risk of bias, the results showed that frailty assessment with CFS to some extent could predict in-hospital mortality and 30-day mortality. The mortality increased with each unit of the frailty scale. DeGeer et al. 2020, found that frailty assessment with CFS could predict 30-day mortality with an AUC (area under curve) of 0.74 (95% CI, 0.9 to 0.79) [1]. Two studies analysed the optimal threshold value for predicting mortality, with the aim of being able to use the scale dichotomously and found that mortality substantially increased from scale-point CFS 5 and above [1,5].

In the study by Hope et al. 2019, the risk of disability at six months was associated with an increase per unit on the CFS scale [6]. In the study by Brummel et al. 2017, the risk of disability varied depending on the assessment tool used to assess disability level, as well as on the time of assessment. Regarding prediction of quality of life, the results presented by Brummel et al showed varying results [4].

It should be noted that no study showed that CFS was able to predict either outcome (mortality, disability or quality of life) of all individuals. This implies that, there will be patients who are assessed as having a high level of frailty but do not die within 30 days, as well as patients who are assessed as having a low level of frailty but that nevertheless die.

Table 1. Identified studies according to age and outcome (reference, risk of bias)

a Measures disability or death at 6 months as a combined endpoint.
b Also presents 1-year mortality.
c Also presents 6 months mortality.
d These articles are based on the same patient material.
  Outcome
Age ICU mortality
In-hospital mortality
30-day mortality 90-day mortality Disability Quality of life Test-retest reliability
≥18 years
7 studies
Shears et al
2018, [3]

Moderate Fernando et al
2019, [11]

High
Montgomery et al
2019, [12]

High
De Geer et al
2020, [1]

Low
Brummel et al
2017, [4]

Moderate
Brummel et al
2017, [4]

Moderate
Hope et ala
2017, [13]

High
Brummel et al
2017, [4]

Moderate
Shears et al
2018, [3]

Moderate
≥50 years
6 studies
Bagshaw et alb
2014, [8]

Moderate (for this outcome)
Darval l et alc
2019, [14]

High
Kara et al
2018 [15]

High
Tipping et al
2019, [16]

High
    Hope et ald
2019, [6]

Moderate
Bagshaw et al
2014, [8]

High (for this outcome)

Hope et ald
2019, [9]

Moderate

≥60 years
1 study
Pugh et al
2019, [10]

High (for this outcome)
        Pugh et al
2019, [10]

Moderate (for this outcome)
≥65 years
3 studies
Langlais et al
2018, [7]

Moderate
Fernando et al
2019, [17]

High
Le Maguet et alc
2014, [18]

High
         
≥70 years
1 study
  Silva-Obregon et al
2020, [19]

High
Silva-Obregon et al
2020, [19]

High
Silva-Obregon et al
2020, [19]

High
   
≥80 years
3 studies
Guidet et al
2020, [2]

Low
Flaatten et al
2017, [5]

Low
Darvall et al
2019, [20]

High
Guidet et al
2020, [2]

Low
Flaatten et al
2017, [5]

Low
      Guidet et al
2020, [2]

Low
Age not reported
2 studies
Fisher et al
2018, [21]

High
        Pugh et al
2017, [22]

High

Appendices

Appendix 1 Critical appraisal checklist.pdf

Appendix 2 Excluded studies.pdf

Appendix 3 Table of included studies.pdf

Appendix 4 Studies appraised as high risk of bias.pdf

Download report

SBU Enquiry Service Consists of structured literature searches to highlight studies that can address questions received by the SBU Enquiry Service from Swedish healthcare or social service providers. We assess the risk of bias in systematic reviews and when needed also quality and transferability of results in health economic studies. Relevant references are compiled by an SBU staff member, in consultation with an external expert when needed.

Published: 6/12/2020
Contact SBU: registrator@sbu.se
Report no: ut202023
Registration no: SBU 2020/393
https://www.sbu.se/ut202023e