Clinical frailty scale and mortality in COVID-19: A systematic review and dose-response meta-analysis

• Published in: Archives of gerontology and geriatrics Vol. 93; p. 104324
• Main Authors: Pranata, Raymond, Henrina, Joshua, Lim, Michael Anthonius, Lawrensia, Sherly, Yonas, Emir, Vania, Rachel, Huang, Ian, Lukito, Antonia Anna, Suastika, Ketut, Kuswardhani, R.A. Tuty, Setiati, Siti
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01-03-2021
• Subjects: Age; Aged, 80 and over; Coronavirus; COVID-19; COVID-19 - mortality; Frailty; Frailty - diagnosis; Humans; Male; Prevalence; Prognosis; Risk stratification; SARS-CoV-2; COVID-19; Frailty; Coronavirus; Prognosis; Age; Risk stratification
• ISSN: 0167-4943; 1872-6976
• DOI: 10.1016/j.archger.2020.104324

•Each 1-point increase in CFS was associated with 12% increase in mortality.•The dose-response relationship between CFS and increased mortality is linear.•Dichotomization of CFS at a specific cut-off point is required to be clinically useful in deciding patient's care.
National Institute for Health and Care Excellence (NICE) endorsed clinical frailty scale (CFS) to help with decision-making. However, this recommendation lacks an evidence basis and is controversial. This meta-analysis aims to quantify the dose-response relationship between CFS and mortality in COVID-19 patients, with a goal of supplementing the evidence of its use.
We performed a systematic literature search from several electronic databases up until 8 September 2020. We searched for studies investigating COVID-19 patients and reported both (1) CFS and its distribution (2) CFS and its association with mortality. The outcome of interest was mortality, defined as clinically validated death or non-survivor. The odds ratio (ORs) will be reported per 1% increase in CFS. The potential for a non-linear relationship based on ORs of each quantitative CFS was examined using restricted cubic splines with a three-knots model.
There were a total of 3817 patients from seven studies. Mean age was 80.3 (SD 8.2), and 53% (48–58%) were males. The pooled prevalence for CFS 1–3 was 34% (32–36%), CFS 4–6 was 42% (40–45%), and CFS 7–9 was 23% (21–25%). Each 1-point increase in CFS was associated with 12% increase in mortality (OR 1.12 (1.04, 1.20), p = 0.003; I2: 77.3%). The dose-response relationship was linear (Pnon-linearity=0.116). The funnel-plot analysis was asymmetrical; Trim-and-fill analysis by the imputation of two studies on the left side resulted in OR of 1.10 [1.03, 1.19].
This meta-analysis showed that increase in CFS was associated with increase in mortality in a linear fashion.