Comparison of two frailty identification tools for critically ill patients: a post-hoc analysis of a multicenter prospective cohort study

Abstract Purpose We aimed to describe the association of two frailty screening tools, the validated Clinical Frailty Scale (CFS) score and the recently described modified Frailty Index (mFI) in critically ill patients. Materials and methods We performed a post-hoc analysis of a multicenter cohort of patients admitted to six Canadian Intensive Care Units (ICU) between 2010 and 2011. Frailty was screened using the CFS and the mFI. Concordance between these tools was evaluated, as well as discrimination and predictive ability for clinical outcomes after adjustments. Results The cohort included 421 patients. Prevalence of frailty was 32.8% with the CFS and 39.2% with the mFI. However, concordance between the two tools was low [(intraclass correlation of 0.37; 95% confidence interval [CI] 0.29-0.45) and partial Spearman rank correlation of 0.38 (95% CI 0.29-0.47)]. Hospital and 1-year mortality, as well as dependency after discharge and hospital readmission, were greater for frail compared to non-frail patients screened with the use of both tools. Conclusion While the CFS and mFI showed low concordance, both showed good discrimination and predictive validity for hospital mortality. Both tools identify a subgroup of frail patients more likely to have worse clinical outcomes.