Assessing frailty in the intensive care unit: A reliability and validity study

To describe pre-ICU frailty in critically ill patients using the Clinical Frailty Scale (CFS).

We included patients ≥18years admitted to 2 ICUs in Hamilton, Canada. The ICU Research Coordinator (RC) generated 3 CFS scores using: 1) chart review, 2) family interview, 3) patient interview. Subsequently, an overall impression was captured in a final score. Mean differences were calculated to assess the RC intra-rater reliability and inter-rater reliability of chart reviews by the RC, Occupational Therapist (OT), and Geriatrics Resident (GR). Scores were also compared between younger and older patients. We also analyzed the relationship between CFS scores and mortality.

We prospectively enrolled 150 patients (mean age 63.8 [SD 15.3] years, APACHE II score 21 [SD 7.3]). CFS were similar between RC, OT, and GR chart reviews (p>0.05 for all comparisons). There was no difference between RC chart review and RC final score, or between RC patient interview and RC final score. Scores following the RC family interview and the RC final score were significantly different (-0.24, 95% CI -0.38, -0.09, p<0.01). Each 1-point increase in the final CFS scored by the RC was weakly associated with ICU mortality (odds ratio 1.18, 95% CI 0.84-1.66, p=0.33), and hospital mortality (OR 1.19, 95% CI 0.89, -1.59, p=0.24).

CFS scores can be generated using medical chart review and can be reliably completed by ICU clinicians and research staff.