Agreement between patient self-reported and assessor-rated frailty using the Clinical Frailty Scale

Guidelines recommend routine preoperative frailty assessment; however, existing data suggest that adherence is low. The ability with which patients can evaluate their frailty before surgery is unknown. Our objective was to evaluate the agreement of patient self-reporting of frailty compared with trained assessors using the Clinical Frailty Scale (CFS).

This was a sub study of a multicentre prospective cohort study of people ≥65 yr of age having elective noncardiac surgery. Preoperatively, patients and trained assessors independently rated patient frailty using the CFS. Agreement between raters on the ordinal scale was estimated using weighted kappa. We evaluated agreement after dichotomising scores (CFS≥4) using Cohen's kappa. We also described the discriminative ability of self-reported and assessor-reported CFS scores in predicting the occurrence of death or new disability 90 days after surgery.

There were 531 participants with both self-reported and assessor-reported CFS. Agreement was moderate (weighted kappa=0.58 [95% confidence interval, CI 0.53-0.63], Cohen's kappa=0.61 [95% CI 0.54-0.68]). Assessor-rated scores were systematically higher than self-rated scores by a mean of 0.45 points (95% CI 0.39-0.52; P<0.001). When predicting death or new disability, paired receiver operating characteristic curve analysis revealed no difference in discriminative capability of self-rated or assessor-rated CFS assessment (c-statistic difference 0.010 [95% CI -0.072 to 0.092; P=0.81]).

Patients self-report their frailty status using the CFS with moderate agreement to a trained frailty assessor. Self-reported and assessor-rated scores had similar predictive capability for postoperative death or new disability.