D25-10 A Protocol to Improve the Validity and Reliability of Clinical Frailty Scale (CFS) Assessments in Critical Care Research

Abstract Rationale The Clinical Frailty Scale (CFS) is the most widely used tool for frailty assessment in the Intensive Care Unit (ICU). Proper administration and scoring is necessary to ensure valid and reliable frailty assessments in critical care research studies, particularly among older adults. We developed a protocol to improve the accuracy and consistency of CFS assessments among research personnel working on a longitudinal study of older ICU survivors. Methods In the LANTERN study, a longitudinal study evaluating determinants of functional recovery among older ICU survivors, CFS assessments are completed at eight time points: during the ICU hospitalization (with a reference timeframe of 1 month prior to admission to assess pre-ICU frailty status), at home after discharge, and during each of six monthly follow-up interviews. CFS scores range from 1 (Very Fit) to 9 (Terminally Ill). We identified key domains of CFS adjudication, including disability in activities of daily living and instrumental activities of daily living, mobility, symptoms of exhaustion, and physical activity. We then determined key thresholds in these domains that would result in a participant meeting criteria for a given CFS category. For example, bathing disability was identified as a key threshold criterion for Moderate Frailty (Category 6) versus Mild Frailty (Category 5), whereas disability in eating is a key threshold criterion for Very Severely Frail (Category 8) versus Severely Frail (Category 7). The protocol was iteratively revised, and a final protocol was developed. Results The final protocol is presented in Table 1. The first column presents CFS scores and definitions. The second column presents corresponding key determinants of each CFS level, along with additional context to guide adjudication. Five raters on the research team were trained in the protocol and reviewed participant CFS scores individually, prior to weekly team meetings. At team meetings, the raters collectively reviewed their CFS scores and determined the correct CFS score using the protocol. Incorrect scores were discussed and education provided to the team, using the protocol as a reference. Between April - October 2025, five raters reviewed 80 CFS assessments from 36 participants. Conclusions The Clinical Frailty Scale protocol is an educational and training tool to improve the validity and reliability of frailty assessment within critical care research teams. Future work includes assessing interrater reliability among research personnel within and between critical care research studies. This abstract is funded by: NIH (National Institutes of Health)