EXAMINING THE PROGNOSTIC VALUE OF FRAILTY DETERMINED BY A MULTIDIMENSIONAL QUESTIONNAIRE, THE FRAILTY ASSESSMENT IN CARE-PLANNING TOOL (FACT), IN CARDIAC SURGERY

Abstract

We assessed whether the Frailty in Care-Planning Tool (FACT) added discriminatory power to a standard cardiac surgery risk prediction model for mortality and major adverse cardiac events (MACE). To date, risk prediction models employed in cardiac surgery do not include frailty. We assessed whether the FACT, and each of its domains, were predictive of MACE, and MACE and/or mortality, after adjustment for the EuroSCORE II, a prominent risk prediction index. Likelihood ratio tests, sensitivity, specificity, and area under the ROC curve were used to compare models. For mortality and/or MACE, as well as MACE alone, certain domains of the FACT (Social, Daily Tasks, and Memory) provide improved fit over the EuroSCORE II alone; however, this was not the case for mobility or the overall FACT score. Traditional risk assessment scores such as the EuroSCORE II may benefit from having a measure of frailty included as a risk factor.