Abstract
An objective of the Duke Activity Status Index (DASI) is to standardize assessments of exertional intolerance. However, patient factors associated with inaccurate assessments of exercise capacity with this tool are not well described. Patients who completed the DASI in preparation for a treadmill cardiopulmonary exercise test (CPET) at our institution from 2022 to 2023 were analyzed. Metabolic equivalents of task (METs) were calculated from the DASI score and measured from CPET. Clinical characteristics were compared using Chi square or Mann-Whitney U tests between those who overestimated versus underestimated exercise capacity via the DASI. Correlations were assessed with a Spearman test, and a binary logistic regression model was fit to the clinical characteristics to identify patient characteristics associated with overestimating METs with the DASI. Overall, 512 patients were included in the current study, and the majority (419 [82%]) overestimated exercise capacity via the DASI. DASI-predicted METs had moderate correlation with CPET-achieved METs (Spearman's rho = +0.620, p <0.001). Patients who overestimated exercise capacity by DASI were more likely to be older, female sex, Black race, on beta blockers, of higher body mass index, and with cardiac comorbidities. The regression model (Χ(2) = 87.6, p <0.001, Nagelkerke R(2) = 0.259) found older age, female sex, diagnosis of heart failure, and congenital heart disease most strongly associated with overestimating exercise capacity (all p ≤0.002). In conclusion, a majority of patients referred for CPET overestimate exercise capacity by DASI score, particularly those with older age, female sex, heart failure, and congenital heart disease.