Abbreviated Duke Activity Status Index for Risk Stratification in Heart Failure

Area of Science:

• Cardiology
• Clinical Assessment
• Health Informatics

Background:

• The Duke Activity Status Index (DASI) is a 12-item questionnaire assessing functional capacity and prognosis.
• There is a need for abbreviated versions of the DASI for efficient clinical application.

Purpose of the Study:

• To develop and validate an abbreviated DASI questionnaire (aDASI).
• To maintain the accuracy and prognostic reliability of the original DASI in a clinical setting.

Main Methods:

• Machine learning (XGBoost algorithm) was used to develop the aDASI from the Cleveland Clinic GeneBank Study.
• The aDASI's prognostic value was validated in heart failure patients from the GeneBank and the independent GUIDE-IT trial.

Main Results:

• The aDASI showed high congruence and accuracy with the original DASI score (R = 0.97, p <0.001).
• Both DASI and aDASI demonstrated strong correlation and predictive value for all-cause mortality in the GUIDE-IT cohort.

Conclusions:

• The abbreviated DASI (aDASI) is a reliable and accurate tool for assessing functional capacity.
• The aDASI effectively predicts all-cause mortality, performing comparably to the original DASI questionnaire.