Abstract 3015: Dynamic Updating of TIMI Risk Score for STEMI: An ExTRACT-TIMI 25 Analysis
Background: The TIMI risk score(TRS)for STEMI captures baseline prognostic information. We incorporated events during the index hospitalization (hosp) to create an updated TRS at discharge.
Methods:Data from 20,479 patients in the ExTRACT-TIMI 25 trial were analyzed(Cox model) to predict 1 year mortality. Events during the index hosp. were used to update the TRS. Patients who died during the index hosp. were excluded. The hazard ratio for each event was used to assign points in the updated TRS (Fig 1⇓).
Results: Events during hosp. changed estimated mortality– c statistic was 0.73 at baseline and 0.76 at discharge(Fig 2⇓):For example,patient A had a baseline TRS of 4, survived the index hosp. without events and the estimated 1 year mortality dropped from 7.1 to 5.5%. Events in patient B were associated with more than 4 fold increase in estimated mortality.
Conclusions: Dynamic updating of the TRS for STEMI with clinical events during the index hosp. refines the estimated 1 year mortality. This may help in targeting complex, expensive therapeutic strategies in those patients with the greatest risk.