Abstract 13767: Comparison of the Intermountain Risk Score to Other Risk Prediction Models
Background: Risk stratification is essential for optimizing patient outcomes. Numerous risk prediction models have been developed to provide tools for clinicians to integrate a variety of factors to determine outcome risk. However, many risk scores (RS) are cumbersome, not easily computed, or queriable. We have recently shown the predictive ability of the easily computed sex-specific Intermountain Risk Score (IMRS), which incorporates age and components of the BMP and CBC, for 30 day, 1 and 5 year death among a variety of groups.
Methods: Patients (N=7,354) of the Intermountain Heart Collaborative Study that had BMP and CBC measurements were studied. A variety of well-known RSs were identified and computed for area under the curve (AUC) within their respective target patient population and outcome. The sex-specific IMRS using the 30 day death equation was then computed within that RS's specific population and for that outcome.
Results: The table lists the RSs that were identified and their respective population, outcome, AUC, and the IMRS AUCs for the same patient population and outcome. The EuroScore, Grace, SAPS II, and TIMI for STEMI were unable to be computed because RS components were non-queriable.
Conclusion: The sex-specific IMRS was shown to be as predictive as many of the other identified RSs even though it was not derived for that specific outcome or population. The IMRS is easily-computed from routinely-collected lab results, age, and gender (unlike some of the other RSs that contain difficult to collect components). Because of its general applicability and its comparability, IMRS should be further studied as a prognostic tool in general care settings.
- © 2010 by American Heart Association, Inc.