Abstract 19748: Predicting Outcome in Adult Congenital Heart Disease: Correlating the Seattle Heart Failure Model with peak VO2 from Cardiopulmonary Exercise Testing
Background: The adult congenital heart disease (ACHD) population continues to grow and has increased risk of mortality and morbidity, including heart failure. Currently, no model exists to identify high risk patients. Cardiopulmonary exercise testing (CPET) is used in risk stratification with peak VO2 (pVO2) as a surrogate marker for outcomes in ACHD. This measurement requires specific equipment and expertise, not widely available outside of tertiary centers. The Seattle Heart Failure Model (SHFM) is a validated prediction model that estimates total mortality in patients with heart failure by using commonly obtained clinical, laboratory, medication, and device variables.
Hypothesis: The SHFM predicts CPET pVO2 in moderate and complex ACHD.
Methods: We performed a retrospective IRB approved study of 106 individuals (>18 years of age) with transposition of the great arteries (TGA), Ebstein anomaly, tetralogy of Fallot (TOF), double outlet right ventricle (DORV) and single ventricle (SV) with a clinically indicated CPET at the Massachusetts General Hospital (MA, USA) and Montefiore Medical Center (NY, USA) between 2008 and 2012. Clinical/laboratory data (within one year of CPET) were obtained for SHFM calculation.
Results: Mean age was 35.6 years (range 18-64), 55.7% are female, average systemic ventricle ejection fraction was 56.3% (18-83%). Medication use: 9.4% of individuals on ACE-I/ARB, 34% beta-blockers, 23.6% diuretics. The mean SHFM 1 year survival was 93.9% (72.5-99.0%); mean pVO2 was 21.2 cc/kg/min (10.2-41.6). The SHFM 5-year model was able to predict an outcome of pVO2<20 in the group as a whole with a C statistic of 0.69. In individuals with TGA (systemic right ventricle) and Ebstein, the predictive ability of the SHFM 5-year model had a C statistic of 0.86 and 0.87, respectively. Peak VO2 and SHFM predicted survival correlated strongly among TGA and Ebstein patients (r=0.65 and 0.57, p < 0.03), but not among SV or TOF patients (r.= -0.02 and 0.12, p > 0.4).
Conclusion: The SHFM is able to offer an increased ability to discriminate among ACHD patients who have a pVO2 < or > 20cc/kg/min. This model is easily created using clinically obtained data. Targeted application of SHFM in the community may have predictive value and guide care in select populations.
- © 2012 by American Heart Association, Inc.