A novel risk score based on exercise capacity to predict all-cause mortality in adults with congenital heart disease

Abstract Background Impaired exercise capacity in adults with congenital heart disease is associated with mortality. Thus, exercise capacity is a potential tool for predicting risk in this population. The present study attempts to further explore the usefulness of exercise capacity in risk assessment of all-cause mortality, together with a limited number of easily accessible clinical variables. Methods The national register of congenital heart disease was searched for adults (18-60 years of age) with at least one bicycle ergometer exercise test and follow up in the register. The exercise capacity was classified based on percent of predicted exercise capacity (EC%pred); normal (> 70%), moderately impaired (50-70%) or severely impaired (<50%). All patients were followed for 10 years or until death, if occurring earlier than 10 years after the exercise test. Data were analyzed using logistic regression modelling. A risk evaluation model was constructed based on percent of predicted exercise capacity, age, and complexity of the heart lesion. Finally, the model was converted into a 5-point (p) risk score used to predict 10-year mortality (EC%pred >70%=0p, 50-70=1p, <50%=2p, age <25=0p, 25-40=1p, 40-60=2p, simple lesion=0p, moderate/severe complexity of heart lesion=1p). Results 1171 patients (mean age at exercise test 30.5 years, 44% women) were included with 70 deaths occurring within 10 years. A score 0-1 corresponded to <1% 10-year all-cause mortality, 2p 5%, 3p 9%, 4p 21% and 5p 33% mortality risk (Table 1). The area under curve for the model was 0.75 (95% CI 0.69-0.81). Conclusion An easily performed risk assessment containing % of predicted exercise capacity, age and complexity of heart lesion performs well in detecting patients at both low and high risk of all-cause mortality within 10 years after an exercise test.