Derivation and validation of a simple clinical risk-model in heart failure based on 6 minute walk test performance and NT-proBNP status--do we need specificity for sex and beta-blockers?

BACKGROUND: It is unclear whether risk prediction strategies in chronic heart failure (CHF) need to be specific for sex or beta-blockers. We examined this problem and developed and validated the consequent risk models based on 6-minute-walk-test and NT-proBNP. METHODS: The derivation cohort comprised 636 German patients with systolic dysfunction. They were validated against 676 British patients with similar aetiology. ROC-curves for 1-year mortality identified cut-off values separately for specificity (none, sex, beta-blocker, both). Patients were grouped according to number of cut-offs met (group I/II/III - 0/1/2 cut-offs). RESULTS: Widest separation between groups was achieved with sex- and beta-blocker-specific cut offs. In the derivation population, 1-year mortality was 0%, 8%, 31% for group I, II and III, respectively. In the validation population, 1-year rates in the three risk groups were 2%, 7%, 14%, respectively, after application of the same cut-offs. CONCLUSION: Risk stratification for CHF should perhaps take sex and beta-blocker usage into account. We derived and independently validated relevant risk models based on 6-minute-walk-tests and NT-proBNP. Specifying sex and use of beta-blockers identified three distinct sub-groups with widely differing prognosis. In clinical practice, it may be appropriate to tailor the intensity of follow-up and/or the treatment strategy according to the risk-group.