BASELINE LIVER RELATED BIOMARKERS AND CLINICAL EVENTS IN A SMALL PROSPECTIVE REDUCED EJECTION FRACTION COHORT

Abstract

A significant biochemical difference between heart failure phenotypes does not necessarily  translate  into  prediction  of  subsequent  events  within  a  phenotype.  Exploratory  outcome  analyses  should therefore account for multiple comparisons. This distinction is particularly important when several biomarkers and several outcomes are  assessed in the same small cohort. Reporting all comparisons, including those that do not support the  expected direction, provides a more balanced estimate of the evidence than selecting only nominally  significant results.