BACKGROUND: Composite outcomes are commonly used in heart failure trials. The aim of this article is to discuss the advantages and disadvantages of composite outcomes and recommend guidelines for reporting them. Examples are used to illustrate key points. METHODS AND RESULTS: A workshop jointly planned by the Heart Failure Society of America and the US Food and Drug Administration was convened in April 2004. One of the panel discussions concerned the use of composite outcomes in heart failure trials. With use of composite outcomes, event rates are higher and if it is reasonable to assume that the treatment effect is similar for each component of the composite outcome, sample size will be smaller than using one of the components as the primary end point. Composites end points are difficult to interpret if effects are not similar for all components or if the effect of treatment is primarily on a more common, less serious component of the composite. Composite outcomes typically only focus on the first occurring event. This can lead to a substantial loss of information in some trials. When composite end points are used, data collection for all components should continue until the end of the trial so that each component can be separately evaluated. CONCLUSION: Composite end points should be used with caution. Additional research is need on optimally weighting components of composite outcomes and to better using the entire event history of patients in heart failure trials. Further analyses of completed trials could be useful in this respect.
BACKGROUND: Composite outcomes are commonly used in heart failure trials. The aim of this article is to discuss the advantages and disadvantages of composite outcomes and recommend guidelines for reporting them. Examples are used to illustrate key points. METHODS AND RESULTS: A workshop jointly planned by the Heart Failure Society of America and the US Food and Drug Administration was convened in April 2004. One of the panel discussions concerned the use of composite outcomes in heart failure trials. With use of composite outcomes, event rates are higher and if it is reasonable to assume that the treatment effect is similar for each component of the composite outcome, sample size will be smaller than using one of the components as the primary end point. Composites end points are difficult to interpret if effects are not similar for all components or if the effect of treatment is primarily on a more common, less serious component of the composite. Composite outcomes typically only focus on the first occurring event. This can lead to a substantial loss of information in some trials. When composite end points are used, data collection for all components should continue until the end of the trial so that each component can be separately evaluated. CONCLUSION: Composite end points should be used with caution. Additional research is need on optimally weighting components of composite outcomes and to better using the entire event history of patients in heart failure trials. Further analyses of completed trials could be useful in this respect.
Authors: Ignacio Ferreira-González; Jason W Busse; Diane Heels-Ansdell; Victor M Montori; Elie A Akl; Dianne M Bryant; Pablo Alonso-Coello; Jordi Alonso; Andrew Worster; Suneel Upadhye; Roman Jaeschke; Holger J Schünemann; Gaietà Permanyer-Miralda; Valeria Pacheco-Huergo; Antònia Domingo-Salvany; Ping Wu; Edward J Mills; Gordon H Guyatt Journal: BMJ Date: 2007-04-02