The perils of proxy outcomes in evidence-based medicine: the case of recombinant factor VIIa

There is currently a major controversy surrounding the off-label use of recombinant activated Factor VII (rFVIIa). The literature offers at best inconclusive support for its use in the management of bleeding in any situation other than in patients with Factor VII deficiency or hemophilia A or B with antibodies to factor VIII or factor IX respectively. This evidence has not, however, slowed the growth in off-label prescribing of rFVIIa in other situations including intracranial hemorrhage, cardiac surgery, trauma, transplantation and prostatectomy. We argue that the controversy surrounding such off-label use of rFVIIa stems in part from different understandings of the purposes and methods of evidence-based medicine (EBM)—in particular the use of gross, proxy outcomes such as mortality to determine the effectiveness of an intervention. We then argue that clinical impression and expert opinion should not be dismissed outright on the basis of inconclusive evidence of ineffectiveness. Those who advocate its continued use, however, have a responsibility to demonstrate the benefits that may be missed in current manifestations of EBM . We suggest that the only justified verdict in the case of rFVIIa is currently 'Not Proven'.