Although clinical trials are being used to evaluate economic outcomes
of new agents, there are methodological problems. Decisions based on t
hese analyses may lead to inefficient use of medical resources. Random
ized clinical trials provide important information on the efficacy of
new pharmaceutical agents for cancer patients. Policy makers are likel
y to require both economic and clinical data in order to approve pharm
aceuticals for widespread use. Clinical trials provide an opportunity
to evaluate economic outcomes for new agents. However, the interpretat
ion of economic analyses of clinical trials raises issues related to p
erspective of the investigators, study design, collection of data on r
esource utilization, and generalizability of data to other settings. I
n this paper, we review these issues and illustrate problems associate
d with analyses of economic data from a recent phase III trial of hema
topoietic growth factors. Clinical results were similar in both Paris
and New York in this phase III trial. However, economic results differ
ed markedly between the hospital in Paris and the hospital in New York
. While significant savings in terms of fewer days in the hospital and
fewer laboratory tests and radiographs for the granulocyte-macrophage
colony-stimulating factor (GM-CSF) patients were noted at the New Yor
k hospital, resource savings were not identified at the hospital in Fr
ance. Caution must be used when reimbursement policies are based on ec
onomic analyses of clinical trials. Policy decisions must be based on
studies that are carefully conducted, analyzed, and interpreted from b
oth a clinical and an economic perspective.