In this paper, a general investment appraisal model is presented which
shows how pharmaceutical companies could take profit considerations i
nto account when making decisions about the design of randomized contr
olled trials. A general model is presented based on the net present va
lue method of investment appraisal. The approach is illustrated with a
hypothetical example which shows how optimal (net present value maxim
izing) designs can be determined based on choices about sample size an
d endpoint measurement. The method could be extended to accommodate co
nsiderations about other trial design features, and could be used to d
etermine a portfolio of studies which maximizes the expected return on
a given development or trial budget. Furthermore, the approach could
be used by pharmaceutical companies to evaluate the incremental costs
and benefits of incorporating non-clinical objectives into trials, suc
h as quality of life research and economic evaluation studies. A numbe
r of practical difficulties would need to be overcome to utilize the a
pproach. Directions for further research are therefore highlighted cen
tred on the key components of the model: a trial cost function, a prod
uct demand function, innovation diffusion processes and Bayesian appro
aches to trial design. (C) 1998 John Wiley & Sons, Ltd.