than electronic databases must be used. One advan-
tage of using clinical trial patients is the certainty of
having patients with the condition of interest – the
trade-off being a concern for the generalizability of
information to the larger population.
Pharmacoeconomic baseline data should not be
considered in isolation, but as one aspect of data
that must be considered as a part of thewhole. Once
the burden of illness information is collected and
analyzed, the development team must move to plan
for ways to measure and document the clinical,
economic and humanistic impact of the new phar-
maceutical entity or other intervention.
23.6 Studies within clinical trials:
techniquesThe information generated from the burden of ill-
ness component of a pharmacoeconomic strategy
will serve as a useful guide for the design of phar-
macoeconomic components within clinical trials.
Obviously, this must be factored against the prior
judgment of whether or not disease-specific QOL
instruments are required at all. Healthcare resource
use, measures of lost productivity and indirect
financial cost measures may be all that is required.
The process of incorporating pharmacoeconomic
measures into clinical trials should begin before a
draft protocol is ever created.
Both the quantity and the types of data able to be
collected will be affected by the nature of the
clinical study: patients may be inpatients or out-
patients, and this in turn will govern the nature of
pharmacoeconomic data that can be recorded. It is
also important whether a clinical trial is intended as
a pivotal trial for registration or not: if a study is
pivotal, then a clinical efficacy measure will have
to be the primary end point. Pharmacoeconomic
parameters can still be incorporated into such a
study as secondary end points, and still provide
valuable information. If, on the other hand, the
clinical research addresses a health system delivery
issue, then the pharmacoeconomic end points may
well be primary, and the study design need not be
constrained by FDA-mandated requirements for
the double-blind, placebo-controlled aspects of
proof of efficacy.
Early phase: feasibility/late phase: data. As the
development moves from early phase II through to
phase IV, the rationale for incorporating pharma-
coeconomic parameters into studies should evolve.
Initially, measures may be used in studies with
small sample sizes to gain experience with certain
instruments, or to determine which instrument is
preferred for use in larger studies. Early on, the
project team may think that everything conceivable
(‘all but the kitchen sink’) is being included in a
study. In some cases, the instrument feasibility
study could be done as a separate study, but the
costs in terms of additional patients needed, and
other resources required, need to be carefully con-
sidered before a decision to reject the inclusion of
several pharmacoeconomic instruments in one
early clinical study.
As the product moves from phase II into phase
III, the number of seemingly redundant instru-
ments should decline as the obvious choice, or
best guess should rise to the top. If the goal of
phase III studies is to file an NDA or gain regula-
tory approval, the studies may not be appropriately
designed to capture the additional information
deemed necessary for the product’s success. In
some cases, separate pharmacoeconomic studies
may be needed prior to marketing.
Good advice is to prioritize at this stage of
development: which pharmacoeconomic compo-
nents critical and which are not for product launch
or shortly thereafter? Thus, there is usually a need
to strike a balance between getting information in a
timely fashion, meeting regulatory demands and
meeting demands of the marketplace; that balance
may often have to be struck pragmatically.Confidence and validity of data
As in any other scientific endeavor, the validation of
the database is as important as its interpretation;
pharmacoeconomic variables require two degrees
of confidence, that is in the accuracy and thevalidity
of what has been measured.
Consider two opposite examples of pharmacoe-
conomic measurement. In one case, patients could
describe their impression of the impact of an inter-
vention on their QOL following completion of a23.6 STUDIES WITHIN CLINICAL TRIALS: TECHNIQUES 297