4 Methodological Challenges for Economic Evaluations
of Companion Diagnostics
Typical economic evaluation requires averaged data, preferences of average
(patient) populations, and the information on real-world treatment effectiveness.
Products of personalized medicine, on the other hand, produce treatment strategies
that are based on the characteristics of individuals and their diseases and in that
sense are much less averaged, at least not averaged for the entire populations.
Hence, the economic evaluations of traditional health care products and the eco-
nomic evaluations of companion diagnostics may not be perfectly aligned in scope,
and the guidelines for traditional economic evaluations and outcomes research will
need to adapt to changes in the type of evidence and the approach applicable to
personalized medicine.^32 In fact, Towse et al. ( 2013 ) argues that the focus of future
research in the area of improving the methodology of HTA and economic evalua-
tions may increasingly lie with addressing the emerging issues for the evaluation of
companion diagnostics. Here we discuss some of the emerging issues in economic
assessment for companion diagnostics.
4.1 Complexity of Personalized Medicine
The use of new companion diagnostic technologies that provide information on the
genetic profile of diseases holds a promise of matching patients with the treatments
that ensure best possible outcomes. This opportunity reveals a complex network of
elements related to diagnostic and preventive procedures, treatment options and
care pathways, amalgamation of biological and other information, risk factors, etc.,
many of which we still do not fully understand, nor do we fully understand their
intricate interdependencies. Patient outcomes may be, for instance, influenced by
different genes and gene variations, and each gene can influence various health
outcomes. In turn, each outcome may be modified by interactions with other genes
and with environmental factors, such as nutrition or smoking habits. It is still
necessary to determine which genetic markers carry the most clinical significance
and to identify many genetic variants that are correlated with particular drug
responses.^33 Naturally, this complexity represents a noteworthy challenge to eco-
nomic evaluations, conventionally applied to pharmaceuticals and “one size fits all”
basis.
The cost-effectiveness studies of companion diagnostics will need to take
account of the prevalence of the biomarker or the genetic mutation in the popula-
tion, i.e., the prevalence of the underlying (genetic) characteristic to be tested using
(^32) Further debate available in Byron et al. ( 2014 ), pp. 1469–1476.
(^33) The discussion on the relation between genetic variations and drug response, see, for instance,
Hamburg and Collins ( 2010 ), pp. 301–304, Ma and Lu ( 2011 ), pp. 437–459.
122 A. Bobinac and M. Vehovec