established the Harvard-Partners Center for Genetics and Genomics (HPCGG),^35
aiming to move genetics and genomics into clinical practice by offering a frame-
work for stimulating research that integrates genetic knowledge into patient care
and developing the information infrastructure that facilitates the application of
genetic and genomic data in research and clinical settings. However, the cost and
the cost-effectiveness of such projects and their future impact on the overall health
care costs (and the exact potential for generating additional health care benefits) are
not clear at the moment, which is not surprising given that there is very little
evidence on the value for money of health information systems in general.^36
The process of defining the technology for evaluation may be more complex for
companion diagnostics than for pharmaceuticals also because the former often
requires the input from many different experts, from laboratory scientists to clini-
cians. Economic evaluations of companion diagnostics may need to account for the
multitude of providers and services required to ensure that genetic tests are cor-
rectly interpreted and used. A companion diagnostic may comprise a chain of
elements, from the diagnostic test itself to the interpretation of the test result and
the treatment decision based on the test. These elements in turn may require a
complex health information system and the services of many specialists (e.g., a
clinician, a laboratory specialist who conducts the test, a clinical geneticist who
assists in developing a diagnosis and treatment strategy, etc.). Many of the services
provided by these specialists may not be eligible for reimbursement or may greatly
vary between different settings (hence hampering transferability). Each of the
individuals, systems, and institutions in the diagnostic chain may also introduce a
behavioral component and a certain standard of quality of their work into the chain,
which may significantly affect the effectiveness of the companion diagnostic
medicine,^37 and this a priori means an increase in the uncertainty surrounding the
potential benefits produced by the test. Phillips et al. ( 2009 ) claim that a substantial
percentage of HER2 tests performed by community laboratories in the US are
inaccurate based on comparisons with higher-volume central or reference labora-
tories, and considering the seriousness of the implications of inaccurate tests for
patients’lives and the impact on the health care system, it is essential to generate
more data on test quality and interpretation. As already noted, economic models
evaluating the cost-effectiveness of companion diagnostics, or treatments, should
not ignore the influence that inaccurate diagnosis has on the incremental cost-
effectiveness of therapy.
(^35) Further information available at their website, athttp://personalizedmedicine.partners.org/.
(^36) The review on the topic of value of information systems is provided by Bassi and Lau ( 2013 ),
pp. 792–801.
(^37) As discussed, from various perspectives by different authors, e.g., Goldberg ( 2009 ), pp. 1–8;
Hamburg and Collins ( 2010 ), pp. 301–304; Annemans et al. ( 2013 ), pp. 20–26.
124 A. Bobinac and M. Vehovec