Biomedical Engineering Reference
In-Depth Information
discrepancy is not readily available and deserves more research. But, the benefits of
DTCA seem to be broader than those captured in a brand sales model. For example,
Osinga et al. ( 2011 ) recently showed that DTCA has a significant positive effect on
stock return beyond and above its revenue effects. Hence, the addressee of DTCA is
not only the patient (or physician), but also the investor. In addition, we know from
brand research that ad expenditures show only a marginal sales effect but contribute
significantly to brand equity which in turn is a driver of firm performance (Srinivasan
and Hanssens 2009 ).
Another striking discrepancy of the Kremer et al. study refers to the fact that
elasticities obtained from individual-level models are considerably smaller than
from aggregate-level models (see also Manchanda and Honka 2005 ). Idiosyncrasies
in data collection may be responsible for this difference. Another possible explana-
tion is the type of response model. Individual-level models typically model brand
choice in a logit framework (e.g., Gonzalez et al. 2008 ; Narayanan and Manchanda
2009 ) while aggregate-level models often resort to the double-log specification
(e.g., Fischer and Albers 2010 ; Osinga et al. 2010 ). Both specifications have very
different implications for the behavior of elasticity and consequently for optimal
expenditure levels (see also Albers 2012 ). Future research may investigate this issue
in more detail.
19.3.3
Recent Model Developments
As Table 19.2 shows, the literature on marketing spend response models for drug
demand is very rich. It is beyond the scope of this article to review all these models
in detail. I rather shall highlight a few recent trends in model development.
The role of quality . Drugs are highly complex products that result from a long and
capital-intensive process of development and approval through national surveil-
lance authorities. Prior to approval the drug needs to show sufficient evidence on its
effectiveness and safety from clinical trials. Drug quality is a multidimensional con-
struct. Relevant dimensions are, for example, the improvement of a medical condi-
tion, the bioavailability, and the number of indications, side effects and interactions
with other drugs. Product management regularly uses results from clinical 1 and sur-
veillance studies in sales folders and talks for physician visits and advertisements in
medical journals.
Azoulay ( 2002 ) proposes new measures of scientific evidence and incorporates
them into a drug demand model. Specifically, he develops flow and stock variables
that measure the research output from clinical placebo and comparative studies.
Analyzing aggregate brand sales data in a discrete choice modeling framework and
1 Product management does not only refer to phase III studies that are required to obtain drug
approval but also to phase IV studies. These studies are carried out after launch and may include
comparative studies to show the superiority of the drug over competitive drugs.
Search WWH ::




Custom Search