Evaluating the Impact of Treatment Effectiveness and Side Effects in Prescribing Choices - Innovation and Marketing in the Pharmaceutical Industry - page 172

cardiovascular risk such as Vioxx instead of a less effective but safer drug? If the
answer is yes, it may benefi t the society that Vioxx remains in the market as long as
patients are well informed of the risk. Only after understanding the willingness of
trade-off between effectiveness and side effects among patients and physicians can
policy makers fully study the welfare impact of regulations on entry or termination
of drugs in the market. From the managerial perspective, to effectively market their
products, pharmaceutical companies have to convey through marketing communi-
cations the benefi ts of their drugs that most physicians and patients consider as
important and that they have advantages over competitors.
Using clinical trial data alone, it is diffi cult for researchers to make inference of
the physician and patient evaluations. This is because participants of clinical trials
are randomly assigned to treatment groups, rather than actively choosing which
treatment to receive. More recently, we have seen an increasing trend in clinical trial
studies that researchers supplement the objective treatment outcome data with
participants' evaluation of treatment outcomes or their compliance behavior to infer
the importance of treatment effectiveness vs. side effects (e.g., Chan and Hamilton
2006 ; King et al. 2007 ; Lamiraud and Geoffard 2007 ; Bordeleau et al. 2010 ).
In addition to clinical trials, pharmaceutical companies and consulting fi rms have
routinely collected physician and patient prescription choice data at both aggregate
and individual level after a drug's launch. This type of data records which treatment
a physician chooses for a particular patient when faced with alternative treatments
in the real market environment. In standard economic models consumer preferences
for multiple product attributes are inferred from observed product choices. Similarly,
physician and patient preferences for treatment effectiveness and side effects may
also be inferred from observed prescriptions. One of the challenges of using this
approach is the limitation of prescription data. To understand the trade-off between
treatment effectiveness and side effects, we argue that prescription choice data alone
is insuffi cient; instead, researchers have to use additional data sources, such as
observed treatment length in Crawford and Shum ( 2005 ) and self-reported switch-
ing reasons in Chan et al. ( 2013 ). Furthermore, unlike some other product catego-
ries such as consumer package goods where consumers have more knowledge of
the product quality and their own preferences, physicians and patients usually have
large uncertainty when evaluating treatment effectiveness and side effects espe-
cially for those drugs new in the market. Physicians typically rely on a variety of
information sources to learn about these attributes, in addition to the public domain
information such as clinical trial reports. To correctly infer the physician and patient
preferences, it is important for researchers to take account of how these uncertain-
ties may impact prescription choices. It is also important for pharmaceutical com-
panies to understand how various types of marketing channels may help to reduce
the physician and patient uncertainty of the important drug attributes that eventually
will infl uence prescription decisions.
This chapter provides an overview of research on evaluating effectiveness and
side effects of prescription drugs. We fi rst briefl y review the standard industry
practice of using clinical trials data to measure the effectiveness and side effects of
a drug. We then discuss how researchers may utilize clinical trials to gauge