Page 40 - GHES-3-2

Global Health Economics and
Sustainability
Random price generators in health policy

Table 1. Examples of results with random price generators on a medical market
Results of the tests Full model with three alternatives Restricted choice set
Main parameters on three and two alternatives (alternatives 1, 2, and 3) (alternatives 2 and 3)
Age × alternative 1 (at least one oral agent)
Age × alternative 3 (no drug)
Obesity × alternative 1
Obesity × alternative 3
Sex × alternative 1
Sex × alternative 3
Price −2.4270 2.8546
Price × Medicare 0.4438 0.4227
Source: Extracted from Huttin & Hausman, 2021 Reprinted with permission of Gudapuris, Copyright 2021, Gudapuris.

level inside the economic system is not part of the study, the rules of the games and impact price competition
which targets individual-level data (providers of care and/ and nudging policies.
or patients). • IT has transformed the implementation of survey
The impact of digitalization on a macro-model instruments and data generation totally or partially;
is complex and challenges the architecture of health thus, random parameter generators may depend on
information systems. It depends on layers of information online systems of instant individual-level data.
for marketing intelligence tools in market access and Therefore, implementing decision tools designed with
counter-detailing techniques professions use to grow implicit and explicit pricing and cost of care information
2
awareness with comparative information on various (e.g., cost sensitivity simulators using discrete choice
products and services. Moreover, multi-stakeholders are experiments and reversed conjoint models) depends
now engaged in such processes (e.g., value assessment on how digitalization impacts the implementation of
frameworks from ISPOR). Other dimensions of the survey instruments, data generation, and software
digitalization in biopharma value chains or the medical development.
informatics of health systems also transform biopharma
R&D operations, manufacturing, and delivery of 5. Limitations and further developments
services, especially after the boost of the COVID-19
pandemic. 3 This paper promotes the use of more knowledge of the
structure of the demand system for medical markets
Additional risks and uncertainties for economic models since the move toward personalized medicine also means
must be considered: an extreme heterogeneity of the demand and requires a
• Transformations in information technology (IT) lead better representation of patient needs in a context where
to different issues around the control and security current trends of technological change lead to singularity
of information processing; primary care settings are (Kurzweil, 2005).
usually not equipped and often rely on the medical
informatics of hospitals or stronger health networks. The experimental study was performed with random
• Algorithmic system designs increasingly enhance, price generators, and parameters were used for one type
supplement, or partially substitute human interactions of choice model, the ML model. The area of stratification
and tasks; however, the change process to control the economics covers methodological developments, especially
impacts of fast digitalization and mobile computing in various approaches for multi-attribute choice models.
on providers of care, patients’ behaviors, and health Further experiments could compare results between ML
systems is under-researched. and different versions of latent class logit models and the
• Software developers play a critical role in designing latest development combining random parameters and
templates and generating new IT platforms that change reducing the number of classes in latent class models.

2 Counter detailing techniques with psychological “cost cues” from Useful economic models have also examined the
behavioral economics help better understand decision shifts and structure of a demand system in oligopolistic markets
identify restraints due to financial restrictions with big societal (Berry, 1984; Berry & Haile, 2014), which are often market
impact and significant effect on disease severity. characteristics of medical markets. Such a model uses
3 See, for instance, ICT diffusion in global value chains for
pharmaceuticals (data extracted from TiVa Dataset within a the mean of consumer utility to compute market shares
collaboration between Prof Huttin and OECD-Science and in different market segments (the mean utility method);
Technology directorate, 2020). Berry et al. (1995; the BLP model) successfully applied

Volume 3 Issue 2 (2025) 32 https://doi.org/10.36922/ghes.3579