Christine Huttin (Aix-Marseille Université) 

give a presentation on : 

Comparative approaches for economic modeling in healthcare markets, relevance of random price generators. 

Abstract : 

OBJECTIVES: this communication aims to discuss how and in which context a random price generator for health care markets can be useful in comparison with other approaches for economic modeling and pricing of medical products and services 
METHODS: a comparison of the different approaches used will be discussed with their statistical limitations: nested logit models, cumulative logit models, latent model of choices , shrinkable estimators with Hierarchical Bayesian approach, combination of robust pricing optimization with multinomial logit choice models and choice models with random price generators; an application in progress of the BLP model on NGS sequencing will also be presented. 
RESULTS: random generators prove to be useful but do not fit any kind of structural conditions to adjust supply and demand of health care markets, especially in more regulated health and economic systems. An approach by sets of alternatives for choice models allows to adjust for heterogeneity of demand but lack of price data in some markets makes it useful to rely on shrinkable estimators with HB approaches. In duopoly cases, the BLP model currently tested, may help investigation of antitrust policies. 
CONCLUSIONS: the complexity of health care markets lead to consider various combinations of models and even participatory modeling to reach growing consensus with various layers of decision makers both on the payer side and economic actors often switching to data driven models for both drug and vaccines supplies.