Economics of Antibiotics: Incentives for Innovation and Implications for Health Care Cost – Antibionomics

(1) Estimation of patient demand
Our demand model will be derived from a random utility specification that allows for patient heterogeneity in their preferences, and that accounts for the true price faced by patients (which is usually lower than the price received by producers, due to insurance). We will also allow for patients to consider the emergence of antibiotic resistance when making their consumption decisions, i.e. they may account for the possibility that current use of an antibiotic may affect its future efficacy.
(2) Physician behavior
We will estimate physician «demand« (prescription choices for their patients) using a model similar to that for patients, allowing for physician heterogeneity (particularly in their responsiveness to scientific information) and potential myopia.
(3) Producer and regulator behavior
We will compare settings in which the producer faces few regulatory pricing constraints (such as the US and Germany) with others in which price controls play an important role (such as France), to determine whether regulators and producers account for the externality of overuse when choosing price.
The estimation approach will apply techniques described in earlier work on dynamic games. In the case of a durable good, consumers vary in their price elasticity, and firms would like to price discriminate over time. For antibiotics, the dynamics are not driven by a patient’s expectations about price, but rather the desire of producers to reduce the risk of resistance through deferring sales from today to the future.
(4) Investment in new antibiotics
We will model R&D investment as a function of expected revenues, resistance, and public funding of research using tools to estimate dynamic oligopoly games.

n/a

n/a

n/a

The antibiotic market is characterized by at least two potential market failures. First, if neither producers nor prescribers/patients internalize the negative externality of resistance stemming from inappropriate use, existing treatments may quickly lose efficacy. Second, addressing inappropriate use may undermine incentives to invest in new antibiotics. For example, restricting the use of a new antibiotic in order to delay the emergence of antibiotic resistance may depress expected profits, thus deterring pharmaceutical firms from investing in this area. Traditional models to promote innovation, in particular the patent system, are ill-suited to this context because producers have incentives to maximize the use of their treatments during the period of patent protection. The overuse that potentially results contributes to antibiotic resistance.

Since many of the infectious diseases treated by antibiotics have the potential to become pandemics, the public health consequences of these market failures may be severe. Indeed, the European Commission has identified antimicrobial resistance as a “major European and global challenge,” associated with the deaths of 25000 patients per year in the European Union and 1.5 billion euros of extra healthcare costs and lost productivity.

We will estimate structural models explaining the behavior of patients, physicians, payers, and pharmaceutical firms in the consumption and production of antibiotics. First, we will estimate patient demand for antibiotics using patient-level panel data on health expenditures extending the usual models of demand to a dynamic setting with consumption externalities. Second, we will estimate physician choices for antibiotic prescriptions using prescriber-level panel data, incorporating the dynamics of physician learning as well as consumption externalities. Third, we will consider two models of price-setting: one in which the producer of an antibiotic chooses the price, and the other in which the price is set by a national regulator. In both models, we will explore how the decision-maker accounts for the emergence of antibiotic resistance in the value and price of the product. Finally, we will estimate a model of R&D investment in new antibiotics that accounts for the behavior of patients, physicians, and payers, and in which we will also allow for government research funding that is endogenously chosen.