Seminar: Machine Predictions and Human Decisions with Variation in Payoffs and Skill: The Case of Antibiotic Prescribing

Artificial Intelligence has the potential to improve public health interventions but their effectiveness can remain limited if humans hold private clinical information and follow private objectives. Using the empirical example of antibiotic prescribing for urinary tract infections, we show that automated prescribing with the aim to reduce antibiotic use would fail to improve on physician decisions. Instead, optimally delegating a share of decisions to physicians, where they possess private diagnostic information, effectively utilizes the complementarity between algorithmic and human decisions. Combining physician and algorithmic decisions can achieve a reduction in inefficient overprescribing of antibiotics by 20.3 percent. We explore potential drivers of these improvements, including diagnostic information and physicians’ objectives when prescribing antibiotics.

Hannes Ullrich is Deputy Head of the Department Firms and Markets at the DIW Berlin and Associate Professor of Economics at the University of Copenhagen. He is also a CESifo Research Affiliate and Fellow at the Berlin Centre for Consumer Policies (BCCP). Hannes is an applied microeconomist with research interests in empirical industrial organization, health economics, and personnel economics. His work has been published in The Economic Journal, Management Science, the Journal of Health Economics, the Journal of Human Resources, and other peer-reviewed outlets. He holds a Starting Grant from the European Research Council for research on antibiotic prescribing and resistance.

Alternatively, join online:

Microsoft Teams

Meeting ID: 372 159 458 89
Passcode: SD63UV7z