Pre-conference workshops
5 July | Full day (09:00 am - 04:00 pm)
Room: GSH3.Ur1
Arne Risa Hole, Universitat Jaume I, ES
The aim of this course is to give the participants a solid grounding in the econometric methods used to analyse discrete choices. The lectures will cover the theory behind the methods and give examples of their use in health economics and related fields. The practical work will give participants the opportunity to use the methods to analyse real-world datasets. The intended learning outcome is that by the end of the course the participants will be able to apply the methods in their own research.
No prior knowledge of discrete choice methods is assumed. The practical work will be carried out using the Stata statistical software package, and participants are required to bring their own laptop with Stata (version 13 or newer) installed.
5 July | Full day (09:00 am - 04:00 pm)
Room: GSH3.Ur2
Richard van Kleef, Erasmus School of Health Policy & Management, NL
Many payment systems in health insurance and healthcare rely on risk adjustment to avoid incentives for risk selection and/or achieve fairness objectives. Over the past four decades, risk adjustment algorithms have evolved from simple demographic models to sophisticated morbidity models using health indicators based on diagnoses and other information derived from (prior) healthcare utilization. Despite these improvements, however, substantial selection incentives and fairness issues remain. Around the world, regulators of healthcare systems and payers of care (typically governments or insurers) face the challenge of further improving their risk adjustment methods. In this pre-conference session, we will discuss emerging topics and new research.
Please click here for more information about the workshop.
The objective for the session is three-fold:
- Summarize recent research on risk adjustment. Topics will include the selection and operationalization of risk adjuster variables, and innovative methods for estimating payment weights using machine learning and constrained regression. We will also discuss the economic relevance of choosing an ‘appropriate’ set of evaluation metrics and the ‘right’ objective function, and how this choice can be accounted for in empirical methods for risk adjustment.
- Summarize recent research on risk sharing. It is well-known that risk sharing (i.e., retrospective payments based on realized spending) can be a useful supplement to risk adjustment. In this session we will discuss a series of novel risk sharing methods that target retrospective payments to patients for whom risk adjustment falls short. Examples include high-risk pooling using machine learning, residual-based (re)payments, and two-sided reinsurance. We will also discuss how risk sharing can be taken into account when estimating risk adjustment weights.
- Connect researchers interested in risk adjustment and risk sharing and identify topics for future research and collaboration. Most of the literature summarized under objectives 1 and 2 is from the field of health insurance, in which research on risk adjustment and risk sharing is most developed. Many of these topics, however, are relevant for other fields/applications too. An important goal of this session is to bring together researchers from these different fields and identify topics for future research. Three applications will be considered: risk equalization in health insurance; risk adjustment of provider payments; and risk adjustment of regional funds.
Speakers:
Professor Shuli Brammli (Hebrew University of Jerusalem)
Professor Randall Ellis (Boston University)
Dr. Lukas Kauer (University of Lucerne)
Dr. Emmanouil Mentzakis (University of Southampton)
Professor Francesco Paolucci (University of Bologna)
Professor Wynand van de Ven (Erasmus University Rotterdam)
Dr. Richard van Kleef (Erasmus University Rotterdam)
Professor Juergen Wasem (University of Duisburg-Essen)
5 July | Morning (09:00 am - 12:00 pm)
Room: GSH1.Aud2
Luigi Siciliani, University of York, UK
This course provides an overview of microeconomic models that can be used to investigate the demand for and the supply of health services. The course is aimed at empirical researchers who would like to motivate, guide and complement their empirical analyses by developing a theoretical model.
The module therefore focuses on relatively simple and tractable theoretical microeconomic models that can be used to frame an empirical research question. On the demand for health care and health, the course will describe models that i) adopt a representative consumer (individual, patient) approach, or ii) allow for heterogeneity in preferences and need (e.g. severity, health benefits, distance). Factors affecting demand will include quality, waiting times, and co-payments. Models for the demand for health will include a static version on the Grossman model, prevention, and food consumption.
On the supply side, the course will give an overview of theoretical models that describe provider incentives towards intensity of care, quality and costs under a range of assumptions (altruistic concerns, capacity constraints, non-profit status, competition) and payment systems (e.g. fixed price regulation/capitation, and pay for performance) under different specifications of the demand function. The focus will be on large organisations, such as hospitals, and to a lower extent primary care providers.
By the end of module, the student should have a clearer idea on how to set-up their own theoretical model that matches their empirical analysis, and an idea of key modelling strategies and trade-offs.
5 July | Afternoon (01:00 pm - 04:00 pm)
Room: GSH1.Aud2
Andrew Jones, University of York, UK
This workshop focuses on the principles and practical application of data visualization and how statistical graphics can enhance applied econometric analysis and policy evaluation. Applications include econometric methods for health outcomes and health care costs that are used for prediction and forecasting, risk adjustment, resource allocation, technology assessment, and policy evaluation. Methods for policy evaluation include: regression methods, doubly robust and machine learning approaches; instrumental variables and methods for panel data, including differences-in-differences, synthetic controls and extensions. Practical examples will show how these methods can be applied and the workshop outlines a grammar of graphics in Stata.
Preliminary reading for the workshop material can be found in:
Jones, A.M., “Data visualization and health econometrics”, Foundations and Trends in Econometrics, 9, 1-78, 2017. DOI: 10.1561/0800000033.
The workshop will introduce new material on policy evaluation.