Background:
Management and prevention of chronic diseases require high-quality surveillance. In an era of big data and infodemia, several challenges hamper efficient surveillance, such as the definition and measurement of chronic diseases and how to best use data that are not designed primarily for surveillance. Using three cases studies, we propose a workshop to discuss current challenges in the surveillance of chronic diseases.

Objectives and key questions:
Our objective is to discuss current challenges in the surveillance of chronic diseases. Specifically, the workshop will address these key questions: 1) how to define chronic conditions? 2) how to prevent surveillance bias?

Brief overview:
We will have one short (10 min) introduction and 3 case studies (each 10 min) presentations followed by an active discussion session (20 min): Introduction (Arnaud Chiolero and Cornelia Wagner) Public health surveillance is the ongoing collection and analysis of health-related data, followed by the timely dissemination of information useful for decision. In an era of big data and infodemia, it might appear easy to conduct the surveillance of chronic diseases. However, multiple challenges hamper efficient surveillance, notably related to changing definitions and diagnostic methods of chronic diseases and to the quality of data from multiple sources not designed primarily for surveillance. Case Study 1: Defining Multimorbidity in High-Income Countries (Piotr wilk and Saverio Strange) In 2012, the Public Health Agency of Canada defined multimorbidity as having two or more of ten common chronic diseases, selected based on their duration, high prevalence, significant societal or economic impact, and amenability to primary prevention. In 2017, multimorbidity was defined as the co-occurrence of a least two of five groups of chronic diseases (cancer, diabetes, cardiovascular disease, chronic respiratory disease, mental illnesses). We will discuss the implications of the two definitions of multimorbidity on prevalence estimates and geographic distribution across Canada. Case Study 2: Self-Reported data for the Surveillance of Chronic Diseases (Arnaud Chiolero) Self reported data are commonly used for population health monitoring. Using data of a large school-based study and through a literature review, we will discuss strength and limitation of self-reported data for the surveillance of overweight. We will extend the discussion toward the pros and cons of surveillance of other chronic conditions based on self-reported data. Case Study 3: Surveillance Bias of Cancer – when Appearances are Misleading (Stefano Tancredi) Surveillance bias occurs when differences in the frequency of a condition are due to variations in the modalities of detection rather than to changes in the actual risk of the condition. This is of growing concern because surveillance activity is more and more often based on data not designed primarily for surveillance, notably from healthcare providers. We will show the impact of this bias on cancer surveillance.

Discussion:
(Cornelia Wagner and Arnaud Chiolero)