Enabling personalised cohort studies from large repositories of clinical practice data.

While clinicians provide treatment decisions and prognoses for individual patients, traditional models of evidence are mostly based on averages over potentially heterogeneous populations, and often fail to represent patients with common comorbidities or medications. Conventional subgroup analyses are ineffective because they tend to subdivide populations based on pre-specified single characteristics. Since there are many circumstances under which treatments may work for some patients but not for others, this lack of personalisation results in inefficient and potentially unsafe care.

We work on the extension and validation of methods that can leverage the information contained in large, routinely collected health datasets, by enabling personalised cohort studies on demand. Findings from these studies can be used to crowdsource prioritisation of clinical questions in need of further evidence. They can also provide an avenue to discuss and support treatment recommendations at the point-of-care.

Speakers:
Dr Blanca Gallego Luxan (B.Phys, M.Sc., Ph.D.): Dr Gallego leads the Health Analytics research program at the Australian Institute of Health Innovation, Macquarie University. Her group generates new methods that enable the use of routinely collected health data to improve and transform healthcare delivery. Trained as a physicist, she obtained a PhD in modelling and predicting climate dynamics from the University of California, Los Angeles (UCLA). She later on moved to Australia where she worked on environmental economics before joining the Australian Institute of Health Innovation in 2006. Dr Gallego has extensive international research experience in data analysis and computational modelling. She has been the recipient of three research awards, two research fellowships, and over $4 million in funding.

Dr Thierry Wendling (Pharm.D., Ph.D.): Dr Wendling graduated in 2012 as a pharmacist at the Faculty of Pharmaceutical and Biological Sciences of the University Paris Descartes (France), where he also completed a Master in Pharmacology. He did his Master Thesis in the department of Clinical Pharmacology at Novartis Institutes for Biomedical Research (NIBR) in Basel (Switzerland), during which he studied the pharmacokinetics of a new intramuscular formulation of an antipsychotic drug. He then graduated with a PhD in Pharmacometrics at Manchester Pharmacy School of the University of Manchester (UK). During his PhD (2013-2015), he worked together with NIBR on the development and application of methods for robust mechanistic (i.e. knowledge-based) modelling of clinical pharmacology data. Since March 2016, Thierry is a post-doctoral researcher in Health Analytics at the Australian Institute of Health Innovation at Macquarie University (Australia). His research interests are now in the fields of evidence-based and personalised medicine, pharmacoepidemiology, machine learning, and clinical informatics. He has been focussing on developing statistical methods that can analyse high-dimensional data from electronic health records to accurately estimate the benefits or harms of medical interventions specifically for individual patients.