Event

Please note the change of venue to Seminar room 515, Level 5, Melbourne School of Population and Global Health, 207 Bouverie St Carlton.

In this workshop, we provide a brief, non-technical introduction to the basic steps of clinical risk prediction model research, from gathering data, developing the model, internally and externally validating the predictive performance of the model, to assessing the clinical utility of the model and implementation in clinical practice.

Next, we discuss some common pitfalls and myths regarding risk prediction models. We pay particular attention to calculating the sample size required to develop or validate a model, to handling missing data, and to measuring model performance. We address the role of shrinkage and machine learning in modern clinical prediction modeling.

To conclude, we provide a birds-eye overview of the current clinical risk prediction modeling landscape in the medical literature.

Presented by Laure Wynants

Dr Wynants is assistant professor of Epidemiology at Maastricht University, and is a prominent member of one of the world’s leading centres in the development, validation and impact of clinical predictive models for diagnosis and prognosis.

She is interested in methods to deal with heterogeneity between populations, and in the clinical utility of prediction models. Her applied work deals with models for gynaecological cancers, bloodstream infections, and COVID-19. Her systematic review of prediction models for diagnosis and prognosis of Covid-19 (BMJ, 2020) has already been cited more than 2000 times.

Read more about Laure's work here: https://cris.maastrichtuniversity.nl/en/persons/laure-wynants

Target audience:

This course has been developed with epidemiologists and clinical researchers in mind as a rapid way to learn the cutting-edge issues and latest approaches to model development and application.

For further extension and for those already familiar with clinical prediction modelling, a 2-day risk prediction course will be run as part of the ViCBiostat Summer School (February 2024). This course will go in to greater detail and include practical, hands-on software demonstrations.

Required background knowledge:

Attendees will require a general familiarity with regression models, such as obtained through a Master of Public Health. It is assumed attendees know how to fit a logistic regression model and how to obtain predicted probabilities from the model, and have a basic understanding of measures of diagnostic accuracy (sensitivity, specificity, area under the ROC curve).

Schedule

Ticket prices

Standard: $390 inc. GST

Student: $310 inc. GST

Tickets are available now via https://www.trybooking.com/CKRKC

MCRI employees are required to organise payment via an internal funds transfer. Please contact to arrange this. You will be provided with a code to register once funds have been received.