Pre-conference courses will take place on Tuesday 29th April (daytime).
Courses will cost £50 each and will run in parallel.
The courses available are listed below:
Short Course 1
An Introduction to Clinical Prediction Models and Sample Size Calculations for Model Development & Evaluation
Clinical prediction models are used to estimate an individual’s risk of a health-related outcome to help guide patient counselling and clinical decision making. Examples include QRISK, which is widely used in the UK during primary care consultations to estimate a person’s 10-year cardiovascular disease risk.
In this course, participants will be provided with an introduction to clinical prediction models and how to improve standards, which currently are often poor. A pathway will be described from model development to model evaluation and impact assessment.
We then focus primarily on sample size calculations for model development and evaluation. We introduce the theory behind existing and emerging approaches, and showcase the pmsampsize, pmstabilityss and pmvalsampsize packages in Stata and R that implement them. Hands-on practicals are included, and participants will be supported by a dedicated and experienced faculty. Participants are required to bring their own laptop with either R or Stata installed.
Faculty includes: Dr Joie Ensor, Dr Kym Snell, Dr Lucy Archer, Dr Becky Whittle, Dr Amardeep Legha and Prof Richard Riley from University of Birmingham
Short Course 2
Systematic reviews of prognosis studies
Prognosis studies are abundant in this era of personalized and precision medicine, which all has to do with prognosis research. Hence, systematic reviews of prognosis studies are increasingly required and conducted.
This workshop will introduce participants to the different types of prognosis research and explain the differences between prognosis, diagnosis and intervention studies. We then provide explicit guidance how to define a proper review question using the PICOTS format, to design a review protocol, to search the literature, to extract the data using the CHARMS checklist, to assess the risk of bias in the primary studies using PROBAST+AI and QUIPS, and finally to conduct a meta-analysis of the retrieved data and test for heterogeneity across studies. We will illustrate all this using many empirical examples, and frequently apply small group practicals and discussions.
After this workshop, participants have a good overview of the essentials of prognosis research and systematic reviews of such studies.
Faculty includes: Dr Anneke Damen, Prof Karel Moons from UMC Utrecht
Short Course 3
The Potential and Pitfalls of Predicting Treatment Effects
Many medical interventions are effective on average but ineffective or even harmful for a substantial proportion of patients. The challenge of Precision Medicine is to tailor medical treatment to the individual characteristics of each patient, concentrating interventions on those who will benefit, sparing expenses and side effects for those who will not. The usual method for identification of patients who are likely to benefit – one-variable-at-a-time subgroup analysis – has well-known statistical limitations and limited ability to inform decisions for individuals. Counterfactual prediction models – models that predict outcomes for a particular patient under two or more possible interventions taking multiple patient characteristics into account simultaneously – are a promising alternative but have many pitfalls.
In this half-day course, participants will be provided with an introduction to counterfactual prediction models that aim to predict treatment effects. We will first review fundamental concepts necessary for understanding heterogeneity of treatment effects (HTE). We contrast two forms of predictive approaches to HTE, risk modelling and effect modelling, using examples from the literature. We then focus on practical methods for developing models for predicting treatment effects (with R software), including (penalized) regression and machine learning techniques. Finally, we will focus on the evaluation of model performance.
Faculty includes:
Prof. David M. Kent from the Predictive Analytics and Comparative Effectiveness Center, Tufts Medical Center in Boston
Dr. David Van Klaveren, Department of Public Health, Erasmus MC University Medical CenterThe course can be selected during conference registration and are only open to those registering for the conference. Please note that only one course should be selected as the three options will run in parallel. Limited places available.
For more information, please contact the MEMTAB Event Management Team on academic.conferences@contacts.bham.ac.uk.