Health economic evaluation studies: how to deal with missing data

Written by Jonathan Bartlett on . Posted in News

Report from one of the sessions at the RSS 2017 Conference. More reports of conference sessions are listed here.

This invited session at the RSS 2017 Conference on missing data in health economic evaluation studies was organised by Rumana Omar (UCL) on behalf of the RSS Medical Section. The speakers were Gianluca Baio (UCL), James Carpenter (UCL & LSHTM) and Andrew Briggs (Glasgow University).

Andrew Briggs began the session, explaining reasons why missing data are particularly problematic in health economic evaluation studies. He noted that economic sections of case report forms may sometimes be given less priority, leading to higher rates of missing data in variables required for economic evaluations compared to a trial’s primary outcome variable. He highlighted that a number of principled methods for handling missing data are being increasingly used, such as inverse probability weighting in longitudinal analyses, alongside causal inference methods which can (under assumptions) correct for treatment switching. Importantly, he noted that such methods are gaining increased acceptance by payer authorities in a number of countries.

Gianluca Baio then presented on an overall Bayesian approach to handling missing data in health economic evaluation studies. He highlighted the importance of accounting for the correlation between cost and benefit variables in such an analysis, something which is automatically handled in the Bayesian approach. Gianluca gave particular focus to describing and proposing the use of flexible parametric models which are able to accommodate some of the common complicating features of health economic variables, such as skewed costs, or a large proportion of patients taking on the lowest value of a quality of life score.

James Carpenter then presented a framework for conducting missing not at random sensitivity analyses using Bayesian methods. James explained a key part of this is the process of elucidating priors from subject matter experts, and in particular being able to frame statistical parameters in such a way that collaborators are able to give informed judgements about their beliefs. To this end, James demonstrated a web based prior elicitation tool custom developed for a trial of aneurysm, in which the experts are presented with fictional patients who differ in terms of data being missing or not.

This RSS 2017 Conference session was titled: Health economic evaluation studies: How to deal with missing data

Jonathan Bartlett is a statistical science director in AstraZeneca’s Statistical Innovation Group in Cambridge. He was previously lecturer in Medical Statistics at the London School of Hygiene & Tropical Medicine. He runs a blog at www.thestatsgeek.com and maintains www.missingdata.org.uk

 

 

Medical section RSS Conference

Join the RSS

Join the RSS

Become part of an organisation which works to advance statistics and support statisticians

Copyright 2019 Royal Statistical Society. All Rights Reserved.
12 Errol Street, London, EC1Y 8LX. UK registered charity in England and Wales. No.306096

Twitter Facebook YouTube RSS feed RSS feed RSS newsletter

We use cookies to understand how you use our site and to improve your experience. By continuing to use our site, you accept our use of cookies and Terms of Use.