Isoqol Berlin Conference
Professor Stephen Walters attended the 21st International Society for Quality of Life (ISOQOL) Annual Conference in October 2014 in Berlin, Germany and gave a 3-hour workshop on “Analysing longitudinal quality of life outcome data”.
The aim of the workshop was to be a practical guide to the analysis of longitudinal QoL data. Many studies using QoL outcomes have repeated assessments over time and are longitudinal. In a randomised controlled trial (RCT) and other longitudinal studies there may be a baseline QoL assessment and several follow-up assessments over time.
The workshop described how the QoL data from such studies can be summarised, tabulated and graphically displayed. These repeated QoL measurements, on the same individual subject, are likely to be related or correlated. This means that the usual statistical methods for analysing such data, such as the t-test and linear regression, which assume independent outcomes may not be appropriate. The workshop then showed how repeated QoL measures for each subject can be reduced to a single summary measure for statistical analysis and how standard statistical methods of analysis can then be used.
Finally, the workshop described more complex modelling approaches, based on an extension of the linear regression model which allows for the fact that successive QoL assessments by a particular patient are likely to be correlated. These included marginal (General Estimating equations) general linear models (GLMs) and random-effect (multilevel) GLMs.
The course was well attended with over 45 attendees and well received with the overall satisfaction with the workshop rated as good (29%), very good (51%) and excellent (20%) by the responders to the workshop evaluation.
Comments by the attendees included:
“Great idea to use real data/examples.”
“Clear and concise/very approachable."
“Level of math was great”
“Nice recommendations on how to present data, good examples.”
“Very good session; really nice blend of practical info, stimulating new thoughts about well used approaches, and highlighting of valuable resources.”