Methodological Issues

There are important methodological issues to address if we are to achieve effective research in SuPaC. The major methodological shortcoming of previous research has been small and unrepresentative samples studied. One of the major benefits of this UK wide collaborative is that we will be able to study large samples representative of patients throughout the UK and also take account of stage of illness, setting of care and ethnic and cultural variations. This part of our work will depend on the enhanced support provided by the methodology strand. The COMPASS statisticians have expertise in qualitative methods; epidemiology, complex interventions, trials and implementation of organisational change, and will ensure high methodological quality in all three strands of research.

Issues to be addressed are:

• Comparing and standardising measures,
• Analysis of longitudinal data,
• Imputation of missing data and response shift and study design and analysis where there is a high rate of attrition and missing data