Application of the Benefits Analysis Tools for MRC population health studies

1. Application of the Benefits Analysis Tools for MRC population health studies • Professor Dipak Kalra • Centre for Health Informatics and Multiprofessional Education • (CHIME) • University College London • d.kalra@ucl.ac.uk

2. Background:MRC Data Support Service (2009-11) • Funded by MRC to better understand the different approaches to data sharing across its population health studies • identify drivers • identify approaches • identify barriers • provide support to studies where needed • contribute to a review of MRC policy and to future policy guidance from MRC • 6 exemplar studies were investigated in detail, working closely with study Directors, PIs and data managers • Today’s review of the benefits tools draws on all six, and from less detailed review of other population health studies • (a kind of virtual study based on aggregated and anonymised insights)

3. MRC study data sharing context • (Not like a central data archive) • Each MRC study: • defines the overall research mission • might have other funding parties e.g. CRUK, international • undertakes the data collection • has to maintain cohort relationships for ongoing collection • is responsible for maintaining confidentiality • designs the data schema • documents some metadata, but the depth needed for shared use is often complex • undertakes most of the data cleaning and derivations • analyses and publishes its own research on the data • shares voluntarily, as a complement to its core mission • The wide range of study types and their maturity means the opportunities and benefits will vary between studies

4. Stimulating new networks and collaborations • Localised expression: Enriched quality and scope of grant applications, greater chance of funding and/or a larger award • Action: Form collaborations with other research groups for joint applications for funding and/or share resources from individual grant awards • KRDS Outcome type: Direct benefit, in 2-4 years • Stakeholders: Internal: PIs, academics; External: grant funders • Quantitative benefit: Larger scale of research funding through larger grant or pooling of grant incomes • Qualitative benefit: Increased visibility and kudos • Weighting: 3

5. Re-purposing and re-use of data • Localised expression: Lower costs of data collection and curation, faster route to usable data • Action: Agree shared use of pre-existing data held by another study to avoid new data collection • KRDS Outcome type: Indirect benefit, in 1 year • Stakeholders: Internal: PIs, academics, data managers • Quantitative benefit: Cost saving, shorter time to publications • Qualitative benefit: ~ • Weighting: 4

6. Conclusions • The Benefits Framework and the Impact Tool can accommodate the kinds of benefit from good data curation practice and from data sharing in MRC population health studies • Detailing the active steps to realise each benefit, when the result might be realised and who benefits seem to be useful ingredients for putting forward a case for funding or for prioritising resource utilisation with a study • Whilst initial population might be done by one person, completing the spreadsheet and working out weightings might be nicely undertaken in a team workshop