Applying the Target Trial framework to routine data: pre-analysis challenges (NIHR Statistics Conference, June 2025)

<p dir="ltr">A recording of the presentation titled "Applying the Target Trial framework to routine data: pre-analysis challenges" delivered at the NIHR Statistics conference (June 2025, https://statistics-group.nihr.ac.uk/event/nihr-statistics-conference-2025) by Matthew Franklin (University of Sheffield, UK), Amy Chang (University of Sheffield, UK), and David Lugo Palacios (LSHTM, UK). </p><p dir="ltr">The presentation covers the experiences and challenges of the presenters when applying the Target Trial framework to routinely collected care data, mainly covering mental health (Matthew, NIHR Target Trials in Mental Health programme, https://target-trials-mental-health.sites.sheffield.ac.uk) and cancer (David, NIHR SORT, https://www.lshtm.ac.uk/research/centres-projects-groups/sort; Amy, ReCREATE Group) routine data.</p><p dir="ltr">12 key areas are discussed: Metadata, Directed Acyclic Graphs, Statistical Analysis Plans, Third-party data providers, sample size, benchmarking criteria, use of data extracts Vs secure data environments (SDEs), risks for PhD students, pilot analyses, timing to get data, convincing funders, and costs.</p><p dir="ltr">The presentation has been edited and shared with explicit permission by the organisers of the NIHR Statistics conference, as provided via email to Matthew Franklin on 02/07/2025. We thank the organisers of the NIHR Statistics conference for allowing us to present our work and share the recording. You can find more about the NIHR Statistics Group here: https://statistics-group.nihr.ac.uk/. </p><p dir="ltr">The recording is shared under a CC BY-NC-ND (Creative Commons Attribution-Non Commercial-No Derivatives) licence. It's a license that allows others to copy and redistribute the material in any medium or format, but only for non-commercial purposes and without modification. Specifically, it requires that the original author be credited, the material is not used for commercial gain, and no derivative works are created. </p>