James Carpenter, Professor of Medical Statistics

James leads the Methodology Analysis programme and has interests across the Methodology theme.

The motivation for his research is finding practical methodological solutions to challenges in Phase III clinical trials and observational research, principally:

  • Choosing the appropriate estimand for a trial (i.e. the patient population, the quantity we wish to estimate and how we handle post-randomisation events in the analysis)
  • Missing data: in particular, practical approaches for exploring and communicating the sensitivity of trial results to a range of appropriate assumptions about the missing outcome data
  • Statistical methods in meta-analysis
  • Using routinely collected health care data to inform trial outcome measures.

James works at the MRC Clinical Trials Unit at UCL on a 50% secondment from the Department of Medical Statistics at the London School of Hygiene & Tropical Medicine.


Selected publications

Cro, S., Carpenter, J. R. and Kenward, M. G. (2018). Information-anchored sensitivity analysis: theory and application. Journal of the Royal Statistical Society, Series A. https://doi.org/10.1111/rssa.12423

Freeman, S. C., Fisher, D., Tierney, J. F. and Carpenter, J. R. (2018) A framework for identifying treatment-covariate interactions in didividual participant data network meta-analysis. Research Synthesis Metods. htpps://doi.org/10.1002/jrsm.1300

Mason, A. J., Gomes, M., Grieve, R., Ulug, P., Powell, J. T. and Carpenter J. R> (2017) Development of a pratical approach to expert elicitation for randomised controlled trials with missing health outcomes: application to the IMPROVE trial. Clinical Trials, 14(4), 357-367. https://doi.org/10.1177/174077451771144

Fisher, D. J., Carpenter, J. R., Morris T. P., Freeman, S. C. and Tierney, J. F. (2017). Meta-analytical methods to identify who benefits most from treatments: daft, deluded, or deft approach? BMJ 356:j573. https://doi.org/10.1136/bmj.j573 

Carpenter, J. R. and Kenward, M. G. (2013) Multiple imputation and its Application. Wiley (Chichester). https://doi.org/10.1002/bimj.201300188





Research Interests

  • Analysis of data with missing observations (both outcomes and covariates), and in particular the method of multiple imputation
  • Practical approaches for exploring the robustness of trial inferences to different assumptions about the distribution of missing values
  • Statistical methods for meta-analysis and network meta-analysis
  • Estimands, and the ICH-E9 addendum
  • Use of routine electronic health records to facilitate phase three trials in secondary care settings
  • Novel phase three designs


Research Areas



Related Studies



UCL Profiles