Biostatistics Team at the Department of Global Public Health

Research 

Dose-Response Meta-Analysis

Development of statistical models and related software for model selection, visualization of marginal and conditional estimated dose-response relationships, assessing goodness of fit, involving multiple studies of either individual or summarized data.

Missing data and multiple imputation

Current development, evaluation, and implementation of flexible two-stage imputation approaches that can handle missing data in multicenter studies (pooling projects) and federated analyses where individual-level data between study sites cannot be shared.

Education & Training

We are leading multiple Master and PhD courses in biostatistics at KI including classes on programming in Stata and Python. We also conduct seminars and workshops during the year that are attended by students, researchers, and faculty members across KI. In all teaching activities we aim to involve innovative teaching methods for which we have received funding from the KI pedagogical research fund.

Karolinska Institute’s Pedagogical Project Funding 2025 (FoUI-1004095) on “Simulation-based teaching and learning statistics in educational programs”. Main applicant: Nicola Orsini. Co-applicants: Robert Thiesmeier and Karin Båge. 

Software & Tools

Documentation and dissemination of our research is an integral part of our work. Multiple software packages have been developed and are accessible for download in Stata and R.

Collaboration & Support

We collaborate with multiple other groups at KI and international partners in Denmark, UK, Tanzania, and the USA.

We host open office Biostatistics Corner for students at the Department of Global Public Health for questions and queries.

We are also part of the SFO Biostatistics Seminar Series that organizes monthly seminars for biostatisticians and epidemiologists by inviting prominent speakers from around the world.

Selected references

Research articles

• Thiesmeier, R., Hofer S., Orsini, N.  Multiple imputation for systematically missing effect modifiers in individual participant data meta-analysis. Statistical Methods Medical Research. 2025. In press. doi: 10.1177/09622802251348800
• Thiesmeier, R., Madley-Dowd, P., Orsini, N., & Ahlqvist, V. H. Cross-site imputation can recover missing variables in federated multi-centre studies. Journal of Clinical Epidemiology 2025. 111820. doi: 10.1016/j.jclinepi.2025.111820. PMID: 40348146
• Thiesmeier R,  Bottai M, Orsini N. Systematically missing data in distributed data networks: multiple imputation when data cannot be pooled. Journal of Statistical Computation and Simulation 2024 94;17 3807-3825. DOI10.1080/00949655.2024.2404220.  
• Thiesmeier, R., Bottai, M., & Orsini, N. (2024). Imputing missing values with external data. arXiv preprint arXiv:2410.02982.

Educational articles

• Orsini, N., Thiesmeier, R., & Båge, K. (2024). A Simulation-Based Approach to Teach Interaction Effects in Postgraduate Biostatistics Courses. Journal of Statistics and Data Science Education, 32(4), 395–404. https://doi.org/10.1080/26939169.2024.2394536
• Thiesmeier R, Orsini N. (2024) Rolling the DICE (Design, Interpret, Compute, Estimate): Interactive Learning of Biostatistics With Simulations. JMIR Med Educ.  Apr 15;10:e52679. doi: 10.2196/52679. PMID: 38619866; PMCID: PMC11058551.
• Thiesmeier, R., Orsini, N., (2024). Teaching Statistics in Health Sciences: The Potential of Simulations in Public Health. CHANCE, 37(2), 34-39.

Contact