Hello and welcome!
My name is Martje Rave, and I am a statistician and data scientist with a PhD in Statistics from LMU Munich. I currently work at Dansk Gigthospital, where I support research in rheumatology using clinical data and Danish registry data.
A large part of my work is about helping researchers make sense of data. I contribute throughout the research process – from turning a clinical idea into a research question and planning an appropriate study, to analysing the data, interpreting the results, and making sure that the conclusions are supported by what the data actually show.
My current projects cover a broad range of questions in rheumatology. For example, I work on studies investigating the association between rheumatic diseases and cardiovascular disease, the implementation of European treatment recommendations for axial spondyloarthritis (axSpA), and the reliability of diagnosis codes in Danish registry data.
Although the projects can look very different, the underlying idea is usually the same: we have a question, we have or need data, and we use those data to find the most reliable answer we can. I am particularly interested in transparent and reproducible research and in making the strengths, limitations, and uncertainty in data as clear as possible.
My statistical background is in regression modelling, Bayesian methods, spatial and temporal data analysis, and the analysis of complex health data. During my PhD, I worked on modelling ICU occupancy and patient flows during the COVID-19 pandemic and contributed to forecasting ICU demand for the Bavarian Health Authority (LGL). I also developed methods for reconstructing incomplete patient pathways from missing ICU admission data.
I primarily work in R, but also use SQL, Python, SAS, and Power BI depending on the project.
Alongside my research work, I have experience teaching regression methodology, Bayesian modelling, and spatial statistics at Bachelor’s and Master’s level, supervising Master’s theses, and teaching seminars on Open Science and statistical modelling.