Team Responsible Health Data Analysis

The Responsible Health Data Analysis (RHDA) Team works to improve the quality, transparency, and rigor of health data analysis. We support researchers in formulating clear research questions, applying suitable methods, and transparently reporting interpretable findings.

We emphasize the importance of making the research question precise, understanding how the data were generated, and making explicit under which assumptions the conclusions are valid across all tasks of data science (descriptive, predictive, and causal). To achieve this aim, we combine quantitative methods with domain expertise to guide all steps of the analysis process.

Causal inference is a particular focus of our work since most questions in health research are fundamentally causal in nature. When randomized controlled trials are not feasible for ethical or practical reasons, careful analysis of observational health data is essential for informing clinical and policy decisions. In the era of electronic health records, open data and record linkage, causal inference tools are crucial to draw correct conclusions from observational data.

In addition to applied research projects, we conduct meta-research on how specific methods are used, reported, and interpreted in practice. A key pillar of our work involves the translation of cutting-edge methodological developments into applied work. By providing tailored, field-specific examples, we seek to make common methodological issues visible, raise awareness, and promote uptake of innovative methods within research communities.

We believe that an important way to improve quality in health research is to train peer-reviewers and journal editors to scrutinize aspects of methodology and promote rigorous examination practices. We offer Peerspectives workshops, designed to guide early-career researchers through their first peer reviews in a mentored environment. With support from the Berlin University Alliance’s Objective 3: Advancing Research Quality and Value, we recently launched the Berlin Editors Exchange Network (BEE-NET), a platform for editors of all disciplines to discuss current challenges and best practices in scientific editing.

For general inquiries pertaining to our research and activities, please contact RHDA-Team@charite.de

Current Projects

Einstein Circle: Causal Inference from Observational Health Data

Peerspectives

Berlin Editors’ Exchange Network