Data Science & Statistics
My work covers clinical trials, observational studies, real-world evidence, claims, patient-reported outcomes, and biomarker data. I have worked in biopharma, contract research, and real-world data organizations.
Analysis projects
I often lead projects from the initial question through the final analysis and delivery. That includes working with clinical and scientific partners to define the problem, assessing the available data, choosing methods, reviewing results, and explaining the findings and limitations.
Depending on the question, the work may lead to a reproducible report, statistical model, publication, dashboard, or reusable software tool.
Methods
My statistical and data science experience includes:
- regression modeling, including linear, logistic, Poisson, zero-inflated, and Firth regression
- longitudinal and correlated-data methods, including mixed-effects models, generalized estimating equations (GEE), and robust sandwich variance estimators
- survival and time-to-event analysis
- causal inference and propensity score methods
- subgroup and heterogeneous treatment-effect analyses
- predictive modeling and machine learning, including penalized regression and tree-based methods
- Bayesian analysis
- missing-data methods and multiple imputation
- simulation and disease-progression modeling
- dimensionality reduction, clustering, and high-dimensional biomarker analysis
- statistical visualization
- data validation and reproducible analytical workflows
Clinical and real-world evidence
Clinical-trial and observational data require different assumptions and different ways of interpreting results. I have worked with completed clinical trials, registries, claims, observational cohorts, and large biomedical datasets. Across these settings, I focus on making the analysis question, assumptions, uncertainty, and limitations clear.
Working with teams
I work closely with clinicians, biostatisticians, data scientists, and data engineering partners. In technical-lead roles, I have also helped establish practices for code review, validation, documentation, and reproducibility, and have mentored analysts and data scientists.