OE.

Healthcare
data work

Case studies and technical work built around practical questions in healthcare data.

Project archive

Read each case study here, then follow its repository link for code and supporting files.

Healthcare EHR analysis

Cleaned and explored more than 5,000 synthetic healthcare records, then built Power BI views of patient demographics, utilization and test outcomes.

Power BIExcelPower QueryDAX

This case study follows a synthetic electronic health record dataset from preparation through reporting. The analysis focuses on how patient demographics, service utilization and test outcomes can be organized into views that are easier to review.

Approach

Records were cleaned and transformed with Power Query and Excel, then explored in Power BI. DAX measures support the report views. The dataset is synthetic and does not contain real patient information.

View repository
Anatomical eye illustration from an 1806 medical text
“Eye (11291008035),” University of Liverpool Special Collections, CC BY-SA 2.0

Hospital operations

SQL Server analysis of hospital operations, appointments and patient utilization, framed around questions healthcare teams can act on.

SQL ServerHealthcare operations

This project uses SQL Server queries to explore a hospital management dataset, including appointments and patient utilization. The analysis is framed around operational questions that can help teams decide what to investigate further.

See the repository for the query work and the project's source details. Interpretations should be read with the dataset's scope and limitations in mind.

View repository
A page of nineteenth-century anatomy text and eye development diagrams
Quain’s Elements of Anatomy, 1882 · Internet Archive Book Images, public domain

Clinical data quality concept

A proposed SQL and Python workflow for finding missing fields, inconsistent codes and duplicate encounters before reporting. This concept is not yet a published project.

SQLPythonData quality

This is a proposal for a future project, not a completed analysis. The intended workflow would check missing required fields, inconsistent codes and duplicate encounters before reporting, then provide an exception list for review by a clinical data owner.

Optometry
data check

Run a fictional eye-care visit through a small completeness and identifier check.

Sample eye-care visitReady to check
Patient ID
PT-2048
Encounter date
2026-06-18
Diagnosis code
H52.13
Lens prescription
−1.50 / −1.25 D
Follow-up
2026-07-16

Select a sample, then check its required fields and identifier.