Diagnosis Classification
Patients are classified as breast cancer, colon cancer, or a combined category when both diagnoses are present.
← Back to selected work
Featured Case Study · R
Oncology Treatment Analytics
Transforming synthetic patient-level demographics, diagnoses, and treatment records into structured, review-ready clinical insights.
Overview
From raw records to an interpretable patient journey.
The project joins demographic, diagnosis, and treatment data at the patient level, applies documented derivation rules, and produces publication-ready summaries for clinical review.
Time to Treatment
The workflow derives the interval between the earliest diagnosis date and first observed treatment date.
Regimen Inference
Rule-based treatment episodes identify the first-line regimen from temporally grouped drug starts.
Baseline Summary
Age, sex, stage, and region are summarized by first cancer diagnosis using publication-ready tables.
Data quality review
Analytical outputs include the exceptions—not only the clean results.
Negative time to treatment
A treatment preceding the earliest recorded diagnosis was flagged as a potential incomplete-history or date-consistency issue.
Missing treatment record
A diagnosed patient without an observed treatment was documented for follow-up rather than silently excluded.
Traceable definitions
Each derived outcome is paired with a written definition and visible R implementation.
RtidyverselubridategtsummarygtQuarto