Clinical and population evidence
I started with clinical records because that is where care often leaves its trace: visual acuity entries, diagnoses, referral decisions, and incomplete documentation. The work then widened into provincial access analytics, where the same discipline is applied to health-system questions.
A system-level access study that frames health care as a prioritization problem: which regions, populations, and service areas show the strongest signals of friction, and what changes when the weighting model changes?
The foundation layer: recoding visual acuity, diagnoses, and referral fields into a record that can support analysis without hiding uncertainty. This is the unglamorous work that makes every downstream claim safer.
After the record is cleaned, the work moves from rows to population signal: what conditions appear most often, how visual impairment is distributed, and where outreach data can guide clinical attention.
A move from description to action: using clinical findings to structure referral risk, triage logic, and follow-up decisions for outreach settings where the next step matters as much as the diagnosis.
A longitudinal outcomes track for glaucoma surgery: pressure, acuity, and survival endpoints prepared as evidence rather than presentation — cleaning first, outcomes second, survival modeling last.
A digital-health readiness assessment that connects informatics operations to adoption maturity: where the record system is strong, where the workflow still depends on manual work, and what stage progression would require.