Population Scientist · Kepka Group · HCI · He/Him
Some questions can't wait for graduate school.
Engineering mindset
The work I want to do needs two things at once: the analytic discipline to model a problem honestly, and the writing to make the answer land with the people who can act on it.
I try to keep both running — rigorous regression and a clear sentence, a careful study design and a human story. The most useful research I've seen refuses to treat that as a trade-off.
Odds ratios, DAGs, logistic regression, R, Python.
CBPR, community surveys, rural outreach. Also a 5k.
Next.js, Python, R — tools chosen for the problem.
5th-grade Mancala champion, Willowcreek Middle School.
Research
Working within the Kepka Group's mission: driving quality improvement and expanding access to cancer prevention for underserved populations.
Grants
Working on a proposal? Want a second read on the strategic framing? Get in touch →
Mission statement
I want to close the distance between what we already know how to prevent and the people still living without that protection. So much harm in public health isn't from what we haven't discovered — it's from what we've discovered and failed to deliver evenly: a cancer-preventing vaccine that never reaches a rural county; a clear message never written in the language someone actually thinks in. My part is small and specific right now — learn the methods well, study communities honestly, and treat the people closest to a problem as collaborators, not data points. Rigor and warmth aren't a trade-off. Prevention done right is invisible; the cancer that never happens leaves no headline — and I'm okay working on things no one will ever have to thank me for.
Heroes
Some I've met. Some I'm working on. If you're on this list and reading this — hi.
Stay in the loop
I help researchers tighten the technical side — study design, regression, sensitivity analyses — to give good work its best shot at acceptance. If that's useful to you, I'd love to hear from you.
Get in touchImproving uptake of the HPV vaccine — a cancer-prevention tool — among rural and underserved communities across the Mountain West, the domain of the coalition Dr. Deanna Kepka founded and leads.
Rural communities face higher cancer mortality and lower screening uptake than urban areas; this area studies why, and what closes the gap.
CBPR is an equitable approach where researchers and community members collaborate on every stage of a study — a core method for reducing cancer disparities and the backbone of culturally tailored outreach in the Mountain West.
Where people live shapes whether they can get screened; this area uses GIS and area-level vulnerability indices to map and explain place-based gaps in cancer prevention.
DHK · University of Utah College of Nursing; Huntsman Cancer Institute, Cancer Control & Population Sciences. Tenured associate professor who leads the 500-member Mountain West HPV Vaccination Coalition and HCI's inaugural Global Oncology Initiative (2025). She showed me what it looks like to meet communities in their own idiom — including Spanish-language fotonovela and radionovela interventions — instead of waiting for them to come to the data.
CPJ · physician-epidemiologist and past president of the American Public Health Association. For teaching a generation to name how social structures, not just individual choices, shape who gets sick.
OB · Bloomberg Distinguished Professor, Johns Hopkins. For decades of clear-eyed writing on cancer disparities and the gap between what medicine can do and who actually receives it.
SG · dean and epidemiologist (Washington University in St. Louis [confirm current institution]). For insisting that population health is about the conditions that make people healthy, and for writing about it for the public, not only the journals.
HR · the late physician and global-health communicator behind Gapminder. For proving that rigorous data and unforgettable storytelling are the same craft when done well — the both-hemispheres ideal I keep chasing.
Reviewers accept manuscripts they can trust, and trust comes from a clearly specified causal question, a defensible design, and honest sensitivity analyses—not from a bigger dataset or a flashier model.
The odds ratio is the workhorse of logistic regression, but it routinely overstates how much an exposure matters when the outcome is common. Here is how to read it honestly.