Nathaniel D. Phillips, PhD

Healthcare Decision Scientist

I design transparent analytics tools that enable people to make better decisions, faster. I do this by combining expertise in real-world data, quantitative methods, and the psychology of decision making.

I've worked at (or for) healthcare organizations ranging from health technology start-ups to big pharma.

My colleagues say nice things about me

"Nathaniel brings technical expertise, reliability, and strategic communication that makes him an exceptional partner on complex real-world evidence projects."

Melyssa Minto

Melyssa Minto, PhD

Computational Biologist, Flatiron Health

Flatiron Health

"Nathaniel is thoughtful, rigorous, and receptive to different perspectives — I would absolutely recommend him as a strategic collaborator."

Dr. Jacob Kannarkat

Dr. Jacob Kannarkat

Physician, Yale University School of Medicine, Talkiatry

Talkiatry

"Nathaniel brings a rare combination of deep thinking, genuine work ethic, and a passion for helping teammates grow."

Veronica Pessino

Veronica Pessino, PhD

President, MBF Bioscience

My work, by the numbers

250M+

Patient records analyzed across EHR, claims, and survey data

500+

Students and professionals trained in data science and decision-making

10+

Organizations supported, from Big Pharma to early-stage startups

8+

Disease areas spanning oncology, neurology, and psychiatry

5+

Peer-reviewed publications in decision science, psychology, and health outcomes

5+

International conferences presented at, including INFORMS, Judgment and Decision Making, and rstudio::conf

My impact on medical decision making

Peer-reviewed clinical research that used my open-source software to build interpretable decision tools for real patients. See the full list on Google Scholar.

An example Fast and Frugal Tree built with FFTrees that diagnoses heart disease from a handful of patient measurements, alongside its ROC curve and classification performance.
A sample Fast and Frugal Tree built with FFTrees diagnosing heart disease from just three patient measurements — transparent logic, clinical-grade accuracy, and a decision anyone can explain at the bedside.

PLOS ONE · 2019

Age at diagnosis, but not HPV type, is strongly associated with clinical course in recurrent respiratory papillomatosis

A cohort of 339 patients analyzed with Fast and Frugal Trees identified age ~5 as the critical threshold demarcating aggressive from indolent disease.

Buchinsky et al.

Read paper

Int. J. Radiation Oncology · 2020

Correlating dose variables with local tumor control in stereotactic body radiation therapy for early-stage non-small cell lung cancer

A multi-center modeling study of 1,500 SBRT treatments used Fast and Frugal Trees alongside logistic regression to identify which dose metrics best predict tumor control.

Klement et al.

Read paper

Resuscitation · 2022

Early risk stratification for progression to death by neurological criteria following out-of-hospital cardiac arrest

Across 1,569 comatose cardiac arrest patients, a Fast and Frugal Tree reached 87% sensitivity and 81% specificity — matching ridge regression while remaining bedside-interpretable.

Coppler et al.

Read paper

Contact Nathaniel

Get in Touch