John Schrom

I'm a machine learning engineer / data scientist / epidemiologist, generally working in the biomedical domain. Below you'll find some of the (public) work I've been up to over the past 10 years. If you're interested in connecting, please email me at or connect with me on LinkedIn.

Principal Data Scientist, One Medical (2017 - present)

Population Health

The utility of population health turns on its ability to improve health outcomes without increasing costs or provider burnout. This project has involved using machine learning and data science to optimize how data is communicated to providers, with the goal of improving patient health outcomes without overburdening clinicians.


  • Schrom JR, Slam A, Liu TK, Bouey C, Berk L, Gilmore A, Lesser L, Behal R. The Impact of Data Communication Style in Quality Reports on Depression Screening in Primary Care. Pending podium presentation at: American Medical Informatics Association Summit: March 2020; Houston.

Program Evaluation & Targetting

One promise of machine learning in healthcare is the ability to target interventions to the patients most likely to benefit. This body of work has involved the evaluation of the effectiveness of existing technological and clinical programs, and the use of machine learning to identify and target eligible patients with programs that will impact their health outcomes.


  • Joshi V, Schrom JR, Munkittrick K, Stearns C, Ivanova S, Hoang D, Lesser L, Fakhouri T, Diamond A. Design and Implementation of an Electronic Survey for Follow-up of Acute Conditions in Primary Care. Podium presentation at: American Medical Informatics Association Meeting: November 2019; Washington DC. [abstract]

  • Schrom JR, Patterson K, Gillmore A, Cohen P, Lesser L. Group Visits Improve Symptoms and Lower Utilization in Primary Care Patients with Anxiety. Poster presented at: Academy Health Annual Research Meeting; June 2018; Seattle. [abstract, poster]

Decreasing EHR Burden

The proliferation of electronic health records (EHRs) has led to an increase in physician burnout, largely attributed to the administrative burden of documenting in EHRs. Machine Learning presents a unique opportunity to improve provider user experience and ameliorate this burden.


  • Schrom JR, Joshi V, Bouey C, Gilmore A. Associating Chief Complaints with Electronic Health Record Activity to Decrease Provider Administrative Burden. Poster presented at: American Medical Informatics Association Meeting: November 2019; Washington DC. [abstract]

  • Schrom JR, Cohen P, Krisch S, Fakhouri T. Modifying the order of medication search results in an electronic health record to increase physician generic prescribing behavior. Poster presented at: American Medical Informatics Summit: March 2019; San Francisco. [abstract, poster]

  • Ingram P, Srinivasan R, Grennan P, Schrom JR. Identifying Non-Clinical Patient Messages Using Naive Bayes. Poster presented at: American Medical Informatics Association Meeting: November 2018; San Francisco. [abstract, poster]

Senior Data Scientist, Grand Rounds (2016 - 2017)

Physician Quality Measurement

Physician quality varies greatly across populations, and patient outcomes can vary based on the quality of the physician they see. This project revolved around building machine learning models in service of identifying high-quality physicians.

Senior Data Scientist, Practice Fusion (2013 - 2016)

Anomaly and Trend Detection

Using techniques borrowed from data mining and statistical process control, I built models to identify trends and anomalies in disease among the overall population as well as specific subpopulations. These were used for both specific conditions (e.g., identifying flu outbreaks) as well as for identifying trends among all conditions.

Clinical Data Warehouse

This project involved creating data models and the subsequent ETLs for the clinical data warehouse. This data warehouse went on to be used to power the life sciences and clinical applications. Once the core ETL had been developed, I used machine learning techniques to quantify the quality of the dataset, and impute missing or incorrectly coded data.

Research Assistant, University of Minnesota (2012 - 2013)

Clinical Phenotyping

Using data from electronic health records, I worked on identifying clinical phenotypes -- or unique manifestations of diseases with distinct treatment efficacies and outcomes -- for diabetic patients using data mining and epidemioogical techniques.


  • Schrom JR, Caraballo PJ, Castro MR, Simon GJ. Quantifying the Effect of Statin Use in Pre-Diabetic Phenotypes Discovered Through Association Rule Mining. AMIA Annual Symposium Proceedings. 2013;2013:1249-1257. [paper]

  • Simon GJ, Schrom JR, Castro MR, Li PW, Caraballo PJ. Survival Association Rule Mining Towards Type 2 Diabetes Risk Assessment. AMIA Annual Symposium Proceedings. 2013;2013:1293-1302. [paper]

Data Analyst, Hennepin County Medical Center (2010 - 2012)

HIV Care Quality & Social Determinents

I worked with providers at Minnesota's largest HIV clinic to measure and close care gaps related to our patients' HIV care. In addition, I led the study design and analysis looking at the role of social determinents of health on care coordination utilization.


  • Schrom JR, Schimotzu S, Poplau S, Larsen K. Improving patient center medical home coordination in a safety net healthcare system among adults living with HIV. Oral presentation at: American Public Health Association Annual Meeting; November 2013; Boston.

  • Schrom JR, Schimotzu S, Poplau S, Larsen K. Predicting Care Coordination Utilization in Minnesota Health Care Homes. Poster presented at: Health Equity Summit; April 2012; Minneapolis.