Inside the Research Informatics Center

The Research Informatics Center staff includes expert programmers, database architects, data scientists, and statisticians.

Photo by Kris Newby


Stated simply, the Research Informatics Center (RIC) helps Stanford investigators access clinical data for research purposes. The RIC staff, which includes programmers, database architects, data scientists, and statisticians, are well versed in the emerging fields of artificial intelligence (AI), machine learning, and neural networks. They are expert data miners and problem solvers, focused on assisting investigators in figuring out what data is needed, how to extract and store it, and ways to analyze the data to obtain answers to scientific questions.

“Our mission is not just to deliver data, but to let investigators know about new approaches to addressing their research questions,” said Daniel Rubin, MD, MS, Professor of Biomedical Data Science, Radiology, and Medicine (Biomedical Informatics) and the leader of the RIC. “The paradigm of data-driven medicine is moving to the forefront, powered by artificial intelligence and access to electronic medical records, images, and omics data.”

Specifically, RIC experts can help researchers:

  • Review the clinical data needs of research projects.
  • Learn about the broad array of clinical data available from the electronic medical record and other systems in Stanford Medicine.
  • Discuss options for clinical data abstraction, reporting, and storage.
  • Apply advanced statistical and bioinformatics tools to extract structured information from a variety of electronic medical record databases.
  • Extract data from Stanford Medicine Research Data Repository (STARR), which includes Electronic Medical Record (EMR) data going back to 1998.
  • Provide data insights through exploration, data mining, and natural language processing capabilities.
  • Search and extract information from clinical texts and access medical images.
  • Set up statistical modeling and visualization tools.

Inspirational big data projects

Analyzing the vast stores of clinical and image data in electronic medical records offers researchers a powerful way to obtain useful medical knowledge, all at a fraction of the cost of prospective research. These up-front analyses are particularly important for generating and prioritizing hypotheses prior to undertaking expensive prospective studies. New analytical tools and methods are opening up exciting new possibilities for better diagnosis, insights into disease progression, and a deeper understanding of cellular and biochemical processes.

A few examples of novel RIC projects include:

  • The extraction of electronic medical record data from more than 12,000 patients with metastatic cancer, which is analyzed by a custom prognostic software model to predict patient survival rates. The model showed a statistically significant increase in accuracy compared to prior prognostic models. 
  • Design of a new bioinformatics software pipeline that rapidly and accurately identified likely sources of bloodstream infections, making it easier to rid patients of dangerous reservoirs of pathogens. In this study, Stanford scientists showed that such infections often originate from patients’ own bodies, usually from their large intestines. 
  • Development of an algorithm to collect and analyze body temperature data from electronic medical records. Based on 230,261 measurements extracted from the Stanford Translational Research Integrated Database Environment cohort (STRIDE), Stanford scientists determined that the average human body temperature has decreased monotonically by 0.03 °C per birth decade from the 37 °C standard set by Carl Reinhold August Wunderlich, a 19th century German physician.
  • A 15-year collaboration with the Bone Marrow Transplant (BMT) program to collect and organize treated-patient data in a core database. In this database, BMT researchers have established more than 100 critical variables for their analytical needs. RIC developers then created an easy-to-use web-based analysis console that allows researchers to slice-and-dice the data interactively, compile summary statistics, and download it for off-line exploration. The analysis console also offers dynamic plotting of a variety of survival and cumulative-incidence curves used in evaluating clinical outcomes and efficacy of treatments. And it enables the BMT group to provide complex reports to both Stanford Hospital departments and external groups.

How to work with RIC

To schedule a face-to-face meeting with a RIC expert, fill out the short consultation form on their website. Before the consultation, the staff recommends that researchers check out its self-help tools, which are designed to help refine research questions, determine patient enrollment criteria, and gain an understanding of the regulatory requirements for data access.

The Cohort Discovery Tool is an online search dashboard that enables researchers to quickly assess whether there are enough Stanford patients that meet the enrollment criteria to justify a specific study. The tool can access structured data such as diagnosis and procedure codes, drug classes and ingredients, and more. It also includes unstructured clinical notes and time-based filtering parameters. Visualization features allow for comparisons segmented by sex, age, race, and location.

Once a study cohort is defined, the Chart Review Tool is designed to facilitate review a deeper review of patient charts. Its powerful search engine enables an investigator to gain insights into target patients and their medical histories to further refine a research proposal.

RIC’s Compliant Access to Clinical Data webpage provides researchers with a quick overview of privacy and approval requirements for clinical data access. The process of obtaining this data is dependent on the intent of the project and the specifics of the data being requested. HIPAA patient privacy regulations apply whenever the intent is to use data for research purposes.

Once a consultation is requested, the RIC will assign a point person to work on each project, to develop an analytic approach and a data collection/storage framework going forward. Searching for gems of clinical wisdom in electronic medical records is often like finding a needle in a haystack.

“Data extraction is not always straightforward,” said Balasubramanian Narasimhan, a Senior Research Scientist in the Department of Biomedical Data Science who works with the RIC staff. “Part of RIC’s task as consultants is to listen to researchers’ wishes and match them to reality.”

RIC services are charged to research projects on an hourly basis, and they are not limited to School of Medicine investigators. Recent clients have included researchers from the Graduate School of Business; Engineering; and Humanities and Sciences.

For more information about RIC services, visit this website or fill out this consultation form. If you have general questions about other Stanford data resources, visit the Stanford Data Science Resources website, a central portal from which data scientists can access advanced tools, data platforms and experts in diverse methodologies for conducting biomedical research.