For certain payers, the average charge for a vaginal delivery at New York Presbyterian Hospital, located in Manhattan’s Washington Heights neighborhood, is nearly five times higher than a birth under the same circumstances at Olean General Hospital, about 350 miles away in western New York.
The nearly $17,000 charge difference between the two hospitals is eye-opening. For that reason, those numbers are the first data sample Marc Triola, MD, presents to medical students when they begin their work in the Health Care by the Numbers (HCBN) curriculum at New York University School of Medicine.
“We need to understand how to navigate and use these data sets,” said Dr. Triola, the associate dean for educational informatics at NYU’s School of Medicine. “If we are not navigators of them, we’ll become victims of them. They are really going to become a critical part of how we drive improving health care for each patient in a health care system and for our country as a whole.”
Dr. Triola presented on the benefits of future physicians’ leveraging big data to develop high-value insights into population health during a recent AMA Innovations in Medical Education Webinar, “Using Big Data to Teach Population Health.”
In order to improve the health system, medical professionals across the continuum must have a grasp of the data associated with it. NYU, a member of the AMA’s Accelerating Change in Medical Education Consortium, is doing this through interactive research projects that are part of a mandatory curriculum for pre-clinical medical students.
NYU uses data available to the public through the New York State Department of Health Statewide Planning and Research Cooperative System. The nearly 40-year-old state program collects a vast array of information—including patient-level demographics, provider details, diagnoses, procedures, costs and charges—for all inpatient admissions to every hospital in the state. To narrow the scope of the data and make it more digestible to students, the university has created a more user-friendly database.
Once students have acquired a background on the available data and how it can be effectively harnessed, they are divided into teams of two and asked to come up with a question that can cover any topic or aspect of care and can be addressed by information in the clinical database. Thus begins the experimental learning aspect of the HCBN curriculum.
The five-week research project requires students to practice skills that are central to population-specific care, such as:
- Developing a testable hypothesis across large numbers of patients.
- Outlining a methodological approach.
- Synthesizing and communicating their findings.
Dr. Triola is a proponent of this type of interactive approach to teaching health systems science.
“One of the big challenges with teaching health systems science is that students need to be engaged and motivated and they need to perceive that, when they are learning, they are participating in an authentic role within their health care system,” Dr. Triola said. “If they learn about health care from a textbook and then go into a hospital and feel very disconnected from all of the policy-level or health management things that are happening there, it doesn’t click.”
HCBN was introduced to NYU’s class of first-year medical students for the first time in 2014. Since then, 225 pairs of students have answered questions using the clinical database the institution created. Students have been encouraged to follow their areas of interest.
Those interests generally seem to be weighted toward disparities in care and geographic variations in quality and availability, Dr. Triola said. For instance, one duo studied how the gender and racial composition of patients admitted with a diagnosis of major depressive disorder changes between various age ranges.
“We viewed it as our mission that the students needed to define a question that they are going to be passionate about,” Dr. Triola said. “This is something that they are going to be interested in, that they want to potentially have as part of their future career or their specialty or subspecialty choice.”
When students present their findings at the conclusion of their research, the results are sometimes presented at conferences or published in peer-reviewed journals. NYU’s instruction in big data has been beneficial for both students and faculty.
"In many cases, these topics and these databases are as new to our faculty as they are to our first-year and second-year medical students,” Dr. Triola said. “Our faculty essentially needed the same orientation that our students did. They needed to understand how to recognize both the power and limitations of these data.”