Data and Medical Education: Reimagining Medical InstructionMarc Triola | Associate Dean for Educational Informatics at the School of Medicine, New York University
Medical schools are uniquely designed to prepare professionals to enter a very specific professional field. As such, it’s critical for medical education leaders to constantly be aware of the needs of the medical labor market and to adjust programming as necessary to ensure future doctors are gaining the skills they will need to adapt to any changes in the space. To this end, the capacity to understand and analyze data has risen to the top of the list of new skills physicians need to learn. In this interview, the first of two parts, Marc Triola will discuss how the NYU School of Medicine is creating opportunities for students to learn about data analysis. In the second part, he will discuss how this change fits into a more broad transformation in the medical education field.
The EvoLLLution (Evo): How are you and your colleagues at NYU School of Medicine using data analytics and metrics to transform the learning experience for medical students?
Marc Triola (MT): Computerized patient data and analytics are transforming the way we care for patients and conduct biomedical research. Hospitals are using large clinical datasets to better understand the value and quality of care delivered in ways they never could before. There is also an explosion of new scientific knowledge, new medicines and technologies that empower patients such as mobile apps and connected health devices. The combined data from all of these sources have the potential to revolutionize the way we coordinate a patient’s care, educate health professions students, and conduct research.
However, our ability to collect data has vastly outpaced our ability to gain insight from or act on it.
All of these advances have led to an environment where clinical data collected from patients, health care providers and hospitals/clinics are ubiquitous and increasingly transparent. This is a challenge and an opportunity for physicians. It is clear that we need new skills in our ability to interpret and use data and analytics to efficiently make clinical decisions, not only about the patient in front of us, but also about our entire panel of patients and the population for which our health care delivery systems provide. These new skills of data literacy will transform our physicians, nurses and health care teams into navigators and leaders of a health care delivery system that is continuously learning from the data it collects about itself.
How do we get there? These are not topics traditionally taught in medical school or residency training. Fortunately medical schools and national efforts such as the AMA’s Accelerating Change in Medical Education (ACE) initiative are collaborating to make these topics a core part of the education for future physicians. At NYU School of Medicine, with support from the AMA ACE program, we have introduced a new Health Care by the Numbers curriculum for all medical students. Key to this curriculum is the use of authentic big clinical data sets abstracted from freely available public sources and de-identified data from our own “data warehouse” via NYU Langone Medical Center’s Clinically Integrated Network.
One of the core activities that every medical student at NYU School of Medicine participates in is a clinical data experience using the New York State Statewide Planning and Research Cooperative System (SPARCS) data. SPARCS includes patient-level data from every inpatient admission to every hospital in New York. We have created a custom database of all 5.2 million statewide inpatient discharges in 2012/13 for our students to use for a research project. Working in teams, students identify a clinical question they want to investigate using the data. Their question can address quality, cost, access to care, and can be related to any of the clinical diagnoses covered by SPARCS. Students are instructed to generate a hypothesis, abstract a large clinical data set, do their analysis, then present to the class potential implications for the health care system based on their findings. Through this process, students not only develop data literacy skills, but they also gain perspective on the power and limitations of these large datasets.
Evo: What were a few of the most significant roadblocks you had to overcome to make data analysis a part of the curriculum at the NYU School of Medicine?
MT: Historically, the biggest obstacle to using clinical patient data in the curriculum was actually getting access to the clinical data. There has been a sea change in the availability and transparency of clinical patient data across medical centers, states and the nation as a whole. Initiatives like Data.gov, the CDC NHANES project, and others have provided a gold mine of de-identified clinical data for health professions education programs and population health researchers.
A second obstacle concerns alignment between the education programs and their affiliated hospitals. In order for a curriculum on clinical data and analytics to be authentic and relevant to our students, it is best delivered in the context of the data about the hospitals and community in which they are learning. Ideally this would mean that clinical data from the hospital flows into the education program (with appropriate protections for patient privacy), and insights from the teachers and learners can then be used to improve the health care delivery system. The education and clinical missions of academic medical centers in this country have coexisted beautifully, but were rarely truly symbiotic given their divergent priorities, stakeholders and needs.
There is a culture-change happening, and many medical centers are now working to blur the lines between the education mission and the clinical mission with the goal of enhancing both. Shared goals, shared pressures, robust electronic data and analytics are bridging the gap between our educational and clinical missions.
This is the first installment of a two-part series by Marc Triola discussing the impact of data specifically, and technological advances in general, on the medical education landscape. In the second installment, he will discuss the broader impact of metrics and IT innovations on the operations of medical schools.