Imagine if data could aid in forecasting a patient’s outcome, enhance hospital efficiencies, or refine human resources in healthcare. A newly released book, “The Analytics Edge in Healthcare,” illustrates that this is already occurring and shows how to expand it.
Written by Dimitris Bertsimas, MIT’s vice provost for open learning, alongside two of his former pupils — Agni Orfanoudaki PhD ’21, associate professor of operations management at the University of Oxford’s Saïd Business School, and Holly Wiberg PhD ’22, assistant professor of public policy and operations research at Carnegie Mellon University — the book delivers a pragmatic introduction to the realm of health care analytics. With a focus on practical applications, the initial section of the book lays down technical groundwork — covering machine learning and optimization — while the latter section offers cohesive case studies spanning various clinical specialities and problem types, employing descriptive, predictive, and prescriptive analytics.
As part of a wider series, “The Analytics Edge in Healthcare” illustrates how to utilize data and models for enhanced decision-making within the healthcare sector, while its predecessor, “The Analytics Edge,” explores the science of data-driven models, facilitating better decisions and providing value to institutions and individuals.
Bertsimas, who also holds the position of associate dean of business analytics and the Boeing Leaders for Global Operations Professor of Management at the MIT Sloan School of Management, is behind 15.071 (The Analytics Edge), a course offered on MIT Open Learning’s MITx, which has engaged hundreds of thousands of online students and inspired the book series. Bertsimas took time away from research and his role at MIT Open Learning to discuss how analytics is revolutionizing the health care industry and share some unexpected applications of analytics in hospitals.
Q: How is the field of analytics transforming how hospitals dispense care and oversee their operations?
A: As someone in the academic field, I have always aimed to teach, publish findings, and apply our research in practice. Consequently, I established Holistic Hospital Optimization (H20) to enhance hospital operations through machine learning, improving patient care. We’ve developed numerous tools at MIT and have implemented these tools in hospitals globally. For instance, we handle patient lengths of stay and their deterioration indexes (a computerized mechanism that forecasts a patient’s risk of clinical decline); we optimize nursing assignments and ensure resources are allocated fittingly; and we enhance scheduling for surgeries. This marks the start of a transition where analytics and AI approaches are increasingly being applied. My aspiration is that this work and this book will hasten the impact of utilizing these tools.
Moreover, I have conducted a nine-lecture course twice alongside Agni and Holly at the Hartford Hospital System, where I recognized that these analytical techniques — typically not incorporated into medical education — can be showcased for healthcare professionals, including doctors, nurses, and administrators. To have a real impact, you must possess effective methods, put them into practice, and educate individuals on their utilization. This aligns well with my responsibilities at Open Learning, where our aim is to educate learners worldwide. Indeed, Open Learning is introducing this fall Universal AI, an engaging online educational platform designed to offer in-depth knowledge of artificial intelligence, preparing a global student body for careers in our rapidly changing job market.
Q: What are some unexpected applications of analytics in healthcare that many might not anticipate?
A: Through analytics, we have successfully decreased the average length of stay for patients at Hartford Hospital from 5.67 days to just five days. We have created an algorithm that estimates the likelihood of patients being discharged, allowing doctors to prioritize those with the highest probabilities and prepare them for release. This enables the hospital to treat a significantly larger number of patients while also decreasing patient hospital time.
Additionally, when hospitals experienced a surge in nurse turnover during the Covid-19 pandemic, we designed an analytics framework that emphasizes equity and fairness, reducing overtime expenses by offering preferred schedules to nurses, which substantially lessens overall turnover. These are merely two instances; numerous others illustrate how an analytical viewpoint towards healthcare and medicine has generated tangible improvements.
Q: In the future, how do you envision artificial intelligence influencing the healthcare landscape?
A: In a profoundly impactful manner — we employ machine learning for improved predictions, but generative AI can clarify them. I already notice a trend in that path. This advancement in AI has enabled these capabilities, which is thrilling. It’s also crucial for the world because of its potential to enhance care and preserve lives.
For instance, during our program at the Hartford Hospital System, we identified a patient whose condition was deteriorating and, through analytics, projected that their situation would worsen. Following this prediction, the doctors examined the patient more thoroughly and found that they were in the early stages of sepsis, a critical condition where the body inadequately responds to an infection. Had we not detected sepsis early, the patient’s life might have been at risk. This had a real impact on preserving someone’s life.
Q: If you were to encapsulate “The Analytics Edge in Healthcare” in a word or two, what would they be, and why?
A: The book signifies a transformative shift in healthcare as it holds the potential to impact the sector in unprecedented ways. It truly articulates my contributions to healthcare and its applications over the past decade.