Better Decisions Through Science // Health Care

With uncertainty at an all-time high, the operations and management science faculty, along with Tuck's international trade economists, are helping a range of industries refine their practice and prepare for what might come next. In this multi-part feature, we highlight current research by Tuck faculty.

Focus // Health Care Operations

---

The Challenge

Manage COVID Vaccine Inventory

Last winter, Michael Gilbert, M.D., the chief medical officer at Catholic Medical Center, in Manchester, NH, was rolling out the first phase of the COVID vaccine to his employees. Dr. Gilbert is an alumnus of the Master of Health Care Delivery Science program, and when a variety of logistical problems came up, he contacted Rob Shumsky, his Operations Management professor and the co-faculty director of MHCDS. In particular, Dr. Gilbert and his team were struggling with whether to use the doses they had in hand for first doses or save them in inventory for second doses. “Our research in the Operations and Management Science group is inspired to help solve open problems,” Shumsky says, “and here was one that fit very well with our expertise in modeling and optimization.”

Shumsky set to work right away, partnering with Dr. Gilbert; Tuck colleague Jim Smith, the Jack Byrne Distinguished Professor in Decision Science; and Anne Hoen, an associate professor of epidemiology, biomedical data science, and microbiology and immunology at the Geisel School of Medicine.

Using data from the vaccination center at Catholic Medical, the researchers constructed a mathematical model to explore the performance of various vaccine inventory management alternatives when faced with supply uncertainties. At the time, the supply of vaccines wasn’t the only unknown in this challenge. The medical community was also unsure of the health consequences of delaying the second dose beyond the recommended three to four weeks. What were the consequences of administering a second dose after five weeks, six weeks, or more?

“The model really highlighted that uncertainty,” Shumsky says, and “it helped us figure out how important that information was.” Surprisingly, the model showed that even over a broad range of health consequences of late doses, the optimal inventory decision was essentially the same. In a working paper reporting this research, they recommended that each week, a vaccination center should complete all second doses that are due, and that if extra doses remain, they should be set aside for the following one to two weeks. Any doses beyond that should be used for first doses.

This prescription was dramatically different from the strategy used at the time by many vaccination centers, where they would save a second dose for every first dose administered. The model showed that this “lockbox” strategy was too conservative and prevented the centers from getting the vaccine to more people, faster.

Many important decisions rest on incomplete information, and for Shumsky, that’s all the more reason to use science and math to guide the way. “Uncertainty might make someone question the point of modeling—if you don’t know anything, you don’t know anything,” he says. “But the great thing about modeling is that you can understand if the uncertainty makes a difference in what the best decision is. You not only find the optimal solution, but you can also see how robust that solution is, given the unknowns.”

 

---

The Challenge

How to Improve the Process of Transferring Heart Attack Patients

Heart disease is the leading cause of death in the United States, and one of the most common symptoms of heart disease is heart attack. Heart attacks are highly treatable, but patients must be treated quickly to maximize their chance of survival. “This is where I see an operations management opportunity,” says Lauren Lu, an associate professor of business administration and the Daniel R. Revers T’89 Faculty Fellow. As an operations management professor, Lu likes to optimize processes and increase efficiencies, and one area she has targeted for research is in the transfer of heart attack patients across hospitals. 

Why transfer heart attack patients? Because many patients live in rural areas and receive their emergency treatment in a local community hospital, far from a tertiary care hospital that can best treat their condition. Lu has written two papers examining the process behind transferring heart attack patients. One looks at where patients in Florida were transferred after a heart attack. She found that Florida hospitals like to transfer their heart attack patients to affiliated hospitals, even when those hospitals might be of lower quality and farther away than non-affiliated hospitals. These “relationship-driven” transfers often sacrifice patient outcomes for the chance to keep the treatment revenue in the same hospital system. Her study calls for a new transfer process—one that prioritizes quality or distance over relationships among hospitals. “Relationship-driven transfers have a lot to do with incentives and protocols,” Lu says. “If you fix the incentive problem, then it’s about designing new protocols to ensure the right hospital is selected for transfer.”

Lu’s second paper on heart attack patient transfers examines the role of technology in the transfer process. During the last two decades, most hospitals have transitioned to electronic health records (EHR), but sometimes hospitals use different EHR systems that can’t talk to each other, i.e., are not interoperable. The Centers for Medicare and Medicaid Services (CMS) has set up incentive programs to promote interoperability. Her paper shows that EHR interoperability has salient benefits for heart attack patients, reducing transfer times by as much as 40 minutes, which led to lower 30-day readmission rates. “The adoption and management of health information technology has life and death consequences,” Lu asserts, “and providers need to make an effort to realize its full potential.”

 

Next

Supply Chain