Public Impact Analytics Science Lab (PIAS-Lab) at Harvard

???? We have open #postdoctoral fellow positions. Info below ??????

??We’re hiring!?? The Public Impact Analytics Science Lab (PIAS-Lab) at Harvard, founded and directed by Professor Soroush Saghafian, has a new round of hiring for postdoctoral fellows in the broad area of Healthcare Analytics! https://lnkd.in/guvRf8aN

? New Paper ?Health Services Research Our (Soroush Saghafian, Robert Huckman, Nicole Hodgson, and I) study examines heterogeneity and impacts of EM physician batch-ordering practices (the practice of placing simultaneous test orders) ?? ???? ???? https://lnkd.in/g_sD5DFa We analyze over 50k Emergency Department encounters from the Mayo Clinic of Arizona, where all patients were randomly assigned to physicians during the study period. Physician's vary in image batch-ordering practices, despite randomly assigned patients. While the decision to batch order for a patient itself is endogenous, we use a physician's “batch tendency” as an instrumental variable to explore the relationship between batching and critical outcomes such as LOS, resource utilization, and 72-hr return to the ED. Findings: having a physician with a batch tendency 1 SD > than the average physician was associated with a 4.5% increase in ED LOS, a 0.2 percentage point decrease in the probability of a 72-h return with admission an additional 8 imaging tests ordered per 100 patient encounters Physicians with a lower batch tendency, meaning that they employ a more judicious and sequential approach to ordering tests, were associated with a shorter LOS and fewer overall imaging tests. This is due to the information gain advantage of sequential test ordering. Our study highlights that patterns of diagnostic test ordering in the ED may have implications for the efficiency of care delivery and patient outcomes. Our findings contribute to the growing body of evidence supporting the use of data-driven approaches in ED management.

Congratulations to Mariam Krikorian Atkinson and Soroush Saghafian for being awarded with Health Care Management Science "Best Paper" of 2023 for their work "Who should see the patient? On deviations from preferred patient-provider assignments in hospitals." https://rdcu.be/dXZ0n The paper deals with the following questions: "What is the impact of following preferred assignments (identified based on the consensus opinion of medical experts) on various performance outcomes? Conversely, when is deviation from such preferred assignments beneficial?" The study used two sources of data from a children’s hospital in the US. The study found that for some types of tasks, it may be beneficial to allow physicians to deviate from preferred assignments, whereas for other tasks, allowing flexibility can be detrimental. These findings may help to provide policy guidance for hospital administrators. https://lnkd.in/ghy6kHS4

The rise in hospital closures, like the recent Steward Healthcare bankruptcy, is a growing concern that warrants our attention. But it isn’t just about the loss of access to care. The aftermath has a wide-ranging impact, affecting physicians, neighboring hospitals, and public health at large. Learn more in this research brief about the work of #MRCBG's Soroush Saghafian (Public Impact Analytics Science Lab (PIAS-Lab) at Harvard) https://lnkd.in/evWiStvm #HealthcarePolicy #HospitalClosures #PublicHealth

Come and see these great Public Impact Analytics Science Lab (PIAS-Lab) at Harvard talks at the INFORMS Annual Meeting next week! We are looking forward to sharing our work with you! - Sunday: Soroush Saghafian will be presenting his work on Ambiguous Dynamic Treatment Regimes https://lnkd.in/gdHtH3aX. We also have joint work between Mariam Atkinson and Soroush Saghafian on Who should see the patient? https://lnkd.in/ghy6kHS4 - Monday: Soroush Saghafian will be participating in a panel on Data-Driven Excellence in Healthcare: Merging Optimization, Stochastic Decision-Making, and Operations Management - Tuesday: Jacob Jameson will be presenting joint work with Soroush Saghafian, Robert Huckman, and Nicole Hodgson on the Variation in Batch Ordering of Imaging Tests in the Emergency Department. Hossein Hejazian will be joint work with Soroush Saghafian on Effective and Efficient Staff in Emergency Departments., Mohammad (Pedram) Zhalechian will present joint work with Soroush Saghafian on Harmonizing Safety and Speed: A Human-Algorithm Approach to Enhance the FDA’s Medical Device Clearance Policy https://lnkd.in/gFjAHZth - Wednesday, Sina Aghaei will present joint work with Soroush Saghafian, Jacob Jameson, and Dana-Farber Cancer Institute on #Ai Assistant for Highly Personalized and Dynamic Causal Decision-Making in Cancer Treatments. Sidian Lin will present her work with Soroush Saghafian on Development and Validation of a Reinforcement Learning Algorithm for Personalized Treatment in Bipolar Disorder

Hospital closures mean more than just the loss of access to medical care: Soroush Saghafian weighs in on the consequences of hospital closures that are currently largely overlooked but need to be part of policy discussions, including impact on physicians and nearby hospitals. https://lnkd.in/eZfm--7U

Who makes better decisions, algorithms or physicians? Yesterday we co-hosted a faculty roundtable with Harvard's Center for Public Leadership where Harvard Kennedy School professor Soroush Saghafian discussed the utility to decision-makers of machine learning and artificial intelligence. His Public Impact Analytics Science Lab (PIAS-Lab) at Harvard was recently awarded a federal grant to leverage AI in the design of personalized cancer treatments, along with tools for physicians to probe and question AI recommendations. Other studies seek to answer questions about emergency room operations, the opioid epidemic, and diabetes screenings. This work is promising for city leaders. Bringing data science into the policy arena, the lab strives to aid public officials in creating better-informed policies. More:?https://bit.ly/4gRJcqJ

Our most recent newsletter is out! https://lnkd.in/guxG9bM3 Featured News~ Public Impact Analytics Science Lab (PIAS-Lab) at Harvard received a nearly $3 million grant from the US Department of Defense for a project that will leverage #AI in collaboration with Dana-Farber Cancer Institute to create personalized treatments for cancer patients. https://lnkd.in/g4mh9D_t Professor Soroush Saghafian, director and founder of the Public Impact Analytics Science Lab (PIAS-Lab) at Harvard, is working on solutions, in partnership with the Mayo Clinic, that would improve the design of artificial intelligence tools in health care. https://lnkd.in/gn2SzHDd

Thank you The Harvard Crimson for covering our latest project at the Public Impact Analytics Science Lab (PIAS-Lab) at Harvard in collaboration with Dana-Farber Cancer Institute: https://ow.ly/1la550Tplx8