HSI Funds Five Research Projects in 2020

HSI held its second call for research proposals to MIT Sloan faculty in March of 2020, and will fund five projects this year. The funds will provide support for a variety of needs, including doctoral student funding, data purchasing, participation incentives, project management funding, and/or travel funding.
We are excited to be able to use HSI funding to support work that crosses disciplines within MIT Sloan, as well as work that falls within more than one of our three focus areas. We will bring you updates about these projects in the future.
Congratulations to all the recipients!The newly funded projects are:

1. "Warehouse Work & Worker Well-being“, by Erin Kelly in Work and Organization Studies, and Hazhir Rahmandad in System Dynamics. Funding will support a cluster-randomized trial in a national e-commerce firm’s network of 28 warehouses. The trial evaluates whether scheduling changes designed to increase workers’ say regarding their schedules will improve employee health and well-being, and benefit the firm through reduced absenteeism, reduced turnover, and perhaps increased productivity. While this study does not take place in the health care sector, its findings may have implications for health care workers and for workers in other industries characterized by demand volatility and highly interdependent work processes. This study aligns well with HSI’s exploration of research in employee population health.

2. “Combining machine learning and behavioral insights to encourage medication adherence”, by Jónas Jónasson in Operations Management, and the MIT Applied Cooperation Team led by David Rand in Marketing and Erez Yoeli, a research scientist in Marketing. This project builds on prior research demonstrating that users of a mobile phone platform to support tuberculosis treatment demonstrated a 67% reduction in treatment non-completion relative to a control group receiving standard care.The new effort consists of two parts: classifier development, and testing those classifiers in a field trial.  The classifiers aim to identify patients who will benefit most from treatment. One is designed for use prior to enrollment to determine who should be granted access. The other will be applied post-enrollment to identify patients who might benefit from increased engagement based on their interactions with the platform, to focus on those most at risk of non-completion.
   
   The field trial will be a set of randomized control trials that compare multiple treatment groups with control groups, to show both the cost and success impacts of using each classifier. The goal is show that using the classifiers will save substantial costs while also improving performance (the second classifier), and will lead to only modest performance reductions relative to granting all those eligible for treatment access to the platform (the first classifier).
   
   We are very excited to be able to use HSI funding to support work that crosses disciplines within MIT Sloan.
3. “The Relationship between Health IT Systems and Healthcare Industry Consolidation, Opioid Epidemic Control, or Pandemic Responsiveness”, by Catherine Tucker in Marketing and her longtime collaborator Amalia Miller at the University of Virginia. The research will use econometric methods and existing American Hospital Association survey data on IT.  Research questions include:
   -  What is the role of healthcare IT in enhancing hospital mergers by expanding access to data, or potentially constraining the benefits of mergers if hospital IT systems prove to be costly and hard to integrate?
   -  How does the ability to document electronic prescribing of controlled substances impact recorded deaths from opioids? More specifically, what is the effect of allowing electronic prescriptions, sharing data with a state’s prescription drug monitoring program, and the combination of these two capabilities?
   -  What is the impact of the types of IT systems adopted by hospitals prior to the COVID-19 epidemic on health outcomes from the pandemics? For example, what are effects of regional health information exchange organizations (RHIOs)?
   This project fits well within our focus areas of both Healthcare Analytics and Healthcare Operations, as it investigates ways in which health IT infrastructure affects care.

We are also continuing to fund two projects that we supported this past year.

1. Predictive and Prescriptive Analytics to Address the Substance Abuse Crisis” by Georgia Perakis in Operations Management and Dessislava Pachamanova from Babson College. The project focus is on both prediction and prescriptive methods to improve treatment of patients with SUD. The analytics take into account patient characteristics and risk profile to prescribe the treatment that would be best for the patient, under both capacity and budget constraints and in a scalable, multi‐staged fashion. The models strongly emphasize interpretability and show promise in experiments with simulated data. The researchers are testing the models with openly available data from the MIMIC dataset, and are working with hospital partners to obtain de-identified data for customized model development and field-testing.
2. “Geisinger’s Fresh Food Farmacy: a Randomized Controlled Trial” by Joseph Doyle in Applied Economics. This study, with Geisinger Health, evaluates health impacts on patients and their families of its Fresh Food Farmacy effort, a food-as-medicine approach to improve health among food-insecure patients with diabetes featuring a diet prescription filled at an FFF each week, along with dietitian advice and care management. Patient recruitment began in April 2019 with a goal of 500 participants. Close to 300 subjects have signed up as of March 2020. HSI funds are supporting research project management and incentive payments to participants for the extra testing needed to measure outcomes for the study.