Select Page

When it comes to building better healthcare infrastructure, not all organizations are created equal. Even a general hospital will receive different kinds of patients and face different challenges depending on the city it’s in, to say nothing of other specialized care facilities. While there will always be approaches in health that are universally accepted, the success of any care facility is highly dependent on its willingness to examine the social factors behind the health of the patients it cares for and adapt to better serve their needs.


This is not unlike the ever-changing goals of nonprofits, where having a cause doesn’t necessarily mean that it fits those that benefit from an organization’s activities. It is the duty of every healthcare organization to detect trends that should be addressed in patient populations, particularly when it comes to social determinants.


Even so, this is a daunting task. The data from electronic health records (EHRs) is out there, but making sense of it can take a lot of time and effort. However, Mount Sinai Health System is leveraging the power of artificial intelligence (AI) to make gathering insights easier.


The challenge with extracting information from EHRs is twofold. First, collecting the information in a consistent manner can be challenging to do across an entire team of medical professionals. While it’s easy to document specific diseases patients might have, there might not be a good way to note other issues that might contribute to a certain condition, such as lack of access to transportation, food insecurity, and behavioral health factors. Secondly, trying to extract data from analytics can potentially be flawed if there are aspects of health that aren’t covered in the structure of an EHR.


Still, hospitals can lean on the verbal descriptions of medical professionals to gain insights into the social determinants of patient health. Language processing algorithms can help detect patterns in patient descriptions and build a better picture of their overall health factors. It’s data that no longer needs to be combed-through by humans, and it helps account for things that might not be recorded in an EHR. 


Mount Sinai is looking to make better valued-based care a reality through these applications of AI. They’re currently in the process of reworking their infrastructure to better make use of their accumulated data. Though it initially began as a smaller project, it has evolved to the point where most of the organization’s data science team is involved.


The initial objectives of this project have been to map out solutions for the most high-impact social determinants. Some of these include economic factors, education, physical boundaries, and health system factors—such as insurance and language barriers. Future phases of the project are set to dive into less major determinants, like legal issues and physical activity. This has been a massive undertaking, requiring data scientists to audit existing information to make sure that it can be interpreted by AI. 


As Mount Sinai’s research progresses, they intend on also reevaluating the ways that physicians write their notes to try to make future data more consistent with the AI infrastructure they’ve created. Even so, the number of insights they’ve gathered through machine learning promise to deliver a better patient care experience for all of their facilities.