How technologies are helping health systems think about-and mobilize on-social determinants of health.
Social determinants of health (SDOH) has become one of the biggest buzzwords in healthcare. In the quest to provide more complete care-and lower healthcare costs-stakeholders across the healthcare industry, are realizing that treating a patient doesn’t have to begin or end in a hospital.
SDOH encompass a variety of factors, from housing availability to water access to income to food stability-not typically factors considered by most stakeholders.
There’s no question that SDOH affect health outcomes, but stakeholders are grappling with how big a role they should take in addressing them. As Benjamin Zaniello, chief medical officer at Collective Medical, a network for care coordination, put it, addressing SDOH is “not simply a healthcare problem. It’s a societal problem. We have not yet fully addressed food and shelter and security in either rural areas like Kentucky or urban areas like Chicago or New York. To expect hospital physicians to fix those problems within their four walls is an unreasonable expectation.”
However, he adds, “health systems do tend to gravitate toward where the money is. There is a widespread perception that focusing on SDOH have a measurable impact on healthcare quality and costs. I think there will be increasing investment in this area.”
The trend is already gathering steam: according to a 2017 Deloitte survey of hospitals looking at SDOH, 88% at least screened for social needs-an indication that SDOH are, if nothing else, coming into focus.
As more attention is placed on lowering costs, experts say SDOH are likely to become a larger part of the healthcare conversation.
Zaniello says that “In the ED, providers and social workers will tell you that social determinants-specifically, challenges with food, shelter and security-are directly tied to the huge volumes of patients admitted to EDs. The expression for a huge volume of ED admissions, for a ‘hot and a cot,’ was always taken as being dismissive of patient’s needs-but it probably most accurately describes them! And now that health systems are increasingly taking on risk, they are more incentivized to leverage their hard-earned insight into the ways social determinants undermine the cost of healthcare.”
AI, often touted as the next great innovation in healthcare, is also being used to address SDOH.
Sashi Padarthy, AVP, Healthcare Business Consulting at Cognizant, a technology consulting firm, says that machine learning will be one of the most important technologies for success. He sees AI being used to detect SDOH factors in EMRs, which can then notify providers and care managers in real time.
Currently, he says, “capture of SD data is still very much based on manual, and often time-consuming processes (e.g., care managers often leverage standard paper-based questionnaires). Recent advancements in artificial intelligence and machine learning-new methods of identifying SDoH-have started to show significant promise. These new methods leverage the data that is already present within a patient’s electronic health record as an input to a machine learning model that analyzes and predicts the likelihood the patient is at risk for a SDOH.”
Padarthy points to Lucina Health’s program that utilized AI to identify mothers at risk for pre-term births, taking medical and socio-economic factors into account to determine the best care plan. The initial results of the program, which began in May of last year, have been largely successful: preterm births have been reduced by 13% for the high-risk group.
Another AI technology aiding communication between CBOs and providers is natural language processing (NLP), a tool already in use in many health systems. Linguamatics, an IQVIA company, is an NLP tool that takes account of SDOH. Linguamatics’ director of clinical analytics, Elizabeth Marshall, points out that “SDOH are much more likely to be documented in clinical notes than anywhere else. This information is captured by a diligent clinician, especially when they believe it to be an issue that may be impeding health conditions for the patient.” She says some clients are beginning to use this information in their analytics-already many organizations are using NLP to track if patients smoke or abuse alcohol.
Currently, says Jacob Reider, MD, CEO of the nonprofit Alliance for Better Health, Albany, New York, SDOH are usually being addressed not by health organizations, but by community-based organizations (CBOs) in conjunction with health organizations. “I think this is appropriate,” says Reider. “We need not seek to see medical care providers grow their skill sets to take on these services. Rather, medical care providers should partner with CBOs to make sure that these needs are addressed.”
Because of this, one of the largest ways technology is helping address SDOH is by improving communication between CBOs and health organizations. Reider points to the success of close-loop referral systems like Unite US and Now Pow, calling them the “most important new category of tools in this space.” These types of systems connect social care and medical care providers. “This makes the right thing to do-screening for SDOH and then referring patients with social care needs to appropriate providers-the easy thing to do,” Reider says. “And that’s precisely what great technology should be doing for us.”
Related article: Three Ways to Better Address Social Determinants of Health
Padarthy calls this type of communication “social prescribing,” saying that it’s likely going to become one of the most important ways technology addresses SDOH. “Just like prescribing medications and sending the prescription to a pharmacy, social prescription allows providers to search and notify the nearest shelter, food bank, or transportation service and send the information to them to make it easy for patient and frictionless,” he says.
He points to companies like Aunt Bertha, which is already involved in the space, or companies like Lyft and Uber that are working on how their services can be used to coordinate transportation.
Another concrete example Padarthy cites comes from United Healthcare, which rolled out ICD-10 codes that providers could use to document SDOH needs among their Medicare population. Between January 2017 and the end of 2018, their providers made 560,000 referrals to social services, equivalent to $250 million in social value.
It’s also important to give patients the ability to communicate needs as well, beyond those given during an examination. Rhonda Mims, executive VP and chief public affairs officer at WellCare, a health plan provider based in Tampa, Florida, describes the success of WellCare’s Community Connections model, which launched in 2011. Part of that model (among other technology advances like capturing transactions in an EHR and sharing that information with partners like state agencies and providers) is a call center. Mims says that in their first year in 2011, the organization received 5,000 calls. Currently, they now field 105,000 calls per year, including requests for food, medication assistance, transportation, and housing support.
In the future, experts say this type of information exchange will be much more commonplace.
Reider says that in 15 years, everyone will control all of their health information with a biometric key. Then social and medical care providers, as well as trusted family and friends, will be able to input information, order services (based on their role), and all communicate with one another more easily. “The core of this infrastructure will be communication rather than documentation,” Reider says, “and yet an accurate picture of the needs of an individual and the services provided will be painted automatically by natural language processing (NLP), machine learning, and perfectly-designed user experiences.”
Remote patient monitoring
The most important currently available technology is mobile, according to Harry Soza, president and CEO of CAREMINDr, a technology company that partners with health plans and providers to advance population health management. “Smartphones and mobile devices are widely adopted and used daily across all socioeconomic levels and ages-including Baby Boomers, who tend to have chronic conditions that can be affected by SDOH,” he says.
Related article: Four Ways Mobile Remote Patient Monitoring Can Improve Star Ratings
Compared with telephonic methods, Soza adds, mobile is far more efficient and results in much more reliable patient information. He says that mobile provides payers and providers opportunities to use targeted outreach methods to educate members both about their disease and also provide information about relevant CBOs. Some-whom Soza refers to as “progressive payers”-are even offering apps to that enable both face-to-face appointments and the ability to flag SDOH issue. Once those issues are flagged, the app can direct them to resources. “Offering this seamless experience will greatly increase the likelihood that members will participate,” Soza says.
Once patients do participate, providers can start to glean more information about patients. Face-to-face interactions via remote patient monitoring (RPM) don’t only involve faces. Mims points out that “RPM allows care to focus on factors that may negatively affect a person’s health that are not controlled inside the hospital. Through phone and video conference, doctors gain perspective on a patient’s life and step in when needed. Also, virtual care allows doctors to see every aspect of a patient’s home setting, which can help assess health issues that a patient might not otherwise reveal.”
Nicholas Hamm is an editor with Managed Healthcare Executive.