Understanding the Data Divide — Problems & Solutions Regarding the SDoH
How the disconnects within healthcare affect our ability to address the SDoH
“The conditions in which we live explain in part why some Americans are healthier than others and why Americans more generally are not as healthy as they could be.”
Improving health access and education is an essential step in addressing the Social Determinants of Health (SDoH). But, in order to thoroughly target the SDoH, a deeper understanding of their origins is needed.
Social and physical determinants of health vary per community and per person, and include such far-reaching topics as public safety, housing, social support, cultural norms, and access to educational, economic, and job opportunities.
According to Healthy People 2020, any resource that enhances quality of life can have a significant influence on population health outcomes. The following place-based framework includes 5 categories for better identifying the source of disparate SDoH:
- Economic Stability: employment, food insecurity, etc.
- Housing Instability: poverty, education, language and literacy
- Social and Community Context: civic participation, discrimination, incarceration
- Health and Health Care: access to health care and primary care, health literacy
- Neighborhood and Built Environment: access to foods that support healthy eating patterns, crime and violence, quality of housing
A comprehensive understanding of how these complex factors influence health is imperative for achieving true preventative healthcare.
Until now, however, disconnects in data gathering, sharing, and analysis across healthcare have delayed the process. As a result, valuable insights are lost which could benefit healthcare and patients, such as:
- Which patients are at risk for readmission?
- Which SDoH are prevalent in our community?
- Which interventions should be prioritized for this patient and their family?
The absence of these insights and many others should not be taken lightly. Without them, providers, insurers, and patients are flying blind.
Following are two current barriers getting in the way of true preventative care, and emerging solutions to build a better (data) base for action:
Problem 1: Disconnected Data
“Interoperability remains a major challenge because even though we have datasets available to us through the federal government or through local organizations and local community groups, that data is often not brought into the EHR system,”
New research from the American College of Cardiology (ACC) shows that failure to use SDoH data when determining patients’ risk for cardiovascular disease leads to health disparities. In one study, Dr. Gmerice Hammond states “If we systematically underpredict risk, we will systematically undertreat.”
The challenge of “siloed data” is one that affects everyone. In particular, the data disconnect has stymied progress in addressing SDoH. Food deserts, lack of medical specialists, patients without access to transportation, and other issues are due in large part to a lack of data sharing across healthcare.
Providers, payers, and patients serve to benefit from using patient data to inform clinical decision making, resource allocation, and community needs. But given the complexity of the structural landscape and lack of interoperability, data that could save lives lie unused and ignored.
Solution 1: Using SDoH data can inform chronic disease risk assessment
“Using a database that tracks health and environmental data at the neighborhood level, including air quality metrics, crime statistics, access to affordable housing, and availability of grocery stores, researchers are uncovering new insights related to diabetes, obesity, and hypertension.”
— Jessica Kent, Top 3 Data Challenges to Addressing the Social Determinants of Health
With more data, more actions can be taken to care for underserved communities. In turn, these data will help providers improve clinical risk prediction, identify high-risk patients, and address their needs sooner rather than later.
Standardization and sharing of SDoH screening and data
In order to increase population health management across communities and geographies, standardized, easily implementable screening and data sharing will allow providers to gain insights into health-related social needs.
Remote Patient Monitoring brings patient data to the hospital
As the decentralization of healthcare continues, remote patient monitoring platforms like Cliexa will allow providers to identify and monitor at-risk patients and better understand and assess their symptoms. By using patient-reported outcomes, Cliexa can also fill in the gaps in the patient experience, informing higher-quality treatments.
Machine Learning for payer risk adjustment
Incorporating machine learning (ML) and social determinants of health indicators improves risk adjustment for health plan payments. In a recent study, ML improved risk adjustment models, and the incorporation of SDoH indicators reduced underpayment in several vulnerable populations.
Natural Language Processing for patient risk reporting
Standardized screening, while necessary, will take some time to develop, let alone be adopted. In the meantime, clinical notes, zip codes, missed appointment records, and other readily-available EHR data can also be used to predict psychosocial risk using Natural Language Processing (NLP).
Pieces Technologies does this by mining untapped data trapped in EHR clinical notes to determine social determinant risks and make recommendations.
Dina, a post-actue care coordination platform, uses NLP to identify risk factors from patient data and their communication platform to identify factors placing patients at risk for readmission.
Great companies leveraging data analytics to bring insights on SDoH:
- Cliexa
- Pieces Technologies
- Dina
- Algorex Health
- Collective Medical
- Salus Board (new company founded by Hector Rodriguez as a part of X Genesis program to combat healthcare resource shortages)
- Wellframe
- Lightbeam Health Solutions
- Clarify Health
- Geneia
- Experian Health
Problem 2: Disconnect between Data and Action
“Providers…often just don’t know what to do with (SDoH data)…Unless that provider has the ability to actively refer the patient to a program, the data really just sits there…We need more opportunities for providers to take action on SDOH so patients can be connected with the social services they need.”
— Joyce Famakinwa, Hospitals, Physicians Aren’t Screening for Social Determinants of Health
Despite a communal want to deal with the SDoH, physicians are limited by the time and resources available to them. According to Kaiser Permanente, 97% of patients think providers should ask about social needs. Yet, another survey reports, 80% of family physicians feel they don’t have the time to do so.
Treating patients with unmet social needs, as a result, is found to cause symptoms of burnout in primary care clinicians.
Medical providers are already overburdened, and it would be unfair to expect a provider to suddenly take on case management responsibilities. Case in point, even when SDoH data are available, 64% of providers don’t believe they have the resources to do anything about social risk factors.
Clearly, there is a disconnect not only of data but of resources. Data should not be a burden or an additional task, but could instead help remove the time and effort needed to address complex social risk factors.
Solution 2: Actionable insights lead to…action
Data insights are only effective if they result in actions that lead to better health outcomes.
I was a Community Health Worker at a Community Health Center (CHC) and was amazed at what a difference on-premise behavioral health and social services made for our patients.
Armed with only a walkie-talkie, I could have a Behavior Health specialist in my patient’s room in seconds. Connecting patients to needed services should be just as easy and could make the difference between action and inaction on the part of a provider and patient.
Technology can and should make care coordination as simple and streamlined as possible, both for better continuity and quality of care.
Simplifying Care Coordination
Kevin Amell of Julota, a company alleviating interoperability challenges in community health management, highlighted the key components of an effective care coordination structure. These include data integration, interoperability, actionable data, and the ability to determine if patients followed up on referrals (also called closed-loop referrals).
Using care coordination software has been proven to significantly cut readmissions. Leveraging data, these companies are able to identify opportunities for interventions to cut costs and improve health outcomes.
Collective Medical takes a data-centric approach to identify at-risk patients. In one instance, it helped Washington State hospitals cut costs by $34 million through an 11% reduction in ER visits by super users and a 14% decrease in non-emergency visits.
Dina, the patient care communication platform mentioned previously, identifies at-risk patients and provides actionable recommendations for referrals to social services and other helpful resources.
Other companies are using similar approaches to help hospitals identify at-risk populations and coordinate solutions to address pressing needs:
All Americans deserve an equal opportunity to make the choices that lead to good health. But to ensure that all Americans have that opportunity, advances are needed not only in health care but also in fields such as education, childcare, housing, business, law, media, community planning, transportation, and agriculture.
COVID-19 has shocked our system and caused us to rethink the way we work, where we interact, and how we act. Ultimately, a better understanding of the SDoH will influence not only healthcare but other industries as we reconsider what should be prioritized.
In the next installment on the disconnects in healthcare, we will dive into how cross-industry partnerships are the future of not only health access, but equity as a whole. Or, check out the full article here.
