How providers can use social determinants of health to improve care
Some providers are hesitant to take on another data asset, such as social determinants of health (SDOH), and put it in the hands of clinicians who already operate in a world busy with electronic documentation.
Most healthcare provider organizations readily admit that they are overwhelmed by data.
Patient encounter data in EHRs, imaging and lab results, and billing data from claims have generated a venerable sea of complex and disparate information, and organizations strive to harness it to enhance decision-making in the optimal treatment of patients.
This is, in part, why some providers are hesitant to take on yet another data asset, such as social determinants of health (SDOH), and put it in the hands of clinicians who already operate in a world busy with electronic documentation and clinical decision support alerts.
SDOH data garnered great interest in its potential for developing both efficacious and personalized treatment plans. Its value has also been lauded for helping to improve the impact of care by addressing barriers keeping patients from fulfilling their care plan.
Research has shown that medical care alone has a very limited effect on overall population health and could be significantly enhanced by pairing it with approaches that address SDOH. The financial drivers for adoption of SDOH data in care management are just as clear and persuasive.
Within the value-based care paradigm, getting paid requires optimizing performance on numerous quality measures. Data experts and professionals in the patient care stream must think outside the box regarding how to best reach patients who face socioeconomic challenges in their day-to-day life that put them at risk for poor outcomes.
At the same time, providers have had little guidance in how to approach using this data within the existing workflows to help clinicians at the point of care.
Some organizations approach incorporating SDOH by sorting through patient demographic data to identify what factors could help them make clinical predictions. This can be challenging because of the large amount of data and uncertainty regarding what is actually useful because not all SDOH data correlates to health outcomes.
Not all data is created equal. For example, ZIP codes and other basic demographic information obtained from EHRs can be useful for broader population health management efforts, but they only show a narrow view into individual patients’ complex lives outside of the care setting. The demographic data is often not standardized or up-to-date; similarly, surveys often generate incomplete data sets and are a tedious way to fact find on a large scale and are not easy to refresh, nor is self-reported data the most reliable option.
In evaluating SDOH data for use in predictive clinical and analytic models and care management, providers should look for non-clinical data sources that are current, comprehensive and longitudinal. For ease of use, social determinant data should be easily standardized and integrated, regularly updated and consistently linked to patient populations for actionable insights.
Another key criterion for selecting SDOH attributes is ensuring they have been clinically validated against actual healthcare outcomes to confirm their predictive power. Organizations will want to confirm that the SDOH information has either been validated or be prepared to validate it through their own internal analytics groups.
Organizations can implement SDOH data in the form of attributes and scores. SDOH attributes can be used in the provider’s own clinical and analytic models. The second way of utilizing SDOH is through predictive health scores. Scores leverage hundreds of socioeconomic attributes to inform providers about potential barriers to patients’ optimal health and wellness. Scores are particularly useful for those organizations that do not want to build their own analytic models.
Regardless of the type of SDOH input, the data can be used to create a specific health alert to flag an area of concern, or generate “high,” “medium” and “low” risk levels per patient. Additionally, it is possible to identify the top three social determinant factors driving a patient’s risk for more pointed care management solutions. Depending on the patient needs and type of care provided, this could translate to spending a few extra minutes addressing challenges during a clinical visit or prior to a hospital discharge.
Intervention, however, does not necessarily fall on the physician during the clinical visit. In small organizations, doctors may drive assimilation of social determinants data. Medium-sized and larger organizations often hire new staff or initiate advanced training for current employees who are daily participants in the patient care flow. Discharge planners, social workers, health coaches or patient experience officers are now utilizing non-clinical data to identify gaps in care for better patient care. They use it to build a bridge from a healthcare setting to public health and community groups with resources for patients who need them.
These emerging healthcare professionals have access to community- or organization-based support resources and can integrate the SDOH insights into care plans targeting unmet needs. By recognizing the need for, and enabling access to services or interventions, providers are driving better health and wellness outcomes. These programs often relate to accessing healthy food, providing reliable housing or transportation resources, and helping patients manage isolation and loneliness. Partnerships with local community-based or government services will help build networks across—and beyond—the continuum of care.
One health system and insurance company, Geisinger, is already reaping the benefits of its Fresh Food Farmacy initiative, which granted free access to a food pantry and education for specific “food insecure” patients struggling with type 2 diabetes. Before the Farmacy, these members cost Geisinger Health Plans an average of $8,000 to $12,000 per person, per month. The payer-side costs have dropped by two-thirds, on average, across the program, since the intervention.
Geisinger has also seen significant improvements in care outcomes in the form of patients’ cholesterol, blood sugars, and triglycerides, improvements that can lower the chances of heart disease and other vascular complications. By zeroing in on food insecurity in a county with particularly high rates of diabetes, food insecurity, poverty, and unemployment, the provider was able to deliver resources that altered the path of treatment and improved the patient condition.
While using SDOH data to affect patient management is relatively new, the practice is here to stay. To ensure the data is used for maximum good in healthcare through safe, secure, ethical, and collaborative means, organizations can consult industry resources such as “The Guiding Principles for Ethical Use of Social Determinants of Health Data.” Providers should communicate with patients about what specific indicators mean, and how they can mutually create an action plan to eliminate barriers to care or reduce access challenges. The responsibility on the provider is indeed increasing, but it will result in a healthier future for all.
Patient encounter data in EHRs, imaging and lab results, and billing data from claims have generated a venerable sea of complex and disparate information, and organizations strive to harness it to enhance decision-making in the optimal treatment of patients.
This is, in part, why some providers are hesitant to take on yet another data asset, such as social determinants of health (SDOH), and put it in the hands of clinicians who already operate in a world busy with electronic documentation and clinical decision support alerts.
SDOH data garnered great interest in its potential for developing both efficacious and personalized treatment plans. Its value has also been lauded for helping to improve the impact of care by addressing barriers keeping patients from fulfilling their care plan.
Research has shown that medical care alone has a very limited effect on overall population health and could be significantly enhanced by pairing it with approaches that address SDOH. The financial drivers for adoption of SDOH data in care management are just as clear and persuasive.
Within the value-based care paradigm, getting paid requires optimizing performance on numerous quality measures. Data experts and professionals in the patient care stream must think outside the box regarding how to best reach patients who face socioeconomic challenges in their day-to-day life that put them at risk for poor outcomes.
At the same time, providers have had little guidance in how to approach using this data within the existing workflows to help clinicians at the point of care.
Some organizations approach incorporating SDOH by sorting through patient demographic data to identify what factors could help them make clinical predictions. This can be challenging because of the large amount of data and uncertainty regarding what is actually useful because not all SDOH data correlates to health outcomes.
Not all data is created equal. For example, ZIP codes and other basic demographic information obtained from EHRs can be useful for broader population health management efforts, but they only show a narrow view into individual patients’ complex lives outside of the care setting. The demographic data is often not standardized or up-to-date; similarly, surveys often generate incomplete data sets and are a tedious way to fact find on a large scale and are not easy to refresh, nor is self-reported data the most reliable option.
In evaluating SDOH data for use in predictive clinical and analytic models and care management, providers should look for non-clinical data sources that are current, comprehensive and longitudinal. For ease of use, social determinant data should be easily standardized and integrated, regularly updated and consistently linked to patient populations for actionable insights.
Another key criterion for selecting SDOH attributes is ensuring they have been clinically validated against actual healthcare outcomes to confirm their predictive power. Organizations will want to confirm that the SDOH information has either been validated or be prepared to validate it through their own internal analytics groups.
Organizations can implement SDOH data in the form of attributes and scores. SDOH attributes can be used in the provider’s own clinical and analytic models. The second way of utilizing SDOH is through predictive health scores. Scores leverage hundreds of socioeconomic attributes to inform providers about potential barriers to patients’ optimal health and wellness. Scores are particularly useful for those organizations that do not want to build their own analytic models.
Regardless of the type of SDOH input, the data can be used to create a specific health alert to flag an area of concern, or generate “high,” “medium” and “low” risk levels per patient. Additionally, it is possible to identify the top three social determinant factors driving a patient’s risk for more pointed care management solutions. Depending on the patient needs and type of care provided, this could translate to spending a few extra minutes addressing challenges during a clinical visit or prior to a hospital discharge.
Intervention, however, does not necessarily fall on the physician during the clinical visit. In small organizations, doctors may drive assimilation of social determinants data. Medium-sized and larger organizations often hire new staff or initiate advanced training for current employees who are daily participants in the patient care flow. Discharge planners, social workers, health coaches or patient experience officers are now utilizing non-clinical data to identify gaps in care for better patient care. They use it to build a bridge from a healthcare setting to public health and community groups with resources for patients who need them.
These emerging healthcare professionals have access to community- or organization-based support resources and can integrate the SDOH insights into care plans targeting unmet needs. By recognizing the need for, and enabling access to services or interventions, providers are driving better health and wellness outcomes. These programs often relate to accessing healthy food, providing reliable housing or transportation resources, and helping patients manage isolation and loneliness. Partnerships with local community-based or government services will help build networks across—and beyond—the continuum of care.
One health system and insurance company, Geisinger, is already reaping the benefits of its Fresh Food Farmacy initiative, which granted free access to a food pantry and education for specific “food insecure” patients struggling with type 2 diabetes. Before the Farmacy, these members cost Geisinger Health Plans an average of $8,000 to $12,000 per person, per month. The payer-side costs have dropped by two-thirds, on average, across the program, since the intervention.
Geisinger has also seen significant improvements in care outcomes in the form of patients’ cholesterol, blood sugars, and triglycerides, improvements that can lower the chances of heart disease and other vascular complications. By zeroing in on food insecurity in a county with particularly high rates of diabetes, food insecurity, poverty, and unemployment, the provider was able to deliver resources that altered the path of treatment and improved the patient condition.
While using SDOH data to affect patient management is relatively new, the practice is here to stay. To ensure the data is used for maximum good in healthcare through safe, secure, ethical, and collaborative means, organizations can consult industry resources such as “The Guiding Principles for Ethical Use of Social Determinants of Health Data.” Providers should communicate with patients about what specific indicators mean, and how they can mutually create an action plan to eliminate barriers to care or reduce access challenges. The responsibility on the provider is indeed increasing, but it will result in a healthier future for all.
More for you
Loading data for hdm_tax_topic #better-outcomes...