The data challenges of SDOH, and how to overcome them

Social determinants of health account for up to 80% of health outcomes, while medical care accounts for the balance, according to County Health Rankings.

Experiences with COVID-19 are a stark reminder that unaddressed social determinants and health inequities play significant roles in the comorbidities that have led to countless deaths attributed to the coronavirus since early 2020.

Rahul Sharma, CEO of HSBlox, a social determinants of health platform vendor, says there are several obstacles the healthcare industry must overcome to fully leverage SDOH, including:

  • Disparate and siloed data from multiple sources across industries, agencies and organizations in the public and private sectors, each with its own structures.
  • Roughly 80% of healthcare data being unstructured.
  • In the realm of social determinants, stakeholders that are not subject to HIPAA regulations.

Healthcare IT News interviewed Sharma to dig into these hurdles to SDOH progress and talk about possible ways of overcoming these hurdles.

Q: Social determinants of health account for up to 80% of health outcomes. How has this manifested itself during the COVID-19 pandemic? What needs to be done?

A: We saw vulnerable populations disproportionately impacted by COVID-19, resulting in higher infection rates and deaths. While many of us leveraged telehealth during the pandemic, members of such populations, many with comorbidities, were likely working on site during business hours and unable to make appointments. Even if they could, they often faced technology barriers like lack of an Internet connection to facilitate telehealth visits.

They also faced barriers on the social care side of the equation. Essential workers like grocery clerks, custodians, housekeepers and others weren’t able to work from home. Furthermore, many take public transportation, and others didn’t have the wherewithal for food delivery to their homes, situations that raised their risks for infection.

In addition, they may have had to leave their kids alone to tend to e-learning, not able to afford a caregiver, or they may have forsaken work to care for children. If they did have a caregiver, it may have been an older relative more at risk of infection. Even if schools provided mobile devices for e-learning, they often lacked Internet, or Internet speeds not high enough for learning, impacting their kids.

All of these events speak to the real need of integrating medical and social care so that we understand people holistically in the context of their lives. Doing so will help us prevent a disproportionate impact on vulnerable populations during the next pandemic.

Q: You’ve indicated there are several obstacles the healthcare industry must overcome to fully leverage SDOH. What are a couple of these obstacles and why are they a problem?

A: The current model of healthcare caters to sick care and crisis management as compared to being proactive and preventive in nature. One of the key reasons for this is the data challenge that exists in the industry – not only pertaining to interoperability, but also to digitization of the wealth of data available in unstructured and semi-structured data sets.

If the pandemic has taught us anything, it’s the importance of having quality, verified data accessible to care providers and public officials to help combat the spread of a communicable disease. Even before the pandemic, efforts to capture and integrate SDOH data across the healthcare continuum have been plagued with interoperability issues, culture gaps and lack of coordination.

Specifically, there are several obstacles to data sharing.

The first is disparate and siloed data from multiple sources across industries, agencies and organizations in the public and private sectors, each with its own unique structures, i.e., lack of standardization plus lack of adoption of standards.

Another is that 80% of healthcare data is unstructured. Such data can take the form of clinician electronic notes, patient-reported information such as portal messages, IoMT (Internet of Medical Things) data, images, audio/video recordings, surveys, and transcripts from telehealth visits that often are hard to access.

A third in the area of SDOH is the fact that certain stakeholders are not subject to HIPAA regulations. Therefore there has to be a mechanism for obtaining and managing patient consent for sharing data with and between stakeholders.

Given the importance of SDOH in determining individual and population health outcomes, it’s clear that payers and providers can benefit from a comprehensive and secure medical and social longitudinal health record that captures SDOH data points to connect patients with community resources that address unmet social needs.

Q: How can healthcare provider organizations overcome these obstacles? And what is the role of health IT here?

A: Technology can make it possible to synthesize data sources of different kinds to address their inconsistencies, help identify errors or misreporting, and integrate credible new feeds. Screening tools, meanwhile, can be augmented with external data sets, such as the 12 Dimensions of the Social Environment, for which the Centers for Disease Control and Prevention has created a directory.

The challenges can be overcome with digitization of the data, forming a longitudinal health record for the patient across different data sets for better predictive and risk score analysis, and by ensuring that data is shared in a secure and permissioned basis only.

One way to ensure authorized disclosure and use of SDOH and other healthcare data is to leverage distributed ledger technology for secure and permissioned sharing in near real time, and at a granular level. This approach ensures the reliability of the data and its source(s), providing the transparency necessary for decision-making. It also ensures that the concerned parties are seeing the same data without any need for reconciliation.

Furthermore, any strategy for sharing data across stakeholders must make protected health and other personally identifiable information a top priority. Safeguards can include biometrics data for proper patient identification verification and matching. Data should be encrypted at rest and in transit.

Moreover, the underlying tech must incorporate a consent management capability to ensure patient-permissioned data sharing with an immutable audit trail of disclosures.

Q: If improving individual and population health while reducing healthcare costs requires that healthcare stakeholders capture and leverage social determinants of health data, what are the next steps the industry needs to take to accomplish this goal?

A: Providers, community-based organizations, payers and other stakeholders need to work together to deploy multi-stakeholder collaboration technology with a longitudinal health record to optimize care coordination and population health measures. Delivering SDOH data at the point of care gives clinicians and other caregivers a comprehensive view of the patient, enabling them to approach care holistically and act on it.

In addition, more hospitals and physician practices need to screen for SDOH factors like food insecurity, housing instability, interpersonal violence, and transportation and utility needs. Digitization of unstructured data sets using AI algorithms can provide a wealth of information for enhancing the longitudinal health record of patients. Machine learning can then be applied to build models for patient medical and social risk scores, cost prediction and patient behavior.

Twitter: @SiwickiHealthIT
Email the writer: bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication.

Cold & Flu Diet & Weight Men's health Nutrition, Food Oral Health Sexual Health Skin Sleep Travel Health Women's Health

Leave a Reply

Your email address will not be published. Required fields are marked *