Overcoming data barriers to unlock the full potential of SDoH in Healthcare

A person’s health and physical wellbeing is often attributed to factors like quality of food, levels of activity and genetics. However, studies have established that health and overall quality of life are affected by several other factors such as place of residence, education level, income status etc. which often go unnoticed during clinical evaluations. 

These factors could be sociodemographic like education and employment, behavioural like dietary patterns, physical activity or even psychological like health literacy and stress. These factors, known as the Social Determinants of Health (SDoH) can not only help healthcare providers manage existing health conditions of patients, but also evaluate the possibility of an individual contracting certain diseases even before they enter a medical facility.  

The World Health Organization (WHO) defines SDoH as “the conditions in which people are born, grow, work, live, and age, and the wider set of forces and systems shaping the conditions of daily life”. Studies show that over 80% of an individual’s health and well-being is influenced by social determinants of health.

Importance of SDoH data in healthcare

A McKinsey survey from 2019 revealed people reporting food insecurity were 2.4x times more likely to report multiple ER visits and 2 times more IP visits over a period of 12 months. Similarly, people with unmet transportation need were two times more likely to report an ER visit and 2.2 times more likely to report an IP visit over a period of 12 months.

Community level SDOH data sources e.g. weather information can be used to identify high risk zones where these non-clinical determinants like air pollution etc, are likely to have a significant impact on health outcomes. For example, in case of countries like India, a relationship can be established between pollution levels and health conditions like lung disorders or even chronic kidney disease. It can be determined that local environments where suspended particulate matter is over a certain level, people with asthma, lung affliction or anaemia, are more likely to face negative health consequences. This information can be used to identify at-risk populations and manage, prevent, and restrict the onset of their diseases.

At an individual level, access to SDoH data can enable healthcare providers to provide health related insights, targeted education as well as more personalized treatment plans. Assessment of individual needs like nutrition etc could be obtained through personal interactions, or by leveraging data sources that provide attributes such as income (e.g. data from credit rating organizations). The individual level needs can then be predicted using analytical methods to derive risk scores for individual needs based on a combination of these internal and external data sources.

Thereby healthcare organizations can leverage social determinants of health for preventive care, ensuring better provision of preventive treatments to improve the overall health of the population. 

Social determinants data can enable smarter investments and operational decisions for population health management programs. The insights generated through SDoH can help in designing effective interventions that reflect the needs and assets of the local community.

Top 3 data barriers to manage SDoH 

The benefits of SDoH can only be availed if there is access to compiled, cleansed, synergized, integrated and readily available data for analysis and decision-making. This can be made possible if we can overcome the following barriers -

Data availability: Availability of comprehensive, granulated data at the micro level is a significant challenge in using SDoH to make preventive health models. Information around sociodemographic factors like income, education, or employment etc is relatively more easily available as compared to behavioural or psychological factors. The only reliable methods of obtaining this information is through surveys on direct interactions but these methods are not easily scalable.

Data accuracy: Despite the data being available for socio-demographic factors, accuracy of available data is difficult to prove. As an individual grows personally, it is of important that the SDoH data must be updated accordingly. If healthcare organisations aim to develop truly effective and personalised care plans for their members, ensuring accuracy of data at any point in time is crucial.

Data interoperability: It is often found that different companies use different screening tools and mathematical models/algorithms to gather relevant data. This further hinders the creation of a reliable, complete SDoH database of individuals making it difficult to establish meaningful relationships/correlations between social determinants and actual health outcomes.

Making the best use of available data

Companies across the globe are getting creative in terms of efforts to determine these environmental factors even though direct information is not available. IoT and wearable devices are playing a big part in piecing the jigsaw puzzle together.

Often, available factors can act as indirect indicators. For example, while it’s difficult to determine a person’s nutritional intake without a survey or a direct interaction, but income information is found to correlate well with it and could be used as an indicator of nutrition. Piecing such data together immediately begins to build a picture that helps anticipate healthcare needs at an individual or a community level. 

Organizations are leveraging advances in data sciences and AI/ML models that can help impute information if it is not available or fill up the missing data elements, using whatever little information is available. Z-codes, launched as part of the ICD 10-CM, help to standardise and store nonclinical data. The Fast Healthcare Interoperability Resources or FHIR platform was also launched with a similar intent of facilitating the exchange of healthcare information electronically between independent clinical systems. 

However, these still do not solve the issue of data availability and reliability, which is really at the core of it and can effectively be solved only through continued partnership of community organisations with healthcare payers and providers.

Road ahead

The potential of SDoH to improve health outcomes while reducing healthcare costs is recognized by stakeholders across the healthcare continuum. Health care organizations are increasingly adopting advanced analytical models to identify the right SDoH to focus upon and formulate measurable goals to demonstrate impact. 

However, the key to unlocking the full potential of SDoH lies in overcoming the data barriers by leveraging innovative and standardized approaches to capture social determinant information at a community and individual level. This can help us deliver the right care in the right setting at the right time thereby helping improve health outcomes.

The article is authored by Anupam Gupta, VP - Advanced Analytics, Optum Global Solutions (India)