Getting started with targeted public health interventions
COMMENTARY | Data sharing that supports more granular analysis into the social factors that drive health inequities will help public health agencies offer targeted interventions to address them.
According to the Centers for Disease Control and Prevention, nearly 80% of an individual’s overall health is determined by factors outside of medical care, such as social connectedness, socioeconomic status and living conditions. These factors, referred to as social determinants of health (SDOH), are the focus of myriad studies and initiatives, many government led. By understanding the social factors that drive health inequities, public health agencies and their partners can offer targeted interventions to address them.
However, most initiatives face challenges when targeting interventions on social factors because SDOH lack standardized data elements, data element value sets and data exchange standards that would allow organizations to easily identify, conduct outreach and deliver services to specific groups that use multiple health and social services. Furthermore, existing SDOH data is often not collected at the individual level in health care provider settings, but rather spread across multiple government agencies.
While national standards for capturing the information needed to assess SDOH at the individual level in health care provider settings are in development (Gravity Project and HL7 are tackling this via workgroups now), connecting individual-level data across government agencies to support such interventions isn’t going to happen overnight. Still, there are strategies that public health agencies can use to accelerate change. Let’s look at three of them.
1. Dig deeper into demographic data
The sharing of de-identified data—which removes direct identifiers like name and address—stratified by demographics across government agencies is a sound starting point for improved health interventions. By sharing this data, agencies can begin analyses that will help them target and effectively use their limited resources to support those with the greatest needs.
However, such analysis is hampered by several factors, including a lack of consistency in value sets for demographic data that includes ethnicity, sex, primary language, disability as well as sexual orientation/gender identity (SOGI) data. Imagine one dataset breaks race into five categories, and another breaks race into seven categories. Those datasets must be standardized for analysis to take place. Additionally, selecting the data elements that are most relevant to a specific community is also challenging, as they are not universal. Altogether, greater sharing and adoption of data element value sets across jurisdictions will allow agencies to dig deeper into demographic data.
2. Focus on platform flexibility and scalability
To ensure relevant, standardized data can be used as it becomes available, agencies should focus on future-proofing their analytics platform. That means ensuring the platform and underlying data model are flexible enough to support changes to how factors like race, ethnicity and sex are classified and stratified. By choosing a platform that is flexible and scalable, agencies will also be able to apply novel statistical methods and wrangle new datasets. As continuous improvements are made in data availability and ongoing changes to how demographic data is stratified, agencies should look for a platform that can accommodate both. Put another way, officials should not simply choose a platform based on the data their agency has today.
Additionally, agencies should make sure their data platform can easily stratify by subgroups. SDOH data focuses on factors like educational access, neighborhood and social context. But slicing and dicing that same data based on gender, language and other demographic factors is crucial to uncovering disparities. Indeed, stratifying process and outcome measures by subpopulation is a common approach for assessing diversity, inclusion and equity. Ideally, calculations for commonly used demographic stratifications and subgroups should be precalculated for faster platform performance.
3. Engage stakeholders
The right platform lets users across a wide range of government missions—from public safety to health care—analyze demographic trends within and across a variety of data elements. But government agencies should also engage a wide range of partners and community stakeholders as they configure and fine-tune their platform. When Virginia’s Framework for Addiction Analysis and Community Transformation (FAACT) program was rolled out, for instance, project leaders hosted discovery sessions with user groups across agencies and across regions. The only way to ensure a wide variety of user types can benefit from the platform is to have them try it and offer feedback.
As a result, the FAACT platform successfully brought together information from multiple federal, state, local and private organizations to create a better way for leaders on the front lines to address the opioid epidemic in a targeted manner. To successfully engage in SDOH analysis and offer targeted interventions, agencies must also be in continuous conversation with partners and stakeholders.
The bottom line is that health is a complicated matter—there’s no way around it. A wide range of factors, from built environment to education, affect any given individual’s health. Connecting individual-level data across government agencies while integrating SDOH and SOGI data isn’t going to happen overnight. But greater data sharing that supports more granular analysis can and must begin today. The health of residents depends on it.
Sarah Samis is vice president of Public Health Products & Platforms at GCOM.