Improving Data Collection and Analysis to Better Identify and Address Health Inequities Across Different Demographic Groups

The only way to improve health equity is to identify all of the areas of inequity and address each one personally. Just like a sick patient in the hospital, the only way for that patient to improve is to identify what is wrong and then provide treatment for their specific ailment. For patients in the hospital, the way to identify and address the problem is through medical examination and tests. As for health inequities, the answer lies within data collection and analysis.

We have been using these tools (data collection and analysis) for a while now, yet health inequity still remains a large problem. So in order to truly tackle health inequities, it may be time that we take a moment to improve data collection and analysis in our organizations. We reached out to our talented Healthcare IT Today Community and asked them — how can data collection and analysis be improved to better identify and address health inequities across different demographic groups? The following are their answers.

Timi Leslie, Executive Director at Connecting for Better Health
Data sharing is critical for equitable access to healthcare. In California, for example, healthcare organizations, social service providers, and technologists are collaborating in a community sandbox, a virtual testing environment used to validate data capabilities across partners. It also serves as a growing reference model for high-priority use cases to improve care coordination and ensure inclusive and complete information for whole-person care.

Diana Zuskov, MPH, Associate Vice President of Healthcare Strategy & Innovation at LexisNexis Risk Solutions
Patient-centered care, use of real-world data, and interoperable data exchange create a perfect opportunity to redefine how we collect and exchange health equity insights to drive more equitable care while maintaining patient privacy and limiting provider burden.

We can’t keep expecting providers to build trust, conduct clinical care, and evaluate complex social needs – especially in underserved areas that are already facing provider shortages and post-pandemic burnout. Reimbursing providers for social needs screening is a critical step in the right direction; however, the limited adoption of social needs screenings and coding shows that adding more for providers isn’t always sustainable or scalable.

Understanding social needs is all about what’s happening outside of the healthcare context. It’s time for healthcare organizations and policymakers to similarly rethink diverse sources of data collection outside the clinical encounter. This could look like ongoing digital health interventions, data exchange with community needs assessments and social needs organizations, or leveraging external data sources to augment clinical insights.

Mary Sirois, Managing Director, Transformation & Innovation at Nordic Consulting
To improve data collection and analysis, organizations should focus on obtaining SDoH data indirectly and efficiently. Organizations should also make it a priority to ensure that bias is not introduced through their AI algorithms. For example, instead of asking sensitive questions like “Do you have trouble paying bills?” during a health risk assessment in patient portals, healthcare providers can use alternative data sources and frame questions to focus on actionable solutions, such as offering financial assistance options. By analyzing SDoH data thoughtfully, organizations can segment patient populations more effectively. They can prioritize high-risk individuals for in-person care while engaging lower-risk patients through virtual care. This helps optimize resource allocation and improve equity in access to care.

Leah Dewey, MPH, Vice President of Clinical and Consumer Engagement Operations at Cotiviti
Building trust with members by communicating with them throughout the year, rather than simply during enrollment, is critical to improving response rates and optimizing data collection. Healthcare organizations can execute digital surveys year-round to gather member data, update contact information, and understand communication preferences. By utilizing this data, healthcare organizations can better target underserved areas and meet members’ communication needs by offering correspondence not only in a variety of languages. In addition, communications should also be tailored to match the culture of members, literacy of recipients, and preferred communication methods such as text, email, or phone.

Sandeep Wadhwa, MD, MBA, Global Chief Medical Officer at Solventum
Improving data collection starts with tracking patients longitudinally—following them across episodes of care rather than viewing isolated “events.” By integrating expanded Z codes to document social risks, we gain a richer and fuller understanding of the patient journey and the broader factors influencing their health. Standardizing coding practices and leveraging grouping technology allow us to document and analyze the social factors impacting health more effectively. This approach not only highlights health inequities but also enables organizations to measure outcomes like preventable complications, readmissions, and mortality. These metrics, combined with cost-per-stay and other efficiency measures, provide a framework for defining episodes of care and driving meaningful improvements in population health.

Dana McCalley, VP of Value-Based Care at Navina
Improving SDoH data collection and analysis begins with clearly defining roles and responsibilities within the healthcare workflow. It’s vital to determine whether data is collected during patient check-ins, through interactions with medical assistants, or via automated systems. Consistency in collection methods ensures that the data is reliable and actionable.

Additionally, SDoH information must be recorded as discrete data points rather than free text to facilitate meaningful analysis. Sensitivity is paramount during the data collection process. Many SDoH questions touch on deeply personal issues, such as domestic violence, incarceration history, or housing insecurity. Patients must feel assured that their disclosures will be handled with sensitivity and confidentiality, without fear of judgment or repercussions. When patients trust the system, they are more likely to provide honest and accurate information.

Once collected, this data enables organizations to identify specific health inequities and tailor interventions that address the unique needs of various demographic groups. For example, a community struggling with housing insecurity might benefit from partnerships with local housing authorities or nonprofits.

Brian Jones, Chief Growth Officer at Avēsis
Accurate and actionable data is the backbone of tackling health inequities. Proper risk stratification requires leveraging robust datasets, implementing outcomes-based interventions tailored by demographics, and integrating SDoH metrics into electronic health records. These efforts allow healthcare organizations to uncover gaps, target interventions effectively, and move beyond one-size-fits-all approaches.

Samir Awasthi, MD, PhD, Vice President of Clinical Research at Anumana
In healthcare today, we collect vast amounts of untapped data from diagnostic tools like electrocardiograms (ECGs) and imaging. AI is transforming care by enabling clinicians to leverage this data for faster, more effective decisions, accelerating patient access to the right treatments.

To achieve health equity, it’s essential to make these innovations accessible in underserved communities where advanced diagnostic tools are often limited. ECGs are fit for this purpose – they are widely accessible, standardized, and routinely used. AI-powered ECG solutions can reach deep into communities to enable earlier detection of cardiovascular disease, often missed until advanced stages that require costly treatments.

The recent CMS OPPS payment decision for AI-enabled ECG technology to detect low ejection fraction – a key yet often silent marker of heart failure – is a promising milestone. The ruling paves the way for equitable access to lifesaving technology by addressing financial barriers to adoption. To close diagnostic gaps, payors must appropriately support the use of tools like ECG-AI in all care settings, ensuring healthier futures for patients regardless of their zip code.

Dominique Demolle, Co-Founder and CEO at Cognivia
With the mandate to implement Diversity Action Plans in the clinical development of new drugs or medical devices, the first step toward inclusivity has been made. However, accessing patients and their data remains one of the biggest hurdles, primarily due to socio-economic factors and a lack of clinical research awareness among different communities. Identifying clinical sites close to these communities and involving them in patient recruitment and engagement will be essential. Providing affordable collection devices and platforms, and ensuring data interpretation accounts for the increased variability due to population heterogeneity, will be crucial. There is no doubt that data modeling and predictive algorithms will be more necessary than ever to understand sources of variability and to match real-world data as early as possible.

Krinx Kong, Chief Commercial Officer at Cognivia
We can enhance our efforts by utilizing inclusive data ecosystems that create AI platforms combining various data sources, including social determinants of health and community information. It’s crucial to develop real-time dashboards that monitor health disparities and highlight the needs of underrepresented groups. Additionally, partnering with community organizations can help design data collection tools that capture marginalized populations’ cultural and socioeconomic nuances. Finally, implementing regular audits of AI models is crucial to eliminating biases that may exacerbate healthcare disparities.

Chris Althoff, Former Senior Consulting Partner and Head of Marketing at emtelligent
As healthcare organizations, providers, and payers continue deploying AI technologies to support more scalable, sustainable, and accessible health systems health equity has become a top priority. Medically aligned AI platforms that account for inherent bias in their training data will be better able to analyze and integrate data across socioeconomic, geographic, and demographic lines are key to delivering real-time personalized insights that enable tailored interventions and bridge critical gaps in care for all individuals – regardless of where they live, their income, race, gender, or religion. In doing so, AI is helping providers, payers, and other stakeholders build an equitable healthcare system that leaves no patient behind.

Kota Kubo, Co-Founder and Co-CEO at Ubie
AI is seeing increasing trust and uptake by health systems, giving them the ability to process more patient data than ever. This puts AI in a prime position to help identify and rectify health disparities that might have otherwise been overlooked. However, whether this potential can be realized will depend on if health systems demand that the solutions they use are specifically trained on social determinants of health. The industry needs to grow beyond models that simply replicate the status quo. AI has capabilities far beyond trying to mimic a human doctor. If we don’t take advantage of its processing and analysis capabilities, health disparities and inequity will remain just as they are now. So, the onus is on us to imagine a better world and make the most of this technology’s potential.

Ankit Jain, Co-Founder and CEO at Infinitus Systems
AI has the transformative potential to address health equity by identifying and mitigating disparities in care delivery and outcomes. However, to unlock this potential, we must prioritize the ethical use of AI, ensuring that it amplifies equity rather than perpetuating existing biases. Here are some key ways in which AI can help address health equity:

• Proactive Identification of Health Disparities: AI can analyze vast datasets from electronic health records, social determinants of health (SDoH), and claims data to uncover inequities in care access, treatment quality, and outcomes across demographic groups; these insights can empower health systems to target underserved communities with tailored interventions
• Scalable Collection of SDoH Data: One of the biggest gaps in addressing health equity is the lack of comprehensive SDoH data in the ecosystem; AI-powered tools can analyze unstructured data from clinical notes, patient surveys, and public health databases to extract meaningful SDoH insights; machine learning models can further identify patterns in socioeconomic factors, housing instability, food insecurity, or transportation barriers, enabling scalable and actionable data collection efforts; when source data is missing, voice AI agents can collect such data at scale in ways not possible before
• Personalized Interventions: By integrating SDoH and clinical data, AI can recommend personalized care plans that address barriers to care, such as transportation challenges, food insecurity, or language barriers, ensuring more equitable patient support
• Improving Resource Allocation: AI can optimize resource distribution by predicting where healthcare services are most needed, ensuring underserved populations have better access to care

So many fantastic answers here! Huge thank you to Timi Leslie, Executive Director at Connecting for Better Health, Diana Zuskov, MPH, Associate Vice President of Healthcare Strategy & Innovation at LexisNexis Risk Solutions, Mary Sirois, Managing Director, Transformation & Innovation at Nordic Consulting, Leah Dewey, MPH, Vice President of Clinical and Consumer Engagement Operations at Cotiviti, Sandeep Wadhwa, MD, MBA, Global Chief Medical Officer at Solventum, Dana McCalley, VP of Value-Based Care at Navina, Brian Jones, Chief Growth Officer at Avēsis, Samir Awasthi, MD, PhD, Vice President of Clinical Research at Anumana, Dominique Demolle, Co-Founder and CEO at Cognivia, Krinx Kong, Chief Commercial Officer at Cognivia, Chris Althoff, Former Senior Consulting Partner and Head of Marketing at emtelligent, Kota Kubo, Co-Founder and Co-CEO at Ubie, and Ankit Jain, Co-Founder and CEO at Infinitus Systems for taking the time out of your day to submit a quote! And thank you to all of you for taking the time out of your day to read this article! We could not do this without all of your support.

How do you think data collection and analysis can be improved to better identify and address health inequities across different demographic groups? Let us know over on social media, we’d love to hear from all of you!

The only way to improve health equity is to identify all of the areas of inequity and address each one personally. Just like a sick patient in the hospital, the only way for that patient to improve is to identify what is wrong and then provide treatment for their specific ailment. For patients in the hospital, the way to identify and address the problem is through medical examination and tests. As for health inequities, the answer lies within data collection and analysis.

We have been using these tools (data collection and analysis) for a while now, yet health inequity still remains a large problem. So in order to truly tackle health inequities, it may be time that we take a moment to improve data collection and analysis in our organizations. We reached out to our talented Healthcare IT Today Community and asked them — how can data collection and analysis be improved to better identify and address health inequities across different demographic groups? The following are their answers.

Timi Leslie, Executive Director at Connecting for Better Health
Data sharing is critical for equitable access to healthcare. In California, for example, healthcare organizations, social service providers, and technologists are collaborating in a community sandbox, a virtual testing environment used to validate data capabilities across partners. It also serves as a growing reference model for high-priority use cases to improve care coordination and ensure inclusive and complete information for whole-person care.

Diana Zuskov, MPH, Associate Vice President of Healthcare Strategy & Innovation at LexisNexis Risk Solutions
Patient-centered care, use of real-world data, and interoperable data exchange create a perfect opportunity to redefine how we collect and exchange health equity insights to drive more equitable care while maintaining patient privacy and limiting provider burden.

We can’t keep expecting providers to build trust, conduct clinical care, and evaluate complex social needs – especially in underserved areas that are already facing provider shortages and post-pandemic burnout. Reimbursing providers for social needs screening is a critical step in the right direction; however, the limited adoption of social needs screenings and coding shows that adding more for providers isn’t always sustainable or scalable.

Understanding social needs is all about what’s happening outside of the healthcare context. It’s time for healthcare organizations and policymakers to similarly rethink diverse sources of data collection outside the clinical encounter. This could look like ongoing digital health interventions, data exchange with community needs assessments and social needs organizations, or leveraging external data sources to augment clinical insights.

Mary Sirois, Managing Director, Transformation & Innovation at Nordic Consulting
To improve data collection and analysis, organizations should focus on obtaining SDoH data indirectly and efficiently. Organizations should also make it a priority to ensure that bias is not introduced through their AI algorithms. For example, instead of asking sensitive questions like “Do you have trouble paying bills?” during a health risk assessment in patient portals, healthcare providers can use alternative data sources and frame questions to focus on actionable solutions, such as offering financial assistance options. By analyzing SDoH data thoughtfully, organizations can segment patient populations more effectively. They can prioritize high-risk individuals for in-person care while engaging lower-risk patients through virtual care. This helps optimize resource allocation and improve equity in access to care.

Leah Dewey, MPH, Vice President of Clinical and Consumer Engagement Operations at Cotiviti
Building trust with members by communicating with them throughout the year, rather than simply during enrollment, is critical to improving response rates and optimizing data collection. Healthcare organizations can execute digital surveys year-round to gather member data, update contact information, and understand communication preferences. By utilizing this data, healthcare organizations can better target underserved areas and meet members’ communication needs by offering correspondence not only in a variety of languages. In addition, communications should also be tailored to match the culture of members, literacy of recipients, and preferred communication methods such as text, email, or phone.

Sandeep Wadhwa, MD, MBA, Global Chief Medical Officer at Solventum
Improving data collection starts with tracking patients longitudinally—following them across episodes of care rather than viewing isolated “events.” By integrating expanded Z codes to document social risks, we gain a richer and fuller understanding of the patient journey and the broader factors influencing their health. Standardizing coding practices and leveraging grouping technology allow us to document and analyze the social factors impacting health more effectively. This approach not only highlights health inequities but also enables organizations to measure outcomes like preventable complications, readmissions, and mortality. These metrics, combined with cost-per-stay and other efficiency measures, provide a framework for defining episodes of care and driving meaningful improvements in population health.

Dana McCalley, VP of Value-Based Care at Navina
Improving SDoH data collection and analysis begins with clearly defining roles and responsibilities within the healthcare workflow. It’s vital to determine whether data is collected during patient check-ins, through interactions with medical assistants, or via automated systems. Consistency in collection methods ensures that the data is reliable and actionable.

Additionally, SDoH information must be recorded as discrete data points rather than free text to facilitate meaningful analysis. Sensitivity is paramount during the data collection process. Many SDoH questions touch on deeply personal issues, such as domestic violence, incarceration history, or housing insecurity. Patients must feel assured that their disclosures will be handled with sensitivity and confidentiality, without fear of judgment or repercussions. When patients trust the system, they are more likely to provide honest and accurate information.

Once collected, this data enables organizations to identify specific health inequities and tailor interventions that address the unique needs of various demographic groups. For example, a community struggling with housing insecurity might benefit from partnerships with local housing authorities or nonprofits.

Brian Jones, Chief Growth Officer at Avēsis
Accurate and actionable data is the backbone of tackling health inequities. Proper risk stratification requires leveraging robust datasets, implementing outcomes-based interventions tailored by demographics, and integrating SDoH metrics into electronic health records. These efforts allow healthcare organizations to uncover gaps, target interventions effectively, and move beyond one-size-fits-all approaches.

Samir Awasthi, MD, PhD, Vice President of Clinical Research at Anumana
In healthcare today, we collect vast amounts of untapped data from diagnostic tools like electrocardiograms (ECGs) and imaging. AI is transforming care by enabling clinicians to leverage this data for faster, more effective decisions, accelerating patient access to the right treatments.

To achieve health equity, it’s essential to make these innovations accessible in underserved communities where advanced diagnostic tools are often limited. ECGs are fit for this purpose – they are widely accessible, standardized, and routinely used. AI-powered ECG solutions can reach deep into communities to enable earlier detection of cardiovascular disease, often missed until advanced stages that require costly treatments.

The recent CMS OPPS payment decision for AI-enabled ECG technology to detect low ejection fraction – a key yet often silent marker of heart failure – is a promising milestone. The ruling paves the way for equitable access to lifesaving technology by addressing financial barriers to adoption. To close diagnostic gaps, payors must appropriately support the use of tools like ECG-AI in all care settings, ensuring healthier futures for patients regardless of their zip code.

Dominique Demolle, Co-Founder and CEO at Cognivia
With the mandate to implement Diversity Action Plans in the clinical development of new drugs or medical devices, the first step toward inclusivity has been made. However, accessing patients and their data remains one of the biggest hurdles, primarily due to socio-economic factors and a lack of clinical research awareness among different communities. Identifying clinical sites close to these communities and involving them in patient recruitment and engagement will be essential. Providing affordable collection devices and platforms, and ensuring data interpretation accounts for the increased variability due to population heterogeneity, will be crucial. There is no doubt that data modeling and predictive algorithms will be more necessary than ever to understand sources of variability and to match real-world data as early as possible.

Krinx Kong, Chief Commercial Officer at Cognivia
We can enhance our efforts by utilizing inclusive data ecosystems that create AI platforms combining various data sources, including social determinants of health and community information. It’s crucial to develop real-time dashboards that monitor health disparities and highlight the needs of underrepresented groups. Additionally, partnering with community organizations can help design data collection tools that capture marginalized populations’ cultural and socioeconomic nuances. Finally, implementing regular audits of AI models is crucial to eliminating biases that may exacerbate healthcare disparities.

Chris Althoff, Former Senior Consulting Partner and Head of Marketing at emtelligent
As healthcare organizations, providers, and payers continue deploying AI technologies to support more scalable, sustainable, and accessible health systems health equity has become a top priority. Medically aligned AI platforms that account for inherent bias in their training data will be better able to analyze and integrate data across socioeconomic, geographic, and demographic lines are key to delivering real-time personalized insights that enable tailored interventions and bridge critical gaps in care for all individuals – regardless of where they live, their income, race, gender, or religion. In doing so, AI is helping providers, payers, and other stakeholders build an equitable healthcare system that leaves no patient behind.

Kota Kubo, Co-Founder and Co-CEO at Ubie
AI is seeing increasing trust and uptake by health systems, giving them the ability to process more patient data than ever. This puts AI in a prime position to help identify and rectify health disparities that might have otherwise been overlooked. However, whether this potential can be realized will depend on if health systems demand that the solutions they use are specifically trained on social determinants of health. The industry needs to grow beyond models that simply replicate the status quo. AI has capabilities far beyond trying to mimic a human doctor. If we don’t take advantage of its processing and analysis capabilities, health disparities and inequity will remain just as they are now. So, the onus is on us to imagine a better world and make the most of this technology’s potential.

Ankit Jain, Co-Founder and CEO at Infinitus Systems
AI has the transformative potential to address health equity by identifying and mitigating disparities in care delivery and outcomes. However, to unlock this potential, we must prioritize the ethical use of AI, ensuring that it amplifies equity rather than perpetuating existing biases. Here are some key ways in which AI can help address health equity:

• Proactive Identification of Health Disparities: AI can analyze vast datasets from electronic health records, social determinants of health (SDoH), and claims data to uncover inequities in care access, treatment quality, and outcomes across demographic groups; these insights can empower health systems to target underserved communities with tailored interventions
• Scalable Collection of SDoH Data: One of the biggest gaps in addressing health equity is the lack of comprehensive SDoH data in the ecosystem; AI-powered tools can analyze unstructured data from clinical notes, patient surveys, and public health databases to extract meaningful SDoH insights; machine learning models can further identify patterns in socioeconomic factors, housing instability, food insecurity, or transportation barriers, enabling scalable and actionable data collection efforts; when source data is missing, voice AI agents can collect such data at scale in ways not possible before
• Personalized Interventions: By integrating SDoH and clinical data, AI can recommend personalized care plans that address barriers to care, such as transportation challenges, food insecurity, or language barriers, ensuring more equitable patient support
• Improving Resource Allocation: AI can optimize resource distribution by predicting where healthcare services are most needed, ensuring underserved populations have better access to care

So many fantastic answers here! Huge thank you to Timi Leslie, Executive Director at Connecting for Better Health, Diana Zuskov, MPH, Associate Vice President of Healthcare Strategy & Innovation at LexisNexis Risk Solutions, Mary Sirois, Managing Director, Transformation & Innovation at Nordic Consulting, Leah Dewey, MPH, Vice President of Clinical and Consumer Engagement Operations at Cotiviti, Sandeep Wadhwa, MD, MBA, Global Chief Medical Officer at Solventum, Dana McCalley, VP of Value-Based Care at Navina, Brian Jones, Chief Growth Officer at Avēsis, Samir Awasthi, MD, PhD, Vice President of Clinical Research at Anumana, Dominique Demolle, Co-Founder and CEO at Cognivia, Krinx Kong, Chief Commercial Officer at Cognivia, Chris Althoff, Former Senior Consulting Partner and Head of Marketing at emtelligent, Kota Kubo, Co-Founder and Co-CEO at Ubie, and Ankit Jain, Co-Founder and CEO at Infinitus Systems for taking the time out of your day to submit a quote! And thank you to all of you for taking the time out of your day to read this article! We could not do this without all of your support.

How do you think data collection and analysis can be improved to better identify and address health inequities across different demographic groups? Let us know over on social media, we’d love to hear from all of you!