Proximity Matters: Using machine learning and geospatial analytics to reduce COVID-19 exposure risk

By Manjula Julka, MD, MBA, PCCI’s Vice President, Clinical Innovation

By Albert Karam, MS, PCCI’s Director of Data Governance and Analytics

Since the earliest days of the COVID-19 pandemic, one of the biggest challenges for health systems has been to gain an understanding of the community spread of this virus and to determine how likely is it that a person walking through the doors of a facility is at a higher risk of being COVID-19 positive.

Without adequate access to testing data, health systems early-on were often forced to rely on individuals to answer questions such as whether they had traveled to certain high-risk regions. Even that unreliable method of assessing risk started becoming meaningless as local community spread took hold.

Parkland Health & Hospital System (the safety net health system for Dallas County, TX) and PCCI (a Dallas, TX based non-profit with expertise in the practical applications of advanced data science and social determinants of health) had a better idea. Community spread of an infectious disease is made possible through physical proximity and density of active carriers and non-infected individuals. Thus, to understand the risk of an individual contracting the disease (exposure risk), it was necessary to assess their proximity to confirmed COVID-19 cases based on their address and population density of those locations. If an “exposure risk” index could be created, then Parkland could use it to minimize exposure for their patients and health workers and provide targeted educational outreach in highly vulnerable zip codes.

PCCI’s data science and clinical team worked diligently in collaboration with the Parkland Informatics team to develop an innovative machine learning driven predictive model called Proximity Index. Proximity Index predicts for an individual’s COVID-19 exposure risk, based on their proximity to test positive cases and the population density. This model was put into action at Parkland through PCCI’s cloud-based advanced analytics and machine learning platform called Isthmus. PCCI’s machine learning engineering team generated geospatial analysis for the model and, with support from the Parkland IT team, integrated it with their Electronic Health Record system.

Since April 22, Parkland’s population health team has utilized the Proximity Index for four key system-wide initiatives to triage more than 100,000 patient encounters and to assess needs, proactively:

  1. Patients most at risk, with appointments in 1-2 days, were screened ahead of their visit to prevent spread within hospital
  2. Patients identified as vulnerable, were offered additional medical (i.e. virtual visit, medication refill assistance) and social support
  3. Communities, by zip-code, most at-risk were sent targeted messaging and focused outreach on COVID-19 prevention, staying safe, monitoring for symptoms, and resources for where to get tested and medical help.
  4. High exposure risk patients who had an appointment at one of Parkland’s community clinics in the next couple of days were offered a telehealth appointment instead of a physical appointment if that was appropriate based on the type of appointment.

5. Proximity Index to other organizations in the community – schools, employers etc., as well as to individuals to provide them with a data driven tool to help in decision making around reopening the economy and society in a safe, thoughtful manner.

Figure 1 PCCI’s Proximity Index Process

In the future, PCCI is planning on offering

Many teams across the Parkland family collaborated on this project, including the IT team led by Brett Moran, MD, Senior Vice President, Associate Chief Medical Officer and Chief Medical Information Officer at Parkland Health and Hospital System.

About the Authors

Manjula Julka, MD, FAAFP, MBA, is the Vice President of Clinical Innovation at PCCI. She brings more than 15 years of experience in healthcare delivery transformation, leading a strong and consistent track record of enabling meaningful outcomes.

Albert Karam is a data scientist at PCCI with experience building predictive models in healthcare. While working at PCCI, Albert has researched, identified, managed, modeled, and deployed predictive models for Parkland Hospital and the Parkland Community Health Plan. He is diverse in understanding modelling workflows and implementation of real time models.



WEBCAST: New partnership announced between PCCI & Healthbox to lead SDOH innovation

A HIMSS Learning webcast introduces a partnership between Healthbox & PCCI, an alliance formed to drive innovation leveraging #SDOH. In the webcast, healthcare leaders and long-term friends Neil Patel, from Healthbox, and Steve Miff, from PCCI, discuss the exciting partnership between PCCI and Healthbox and SDOH trends. Click on the image below to view the webcast:

Please have a look at the full set of HIMSS Webcasts featuring PCCI and Healthbox discussing how to implement SDOH principles via connected communities of care:




PCCI’s Vulnerability Index shows declines in Dallas County’s COVID-19 risk in August

DALLAS – Parkland Center of Clinical Innovations’ (PCCI) Vulnerability Index has noted a small, but important decline in COVID-19 risk for much of Dallas County between July and August.

Launched in June, PCCI’s Vulnerability Index determines communities at risk by examining comorbidity rates, including chronic illnesses such as hypertension, cancer, diabetes and heart disease; areas with high density of populations over the age of 65; and increased social deprivation such as lack of access to food, medicine, employment and transportation.

Based on tracking its key factors, including changes in confirmed

The most at-risk ZIP code continues to be 75211, around Cockrell Hill. This area was the most as risk area for both July and August.

COVID-19 cases and in mobility, the PCCI Vulnerability Index saw a modest decrease in the COVID-19 risk due to decreased COVID-19 cases across. However, several zip codes showed relative increases in their risk due to increased mobility compared to the same period in 2019 and a higher relative proportion of COVID-19 cases in their area compared to July.

The ZIP code with the biggest jump in vulnerability to COVID-19 infection by the end of August was 75217, the region that includes Pleasant Grove, intersected by Loop 12 and U.S. Highway 175. This ZIP code jumped from 5.3 percent vulnerability rating to 93.5, making it one of the most vulnerable ZIP codes in Dallas County.


The ZIP code 75230, north of Walnut Hill Lane between U.S. Highway 75 and Dallas North Tollway, and 75218 immediately east of White Rock Lake, saw the largest drops in their vulnerability risk going from 45.6 to 12 and 38.9 to 16, respectively.

“The data offered by our vulnerability index is showing that Dallas County is taking incremental, but i

mportant steps in managing COVID-19 risks,” Thomas Roderick, PhD, Senior Data and Applied Scientist at PCCI. “The changes we are witnessing are generally based on two important dynamic elements: recent confirmed COVID-19 cases, and mobility that includes people moving and gathering. For much of the county the number of confirmed COVID-19 cases began to slow in August, compared to July. With no holidays in August, there were fewer large gatherings and travel. These two factors can account for overall decrease across the county. Those areas experiencing increases in vulnerability ratings are largely attributed to yearly increases in mobility and to ongoing COVID-19 hotspots in a particular ZIP code.”

While August’s vulnerability index for Dallas County provides positive signs, Dr. Roderick warns that the Labor Day holiday, school openings, and sporting events may contribute to increases in vulnerability and COVID-19 cases. Additionally, he stresses the importance of personal behavior, such as wearing a mask, social distancing, and following the CDC guidelines.

July 2020

The PCCI COVID-19 Vulnerability Index can be found on its COVID-19 Hub for Dallas County at:

Data Sources:

To build Vulnerability Index, PCCI relied on data from Parkland Health & Hospital System, Dallas County Health and Human Services Department, the Dallas-Fort Worth Hospital Council,   U.S. Census, and SafeGraph.


About Parkland Center for Clinical Innovation

August 2020

Parkland Center for Clinical Innovation (PCCI) is an independent, not-for-profit, healthcare intelligence organization affiliated with Parkland Health & Hospital System. PCCI leverages clinical expertise, data science and social determinants of health to address the needs of vulnerable populations. We believe that data, done right, has the power to galvanize communities, inform leaders, and empower people.




New England Journal of Medicine/Catalyst: Connected Communities of Care in Times of Crisis

The New England Journal of Medicine/Catalyst has published an article from co-authors Keith Kosel, Vice President, Enterprise Relationships at PCCI and David Nash, MD, Founding Dean Emeritus at Jefferson College of Population Health, about how the integration and cooperation among health care organizations that provide clinical care and community-based organizations that address social determinants of health, is of growing importance and can be especially useful during COVID19. NEJM Catalyst is a product of NEJM Group, a division of the Massachusetts Medical Society.
To read the entire article, please click the image below:

Is Your Community Ready to be Connected?

By Keith C. Kosel, PhD, MHSA, MBA

Vice President, Enterprise Relationships

This question initially brings to mind many possibilities such as connection to the latest 5G cellular service, a new super-fast internet provider, or maybe one of the many new energy suppliers jockeying for market share from traditional utility companies. While all of these might represent legitimate opportunities to improve one’s community, here we are talking about a different concept; specifically, whether your community is ready to have a Connected Community of Care (CCC) to advance whole person health.

The image of a CCC may seem obvious. After all, we all live in communities where we have some connections between hospitals, physician practices, ambulatory care centers, and pharmacies to name just a few. But here we are talking about a broader sense of connected community that includes not just health care organizations, but social service organizations, such as schools and civic organizations and community-based organizations (CBOs) like neighborhood food pantries and temporary housing facilities. A true CCC links together local healthcare providers along with a wide array of CBOs, faith-based organizations and civic entities to help address those social factors, such as education, income security, food access, and behavioral support networks, which can influence a population’s risk for illness or disease. Addressing these factors in connection with traditional medical care can reduce disease risk and advance whole person care. Such is the case in Dallas Texas, where the Dallas CCC information exchange platform has been operating since 2012. Designed to electronically bring together local healthcare systems, clinicians, and ancillary providers with over a hundred CBOs, the Dallas CCC provides a real-time referral and communication platform with a sophisticated care management system designed and built by the Parkland Center for Clinical Innovation (PCCI) and Pieces Technologies, Inc.

Long before this information exchange platform was implemented, the framers of the Dallas CCC came together to consider whether Dallas needed such a network and whether the potential partners in the community were truly ready to make the commitments needed to bring this idea to fruition. As more and more communities and healthcare provider entities realize the tremendous potential of addressing the social determinants of health by bringing together healthcare entities and CBOs and other social-service organizations, the question of community readiness for a CCC is being asked much more often. But how do you know what the right answer is?

Before looking at the details of how we might answer this, let’s remember that a CCC doesn’t don’t just happen in a vacuum. It requires belief, vision, commitment― and above all― alignment among the key stakeholders. Every CCC that has formed, including the Dallas CCC, begins with a vision for a healthier community and its citizens. This vision is typically shared by two or more large and influential key community stakeholders, such as a   large healthcare system, school district, civic entity, or social- service organization like the United Way or Salvation Army. Leaders from these organizations often initially connect at informal social gatherings and advance the idea of what if? These informal exchanges soon lead to a more formal meeting where the topic is more fully discussed and each of the participants articulates their vision for a healthier community and what that might look like going forward. This stage in the evolution of a CCC is perhaps the key step in the transformation process, as while all stakeholders will have a vision, achieving alignment among those visions is no small feat. Many hopeful CCCs never pass this stage, as the stakeholders cannot come to agreement on a common vision that each can support. For the fortunate few, intrinsic organizational differences can be successfully set aside to allow the CCC to move forward.

It’s at this point in the CCC’s evolution that details begin to matter in truthfully answering the question, “Is this community ready to be connected?” While there may be agreement among the key stakeholders on a vision, the details around readiness may still divert or delay the best-laid plans. It is safe to say that the key to understanding a community’s readiness to form a CCC lies in the completion of a formal, comprehensive, and transparent readiness assessment. A readiness assessment is a process to collect, analyze, and evaluate critical information gathered from the community to help identify actual clinical and socio-economic needs, current capabilities and resources (including technology), and community interest and engagement. Taken together, a comprehensive readiness assessment can help identify a community’s strengths and weaknesses in preparation for establishing a CCC. A readiness assessment is not a tactical plan for building a CCC, nor is it a governance document that provides how all members of the CCC will relate to each other. Instead, the readiness assessment provides communities interested in establishing a CCC with an honest and unbiased yardstick to measure preparedness. Conducting and using the results of the readiness assessment is one of the best ways to ensure a successful CCC deployment.

A typical CCC readiness assessment covers five areas: (1) community demographics; (2) clinical areas of need (including trends); (3) social areas of need (including trends); (4) technology competency (e.g., what percent of the potential network participants are computer literate?), availability (e.g., what percent of the potential network participants have internet access?), and suitability (e.g., is the internet access, high speed?); and (5) what are the needs of potential network participants and can these be modeled as use cases for the information exchange network? This information is essential to help key stakeholder decision-makers decide to move forward with establishing a CCC and to know what specific challenges may lie ahead.

The collection of this essential information can be done in a number of ways, such as making use of existing publicly reported data or conducting surveys, interviews, focus groups and townhall meetings with community leaders and residents and clinical and CBO leaders and staff. Experience conducting the readiness assessment that provided the foundation for the Dallas CCC showed that no single information-collection method was sufficient to collect the necessary level and robustness of the data. In Dallas, we utilized all five approaches but found that in addition to researching publicly available data, initial surveys, followed by interviews and focus groups, yielded the most voluminous and reliable information to chart the course ahead.

In addition to the various methods to collect this essential information, the key to obtaining useful and reliable information requires a sufficient number of respondents/participants who are drawn from various organizations and organizational levels. Simply put, you must have a large enough sample and you must have diversity within the sample. It’s not enough to just interview leaders of potential network participants, as their understanding of the needs, trends, and capabilities may look very different from that of front-line staff. Similarly, surveying only one category of potential network participants may not provide enough information to  fully understand the socio-economic needs in the community or even the perspectives surrounding the prevalence of chronic conditions. Beyond the qualitative methods involved, it is important to note that if done right, this process takes a lot of time to complete. Cutting corners by reducing the sample size, for example, or doing selective sampling to speed the readiness assessment process along will only cause problems later when this insufficient information results in erroneous decision-making.

Once the data has been collected, it is important to carefully analyze what the data is trying to tell you. Results of the readiness assessment must be shared openly and honestly with all key stakeholders, particularly those serving in a governance capacity. The governance group (a topic for another day) that has formed in parallel with the readiness assessment must be able to evaluate and understand the main messages from the readiness assessment to make an informed decision as to whether to move forward with establishing a CCC. Like the need for alignment around the key stakeholder’s vision for the CCC, there must be universal agreement by the key stakeholders as to the message of the readiness assessment and its implications for the road ahead. As with the vision alignment stage, substantive disagreements among the group at this stage are a sign of trouble ahead unless differences can be resolved.

At this point you might be thinking that this all seems very complicated and fraught with potential land mines waiting to derail your effort to answer the original question “Is your community ready to be connected?” Again, I would emphasize the importance of unwavering commitment and alignment to achieve the vision. But I would also offer advice gleaned from working in the CCC space for the last eight years, which is to get help early and don’t wait until the horse is out of the barn! We have seen first-hand many communities and consultants approach the conduct of a readiness assessment with a cavalier attitude, often exemplified by the statement, “we already know all of this,” only later to have to backtrack their pronouncements at substantial additional cost in time and resources. Fortunately, today there are a number of excellent organizations, including PCCI, with the experience, credibility, and integrity in the CCC space to help you on this journey. Don’t be afraid to seek them out. It will be a wise investment that you will not regret, particularly when you begin to see the results of improved whole person health and well-being in your community.

About the author

Dr. Keith Kosel is a Vice President at Parkland Center for Clinical Innovation (PCCI) and is author of Building Connected Communities of Care: The Playbook for Streamlining Effective Coordination Between Medical and Community-Based Organizations,” a guide that brings together communities to support our most vulnerable. At PCCI, Keith is leveraging his passion for – and extensive experience in – patient safety, quality, and population health by focusing on understanding social determinants of health and the impact of community-based interventions in improving the health of vulnerable and under-served populations.


PCCI updates COVID-19 Vulnerability Index, now accounting for mobility, recent cases

To better help inform the public and public health leaders in Dallas County, Parkland Center for Clinical innovation (PCCI) has updated its COVID-19 Vulnerability Index, adding dynamic factors such as recent mobility and positive COVID-19 test results.


Launched in June, PCCI’s Vulnerability Index determines communities at risk by examining comorbidity rates, including chronic illnesses such as hypertension, cancer, diabetes and heart disease; areas with high density of populations over the age of 65; and increased social deprivation such as lack of access to food, medicine, employment and transportation.

To further give a clearer view of the risk of COVID-19 holds for ZIP codes within Dallas County, the map now calibrates the infection risk on a 1-100 scale. With this scale and added dynamic factors, the map shows clearly where each ZIP code falls in terms of risk for infection for COVID-19. For example, the ZIP code with the highest Vulnerability Index score of 100, is 75211 around Cockrell Hill. One of the lowest ZIP codes, 75247, along I-35 just south of Love Field, has a Vulnerability Index score of .10.

“We continue to evolve the Vulnerability Index to help residents of Dallas County have the best understanding of how COVID-19 affects their neighborhoods,” Thomas Roderick, PhD, Senior Data and Applied Scientist at PCCI. “The Vulnerability Index map is an important element to help public health officials determine where to allocate testing and intervention resources, as well as underscore how important personal behavior, such as wearing mask and social distancing, are for individuals in high-risk areas.”

The PCCI COVID-19 Vulnerability Index can be found on its COVID-19 Hub for Dallas County at:

Data Sources:

To build Vulnerability Index, PCCI relied on data from Parkland Health & Hospital System, Dallas County Health and Human Services Department, the Dallas-Fort Worth Hospital Council, U.S. Census and SafeGraph.

HIMSS Interview with co-authors of new PCCI book, ‘Building Connected Communities of Care’

Hosted by HIMSS, co-authors Steve Miff and Keith Kosel sat down to discuss their new book, “Building Connected Communities of Care.” In this virtual presentation, originally set for HIMSS 2020, Steve and Keith take listeners through the factors in today’s healthcare ecosystems that are driving the need for connected communities.

Please click on the file below to hear the HIMSS interview program.

PCCI’s Vulnerability Index: Taking the fight to COVID-19

Read our blog about how PCCI’s team of clinical and data science experts developed the PCCI Vulnerability Index as a way to assist community and healthcare leaders to address the factors that cause COVID-19’s exponential spread. Click on the image below to view the blog:


Book Excerpt: ‘Building Connected Communities of Care’ Six Tracks Needed for a Connected Community

In a new book, “Building Connected Communities of Care,” published by PCCI, the authors Keith Kosel, PhD, and Steve Miff, PhD, created a playbook that offers a step-by-step program for coordinating medical and community-based resources to change how, where and when healthcare is delivered.

The book is a practical, how-to guide for health systems, payers, communities, philanthropic agencies, foundations, and policymakers desiring to streamline coordination and assistance efforts between medical and social services to reduce costs and improve the health, safety, and well-being of a community’s most vulnerable residents, especially those with chronic diseases and complex social needs.

The book is informed by PCCI’s experience building one of the first Connected Communities of Care in the nation, which was not an overnight proposition. The PCCI experience in Dallas highlighted that dedicated time, clear process, allocated resources and extreme collaboration were all needed  to align diverse and essential stakeholder groups. The playbook organizes activities needed  to build a Connected Community of Care into six specific tracks – all of which must be addressed, and all of which are vital for lasting success.  Following is an excerpt from the book:

Chapter 1, page 3:

The Six Tracks for CCC Implementation  

The figure (left) illustrates the six tracks of activity needed to create a Connected Community of Care (CCC). Development activities begin in the center to develop governance models, procedures and legal policies that reflect the values of the key stakeholders and the specific goals of the CCC.   Each additional Track has defined activities – activities within a Track build as a program moves from planning, to initial launch, to on-going sustainability;


PCCI’s Connected Communities of Care Model 


While PCCI recommends that CCC leaders pursue the Tracks in the following general order, much of the work in various Tracks can (and should) be performed in parallel with work in other Tracks.

PCCI recommends that the CCC leaders assign a designated Track lead for each specific Track. The work in the six CCC Tracks will also require input from additional stakeholders specific to the implementation of that particular Track. The six Playbook Tracks are:

  1. Governance Track. A CCC’s governance structure relies on a collective decision-making model rather than on leadership by a specific individual or organization. This Playbook assumes that a few key community organizations have already formed an initial steering group to make the significant decision to undertake the CCC initiative. It is critical, at least initially, for an empowered, established group of decision-makers to provide leadership through a “readiness assessment” process and during the initial CCC design stages. The “readiness assessment” comprises a set of activities designed to collectively uncover a community’s clinical and social needs and level of preparedness and commitment to hosting a CCC.
  2. Legal/Policy Track. Communities should identify considerations related to contracts, policies, and procedures to provide an overall CCC legal and policy framework for Governance and as part of the development of each Track. The construct of a legal framework requires a review of applicable federal, state, and local law, along with requirements imposed by Funders, Sponsors, and clinical and community Partners. As these requirements and considerations are tightly integrated with the business requirements, PCCI has incorporated some of the Legal/Policy considerations within each respective Track. The CCC’s Legal /Policy Track lead and CCC legal counsel should review all relevant key documents in all Tracks to ensure compliance. To streamline CCC preparation and implementation, PCCI recommends that CCC legal counsel leverage Participants’ existing legal structures, policies, processes, and agreements, where possible.
  3. Technology Platform Track. The Governance Track provides a framework for strategic assessment of CCC technology needs, ranging from required features to market analysis. The Technology Platform Track builds off that strategy and explores in depth the nuances and critical activities necessary to ensure successful deployment of the CCC’s backbone – the data-sharing platform. The technology infrastructure creates an integrated electronic platform to exchange clinical and social information securely between health organizations (i.e., hospitals, clinics) and CBOs (e.g., homeless shelters, food pantries) that are part of the CCC network. Construction of the platform should facilitate future data and external solution integration and provide an information exchange platform on which to customize additional case-management functionalities to meet the CCC users’ service-coordination requirements.
  4. Clinical Providers Track. Although clinical CCC workflows vary across selected clinical sites, the workflows need to converge on the CCC’s common goals. The Clinical Provider Track lead should contemplate the key factors and related nuances in establishing the clinical CCC consortium, including but not limited to the following: executive sponsorship; clear definition of roles and responsibilities; handling of clinical information; the compliance framework; and integration of the new workflows resulting from this work.
  5. Community Partners Track. Community workflows also require consideration of a unique set of circumstances, relationships, and nuances. Even more so than the clinical-provider workflows, community workflows vary widely across CBOs, but ultimately must align to support the global CCC goals. Leadership, staffing, and management models may vary from those of the clinical Partners, thus requiring dedicated, deep expertise from the Community Partners Track lead working to engage CBO Partners.
  6. Program Sustainability Track. Stakeholder and Participant support and revenue generation are two of the most important factors contributing to CCC sustainability. The CCC can garner that support through defining and demonstrating its value in providing better services and outcomes and in creating a vehicle for research and innovation benefiting the entire community. Significant funding may be required to design, build, implement, and sustain your local CCC. Unlike hospital quality-improvement programs that are expected to be deployed and to generate results within annual budgets, CCC deployments require several years to reach scale and maturity in order to produce meaningful Return on Investment (ROI) and Social Return on Investment (SROI) results.

For more information about “Building Connected Communities of Care,” or to get your copy today , go to HIMSS Publishing or

About the authors

Dr. Keith Kosel is a Vice President, Enterprise Relations at PCCI.

Dr. Steve Miff is the President and CEO of PCCI.


PCCI’s Expert Leadership Leveraging Social Determinants of Health

PCCI is a leader in researching, understanding and implementing social determinants of health (SDOH) programs that have taken a theory into practice, helping save lives, improving care and reducing costs. PCCI leverages data science and social determinants of health to better support under-served populations across our communities, and shows how this is possible with a number of successful programs. 

These are outlined in a groundbreaking series of blog posts showing how SDOH can go from a simple buzzword to a way to help change the way healthcare is delivered.

Please click on the links below to see PCCI’s expert perspectives on SDOH:


By Vikas Chowdhry, MS, MBA, is PCCI’s Chief Analytics and Information Officer


By Manjula Julka, MD, PCCI’s VP, Clinical Innovation


By Leslie Wainwright, PhD, PCCI’s Chief Funding and Innovation Officer


By Steve Miff, PhD, CEO of PCCI


By Steve Miff, PhD, CEO of PCCI