Latest News

Turning Primary Care EHRs Into Public Health Assets

By Deborah Borfitz

May 14, 2019 | Population-level health concerns such as lead poisoning and sexually transmitted infections (STIs) are everyday issues for community health centers funded by the Health Resources & Services Administration (HRSA), where one in 12 people in the US receive care. This growing national network of primary care service delivery sites offers culturally competent care to vulnerable populations, including veterans and those living in poverty. These health centers are supported by investments in health information technology infrastructure enabling public health functions to be realized at the point of care, according to Fred Rachman, MD, CEO of AllianceChicago.

AllianceChicago is one of the oldest Health Center Controlled Networks supporting HRSA-funded health centers, making it unique in terms of size and mission, says Rachman. It is national in scope, servicing clinics in 23 states with an array of both critical and nice-to-have operational services.

A key objective is to turn their electronic health records (EHRs) into public health assets.

Mapping Community Resources

AllianceChicago partnered with the University of Chicago on a project funded by an innovation grant from the Centers for Medicare & Medicaid Services to create access to an electronic database of community health resources that would link to the EHRs of local safety net providers on the South Side of Chicago.

The database, known as HealtheRx, geographically mapped relevant community-based resources such as recreation centers and transportation and social services programs. The community resources were aggregated as a separate door-to-door effort by local students with funding from the nonprofit MAPSCorps. "Our role was surfacing that list of community assets to end users in the EHR system," says Nivedita Mohanty, MD, chief research officer of AllianceChicago.

HealtheRx uses ICD diagnostic codes, almost any of which has a connection to social determinants of health, says Mohanty. An example would be diabetics whose care plans would include managing their blood sugar and nutrition, getting routine eye screenings and exercising regularly. "One of the benefits of having a community assets list is that you can find resources in one's neighborhood, overcoming the obstacles that might be presented by social and economic influences."

For the AllianceChicago network participants in the project, the original HealtheRx model was ported to clinical care teams at South Side health centers, but has since expanded considerably, Mohanty says. The intellectual property was transferred to the university and then to the principal investigator, who commercialized it as the NowPow platform that is now used by a variety of health systems in multiple cities.

Predictive Analytics for Lead Hazard Prevention

Another important initiative undertaken by AllianceChicago seeks to mitigate the risk of lead exposure, which disproportionately affects vulnerable populations living in housing built before Chicago banned lead-based paint, says Mohanty. The consequences of exposure are significant, including learning disabilities, behavioral problems, growth deficiencies and attention deficit disorder, but remediation typically doesn't happen until after a child has been poisoned.

In the first and largest component of a project led by the Chicago Department of Public Health and funded by the Robert Wood Johnson Foundation, AllianceChicago collaborated with the Chicago Department of Innovation and Technology and the University of Chicago's Public Policy Department to share data for the purpose of social good, Mohanty says. Aggregating data from different health systems gave stakeholders an idea of where the lead is, allowing them to develop a predictive algorithm "to identify subsequent cases of lead poisoning before lead is detected in a child's blood." This in turn helped inform a better preventive strategy for lead exposure in the Chicagoland area.

In the next phase of the project, its HRSA-funded health centers participating in the pilot will benefit from the algorithm, says Mohanty. "We worked with the city to create an interface so when a patient comes into the clinic we can run the algorithm and it will produce a risk score for that patient and then a visual inspection can be proactively scheduled [with the public health department] at the child's or pregnant woman's home to identify if that patient has visible signs of lead risk and, if so, steps toward remediation can occur."

The algorithm will be deployed in the EHR and target pregnant women and in children under the age of one, she says. The EHR decision support has been tested on test patients' data but is "on its way to being able to provide insights for real patients" living in high-poverty neighborhoods as part of their routine prenatal care and well child encounters. The Chicago public health department has secured resources to expand capacity for visual lead inspections in anticipation of the ramped-up demand.

The collaborative initiative won the Milbank Memorial Fund and AcademyHealth State and Local Innovation Award in 2018 for building the predictive model and all the requisite data sharing that went into it, Mohanty says. The data sharing was largely facilitated by the focus on public health initiatives rather than for other projects sponsored by private interests or solo not-for-profit entities.

"One of the roles of public agencies is to do things that individuals can trust are in the public interest rather than coming out of a profit motivation," adds Rachman, especially on projects involving access to personal health data. The ROI on projects "comes out of having some sustainable public effort that will benefit everyone and also be self-sustaining."

The roadmap for the lead hazard prevention initiative will eventually extend beyond HRSA-funded clinics to include other types of ambulatory practices as well as emergency rooms and hospitals, Mohanty says. Meanwhile, AllianceChicago has developed a similar type of risk prediction tool for diabetes in conjunction with the University of Chicago's Public Policy Department—more specifically, its Data Science for Social Good program that recruits students in computer science fields to help solve public problems.

"We leveraged their brain power to help us identify how to predict diabetes risk because that's a huge focus for all of our health centers," she says. "If we're able to proactively identify these individuals even before [they] are eligible for screening recommendations, we might be able to impact their disease trajectory in developing diabetes. We're actively looking for ways to deploy that algorithm within our own network, either tied to EHRs or to population health management tools."

Case Detection Logic for STIs

In a pilot project funded by the Centers for Disease Control & Prevention (CDC), and supported by the Public Health Informatics Institute, AllianceChicago developed an algorithm for identifying cases of STIs and electronically reporting them to the state public health department, Mohanty says. The algorithm uses case detection logic—e.g., lab results and ICD codes suggesting a patient has gonorrhea or chlamydia. The current process is labor intensive, requiring a member of the care team to fill in paper documents, resulting in under-reporting.

As was demonstrated, electronic case reporting both detected more cases of the two STIs and improved the level of reporting, Mohanty says. Using a second grant from the CDC, AllianceChicago will now track what happens to patients in whom an STI is detected, such as returning to the clinic for treatment.

What If?

In a limited pilot with expansion potential, AllianceChicago recently partnered with the CDC and the Illinois Department of Public Health to create an iPad version of the Behavioral Risk Factor Surveillance System (BRFSS), the world's largest continuously conducted health survey system that collects state data about how residents regard their health-related risk behaviors, chronic health conditions, and use of preventive services. "We did it as a demonstration to show health departments and the CDC what a future way could be to administer the BRFSS," says Mohanty, who remains hopeful that someone will pick up on the idea.

The current survey process is decades old, Rachman says, relying on "random rigid dialing on small samples and takes two years to produce a population-level snapshot. So, we were trying to bring it into the modern age. The real power of the survey is the ability to do internet searches in real time using technology." One day, an aggregated database of survey results could be queried by providers to return information in the context of patients' community and health status—and actions that could be taken to improve it.

Clinical Decision Support Repository

Since 2016, a growing variety of downloadable clinical decision support (CDS) tools have been made available for piloting on the CDS Connect Repository supported by the Agency for Healthcare Research and Quality, says Mohanty. Pain management, and an immunization calculation engine, are among the current domains covered. While available to the public, technical support would be required to get them integrated with an EHR system.

AllianceChicago was involved in the pilot for a cardiovascular disease risk tool that predicts an individual's 10-year risk of having a cardiac event, providing guidance on who should be put on statin therapy for high cholesterol, she says. It is now live at one HRSA-supported health center. An alternative tool is already available in the EHR of the other health centers.