EHR-Based Surveillance of Diabetes Care Disruption During a Public Health Emergency: Neighborhood Patterns Across Chicago Census Tracts
This study utilized EHR data from Chicago health systems to demonstrate that the COVID-19 pandemic caused significant neighborhood-level disruptions in diabetes monitoring and preventive care, revealing that EHR-derived quality indicators can effectively serve as surveillance signals for identifying gaps in documented care during public health emergencies.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Chronic diseases like diabetes are often called "silent" conditions because they can damage the body for years before a person feels sick or sees a doctor. The key to preventing this damage is regular monitoring: checking blood sugar levels, measuring blood pressure, and taking preventive medications. In a stable world, these checks happen during routine visits. But when a public health emergency strikes, like a pandemic, the normal rhythm of care can break. People miss appointments, clinics change their hours, and the systems that track whether patients are getting the right care can go dark. This creates a dangerous blind spot. If a doctor cannot see a patient's latest test results, they cannot adjust treatment, even if the patient's condition is worsening. The question for public health officials is how to spot these gaps in care across a whole city without waiting for a crisis to become a tragedy.
A team of researchers set out to solve this problem by looking at how diabetes care changed in Chicago during the early years of the COVID-19 pandemic. They treated the city not as a single block of data, but as a mosaic of 782 distinct neighborhoods, known as census tracts. Using electronic health records from seven major health systems, they tracked five specific indicators of care for adults with diabetes between 2017 and 2022. They looked for whether patients had recent blood sugar tests, whether their blood sugar was under control, whether their blood pressure was measured, whether their blood pressure was controlled, and whether they were prescribed statins, a type of medication that protects the heart. The researchers compared the years before the pandemic with the years during it to see where the care had stopped being visible in the medical records.
The results revealed a sharp and widespread loss of "clinical visibility," a term the authors use to describe the presence of documented test results in a patient's file. Before the pandemic, about 43 percent of eligible patients did not have a documented blood sugar test in a given year. During the pandemic, that number jumped to nearly 61 percent. This meant that for hundreds of thousands of patient-years, doctors simply did not have the data needed to know if a patient's blood sugar was high or low. Similarly, the number of patients without a documented blood pressure measurement remained stubbornly high, dropping only slightly from roughly 79 percent to 75 percent. The researchers noted that this lack of data is not just a missing number; it represents a patient who might need medication adjustments but is invisible to the system.
While the loss of data was the most striking finding, the numbers for actual health outcomes told a more complicated story. The percentage of patients with documented, poorly controlled blood sugar appeared to improve, dropping from 12 percent to 9 percent. However, the researchers warned that this improvement was likely an illusion created by the missing data. Because so many patients were not being tested, the pool of people with recorded results became smaller and potentially different from the whole group. In the same way, the rate of controlled blood pressure appeared to rise, but the researchers argued this could not be trusted without knowing who was actually getting measured. The one clear negative trend was in preventive care: the prescription of statins for high-risk patients declined from about 66 percent to 60 percent, suggesting that even when patients were seen, preventive measures were being deprioritized.
The study went beyond citywide averages to map these changes onto the actual geography of Chicago. The researchers found that the disruptions were not spread evenly; they clustered in specific neighborhoods. They identified areas where multiple negative trends overlapped, creating "priority zones" where the risk of unmonitored disease was highest. Three neighborhoods—Brighton Park, East Garfield Park, and West Englewood—stood out as having the most severe gaps across multiple indicators. Other areas, including parts of the South and West sides, also showed significant clusters of missed care. These maps provided a visual signal of where the health system's safety net had frayed the most.
The authors emphasized that their findings were descriptive, showing what happened rather than proving exactly why it happened. They could not say for certain if the gaps were caused by patients staying home, doctors changing their documentation habits, or the sheer chaos of the emergency. What they did show was that electronic health records could serve as an early warning system. By monitoring these simple indicators of "visibility," public health agencies and hospitals could identify neighborhoods where routine care had stalled. This approach allows for targeted outreach, such as sending teams to specific blocks to catch up on missed tests or review medications, rather than guessing where help is needed. The study suggests that in times of crisis, the most important thing a health system can do is keep the lights on for the data that tells them who is getting sick and who is being left behind.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.