Health Alerts
Last updated: July 7, 2026
Epic Health Alerts use statistical models to identify counties or states experiencing elevated rates of health conditions based on Cosmos data. Alerts are reviewed by the Epic Research team before publication. For details on our detection process and data source, see the Methodology section below.
Rates with an asterisk are from partial weeks.
Number of Alerts by State (hover for details)
Alerts By Condition
Select a condition or multiple conditions below to display areas with alerts for that condition.
Acute Pharyngitis
An inflammation of the throat (pharynx) that causes sore throat, pain with swallowing, and fever. The cause can be an infective organism or other types of irritants. This alert reflects elevated diagnosis rates above expected seasonal levels.
| State | County | Estimated Onset | Cases per 100k (Latest Week) |
|---|---|---|---|
| Iowa | Pottawattamie | 5/4/2026 | 3191.5* |
| Wisconsin | Jefferson | 5/4/2026 | 630.4 |
Hand, Foot, and Mouth Disease
Hand, foot, and mouth disease is a common and contagious viral illness caused by enteroviruses, most often affecting young children. Symptoms include fever, mouth sores, and a skin rash on the hands and feet. This alert reflects elevated diagnosis rates above expected seasonal levels.
| State | County | Estimated Onset | Cases per 100k (Latest Week) |
|---|---|---|---|
| Nebraska | Douglas | 6/22/2026 | 154.9* |
| Nebraska | Sarpy | 6/22/2026 | 615.1* |
Heat Illness
Heat illness is a group of acute conditions that occur when the body's heat gain exceeds its ability to cool itself. It ranges from heat cramps, heat syncope (fainting), and heat exhaustion to heatstroke, a life-threatening medical emergency. Symptoms can include heavy sweating, muscle cramps, dizziness, nausea, headache, and, in severe cases, confusion or loss of consciousness. This alert reflects an elevated rate of visits for heat-related illness in the affected area, more visits than would typically be expected for this location and time of year.
| State | County | Estimated Onset | Cases per 100k (Latest Week) |
|---|---|---|---|
| Indiana | (State-Wide) | 6/29/2026 | 556.6* |
| Kentucky | (State-Wide) | 6/29/2026 | 779.6* |
| Michigan | (State-Wide) | 6/29/2026 | 420.1* |
| Ohio | (State-Wide) | 6/29/2026 | 567.4* |
| Wisconsin | (State-Wide) | 6/29/2026 | 254.1* |
Toxic Effect of Smoke
Toxic effect of smoke is an acute injury caused by unintentionally breathing smoke from fires or burning materials. Presentations can range from cough, throat or eye irritation, and shortness of breath to more serious breathing difficulty. This Health Alert reflects an elevated rate of visits for accidental smoke inhalation in the affected area, more visits than would typically be expected for this region and time of year. Specific source or cause are not known.
| State | County | Estimated Onset | Cases per 100k (Latest Week) |
|---|---|---|---|
| California | Los Angeles | 6/22/2026 | 40* |
Viral Gastroenteritis
Viral gastroenteritis is an inflammation of the stomach and intestines caused by a viral infection. Good hand hygiene, particularly thorough handwashing with soap and water, is the most effective measure to prevent the spread of viral gastroenteritis. Norovirus is the most common cause in adults, while rotavirus is more common in young children; a rotavirus vaccine is available for infants. Viral gastroenteritis spreads through the fecal-oral route, including contact with an infected person, contaminated food or water, or contaminated surfaces. Symptoms include watery diarrhea, nausea, vomiting, abdominal cramps, and low-grade fever, and typically resolve within one to three days.
| State | County | Estimated Onset | Cases per 100k (Latest Week) |
|---|---|---|---|
| California | Madera | 4/6/2026 | 781.9* |
| Michigan | Oakland | 6/22/2026 | 188.6* |
| Ohio | Lake | 6/22/2026 | 47.9 |
| Ohio | Lucas | 6/22/2026 | 69.2 |
Methodology
- Data Source
- Health Alerts are generated based on analysis in Cosmos. For more information, visit About Cosmos.
Health Alerts are based on data from Epic Cosmos participating organizations. They do not reflect complete case counts for a condition but rather evaluate population level trends of disease activity.
- How Conditions Are Identified
- Our detection process monitors diagnosis rates at the county or state level using ICD-10-CM diagnosis codes. A condition is flagged for review when it meets all the following criteria:
Year-over-year increase. The rate of the condition must be higher than it was during the same period in the prior year. This accounts for normal seasonal variation, such as a rise in flu diagnoses during winter, which would not trigger an alert on its own.
Accelerating growth. The share of the condition (the proportion of this diagnosis relative to all patients seen) must be growing at an increasing rate. Specifically, the growth from last month to this month must exceed the growth from two months ago to last month. This ensures the model is responding to conditions that are increasing rapidly, not simply trending gradually upward.
Statistical significance. Conditions that meet the first two criteria are then evaluated using a Farrington improved algorithm, a well-established method in public health surveillance. This model uses three years of historical data to establish expected baseline rates, further adjusting for seasonality and long-term trends, and confirms that the observed increase is statistically significant and not likely due to random variation.
- Clinical Review
- All statistically flagged conditions are reviewed by the Epic Research team, which includes clinicians and data scientists, before being published. This manual review step assesses whether the alert is clinically meaningful and appropriate for public reporting.
- Statewide vs County-level Alerts
- Under most circumstances, alerts will be issued at a county level, rather than at a state level. However, for conditions that have a small number of cases in affected counties, we may issue a statewide alert rather than a county level alert to help preserve patient privacy. We may also issue a statewide alert when a condition meets the alerting criteria when aggregated across the entire state but few or no individual counties meet the alerting criteria.
- Types of Conditions Detected
- The criteria for determining alerts are designed to surface conditions that are both increased from historical rates and accelerating. Alerts are most likely to appear for acute conditions, communicable diseases, and rare or unexpected spikes rather than for chronic conditions or predictable seasonal patterns.
- Update Frequency
- The Epic Research team typically reviews model output on weekdays. New alerts are published as clinically relevant conditions are identified; there is no fixed publication schedule. Alert activity on this page reflects the most recent completed review.
- Column Definitions
- Cases per 100k (Latest Week) — The number of patients with the condition per 100,000 total patients with an encounter in the county during the reporting week.
Estimated Onset — The date the elevated rate first occurred as detected by our model.