The Eaton wildfire burned from January 7 to 31, 2025, across 14,021 acres in the greater Pasadena, California, area — Altadena, Pasadena and Sierra Madre, home to about 230,000 people. It destroyed 9,419 structures and killed 19 residents. In the first 24 hours about 100,000 people were evacuated, among them an estimated 1,800 residents of 25 assisted living and 10 skilled nursing facilities.
The shelter
A shelter opened the day the fire began, in the Pasadena Convention Center — about 97,000 square feet, divided into halls for the general population, families, people with pets, and dining. About 300 evacuees arrived the first evening and 1,700 by the next; by the last week, February 10–16, about 200 remained.
| Date (2025) | What happened |
|---|---|
| January 11 | the Pasadena Public Health Department has isolation and infection-control protocols ready for norovirus, COVID-19 and influenza, based on local wastewater and lab data and on past shelter outbreaks — including norovirus after the 2018 Camp Fire |
| January 12 | management of the center, including infection control, passes to a nongovernmental organization experienced in running emergency shelters; the health department hands over its protocols and reduces its staff on site |
| January 13 | on-site clinicians report residents with COVID-19, influenza, and vomiting or diarrhea. The health department visits the same day and finds poor hand hygiene and protective equipment use, no separate isolation area, and cleaners that don't kill norovirus, and reinforces the use of EPA List G disinfectants. Separate isolation halls open |
| January 18 | with more illness reported, the city asks the California Department of Public Health for help |
| From then | the city and state health departments visit and audit the shelter daily |
| January 22 | cases fall sharply |
| By January 27 | more cleaning staff on site, the right cleaners in use throughout, and staff respirator use high |
| February 6 | the last reported case, COVID-19 |
| February 12 | all ill residents out of isolation |
| February 16 | the shelter closes |
What spread
Illness was found when residents sought care from on-site clinicians — no one without symptoms was screened — and from staff reports to their employers.
| Illness | Cases | Residents | Staff | Hospitalized | First and last detected |
|---|---|---|---|---|---|
| Norovirus (4 confirmed, 100 probable) | 104 | 66 | 38 | 9 (8.7%) | Jan 11 – Jan 31 |
| COVID-19 | 56 | 48 | 8 | 6 (10.7%) | Jan 11 – Feb 6 |
| Influenza | 29 | 23 | 6 | 0 | Jan 11 – Jan 30 |
| Other respiratory illness (negative or no test) | 30 | 1 | 29 | 0 | Jan 15 – Feb 4 |
Eleven residents had more than one: four norovirus and COVID-19, five norovirus and influenza, two COVID-19 and influenza. COVID-19 and influenza were confirmed with an on-site combined rapid test; after the first few stool tests confirmed norovirus, testing stopped and probable cases counted together with confirmed ones.
Figure 1. Cases by illness and date, residents and staff. CDC
Figure 2. When each infection-control measure began. CDC
What worked
The drop in cases came after isolation and infection-control measures were put in place and agencies coordinated — though, the authors stress, an observational study cannot prove which measures did it. They credit:
- clear roles for infection control, set early and shared among the health departments, the shelter organization and health care partners;
- protocols that carried over when management changed hands and staff turned over;
- surveillance of staff as well as residents, keeping staff identities confidential;
- regular audits and feedback on masks, hand hygiene and cleaning.
Shelters should build standardized infection-control protocols into how they open — before people fall ill — especially when several diseases are circulating and several agencies are involved.
Limits
- Only people who sought care were counted, so milder cases were missed.
- Rapid tests are not very sensitive, so some COVID-19 and flu may have been counted as "other respiratory illness."
- Staff surveillance started late, and some earlier cases may have been missed.
- Demographic data weren't consistently collected.
- People came and went, so daily counts of residents weren't possible.
- It is unclear whether the shrinking population itself reduced cases.
- Antiviral treatment data weren't available.
Sources
Based on Patrick R, Lee K, Kuan M, et al., "Norovirus, COVID-19, and Influenza Outbreaks Among Residents and Staff Members at the Eaton Wildfire Evacuation Shelter — Pasadena, California, January–February 2025," MMWR Morbidity and Mortality Weekly Report volume 75, number 26, Centers for Disease Control and Prevention; a work of the United States government in the public domain.
Licence: CC0 1.0 (public domain) · Adapted from www.cdc.gov
1
0
0
0
Comments
