CO₂ as an airborne-pathogen proxy: what the ratio actually tells you

CO₂ in an occupied room comes overwhelmingly from human breath, and ventilation is what removes it. Airborne respiratory pathogens (the influenza, RSV, and SARS-CoV-2 families, plus measles and many others) follow the same physics: they are produced by occupants and they are removed by the same airflow. They are not identical, because pathogen production varies enormously across individuals and over the course of an infection, and because filtration removes pathogens at rates that depend on particle size, while filtration does not touch CO₂. But to a first approximation, the fraction of indoor air that has already been breathed by someone in the room tracks both, and CDC ventilation guidance and ASHRAE's pandemic-response work both adopted CO₂ as a practical ventilation-rate proxy on those grounds.

The arithmetic, roughly. Outdoor CO₂ is about 420 ppm in 2026 and rising slowly. A resting adult exhales air at roughly 40,000 ppm CO₂. The rebreathed-air fraction in a room is approximately (indoor CO₂ − outdoor CO₂) / (exhaled CO₂ − outdoor CO₂), which for typical numbers gives under 1% rebreathed at 800 ppm, around 2-3% at 1,000 ppm, and 5-6% at 1,500 ppm. Those percentages sound small until you realize that "5% of the air you are breathing has been breathed by someone else in the room recently" is a substantial dose-multiplier when one of those other people is infectious. The shape of the model is the same one CO₂ rebreathing detection uses for sleep and meeting-room analyses; the application is different.

Working thresholds. Under 800 ppm: ventilation is good enough that most respiratory particles dilute out before they accumulate; this is the target for healthcare waiting rooms, classrooms, and any space hosting immunocompromised people during respiratory illness season. 800-1,000 ppm: acceptable for most non-clinical occupied spaces; the WELL Building standard targets 900 ppm. 1,000-1,500 ppm: moderate rebreathing; in respiratory illness season, in spaces with high-risk activities (singing, group exercise, dental work), or with vulnerable people present, this is the band where additional measures (open windows, portable HEPAs, masking) earn their keep. Above 1,500 ppm: significant rebreathing; a typical poorly-ventilated classroom in winter sits here for hours. The thresholds are softer than they look because the relationship is monotonic with no clean cliff.

What the model does and does not tell you. It does not measure pathogen load: a room can be at 600 ppm with one highly infectious singer and still be a high-risk room, or at 1,800 ppm with nobody contagious and not be a risk at all. It does not account for filtration: a room with a Corsi-Rosenthal box or in-duct MERV-13 has a meaningful pathogen-removal rate that CO₂ does not see at all (CO₂ is not filtered, so two rooms at the same CO₂ can have very different airborne pathogen levels). It does not capture proximity: standing 1 meter from an infectious person is much higher-risk than standing 5 meters away, regardless of room average. The right use is as one signal among several, not a single risk score. See classrooms, offices, and demand-controlled ventilation for the deployment side, and reducing CO₂ for the playbook.

This is environmental information, not medical advice. The dashboard's readings help you make decisions about the air in your space. They do not diagnose conditions, interpret symptoms, or replace conversations with your physician. If symptoms persist, worsen, or coincide with a known exposure, talk to a healthcare professional. See the AI's medical-advice scope.