Two rooms on the same corridor, both signed off at commissioning, can behave completely differently once people are inside them. One clears the air it is given. The other holds it. From the doorway, the rooms look identical, and no document in the building distinguishes between them.
Airborne transmission risk is shaped by that difference more than by almost anything else a facilities team controls day to day. Dilution is the primary defense available in an occupied space, and dilution is precisely what carbon dioxide measures. The connection is not a pandemic-era invention. It rests on a body of work stretching back two decades, and understanding both what that work supports and what it does not is the difference between a defensible monitoring program and an overclaim that collapses at the first informed question.
Why Airborne Transmission Risk Comes Down to Air Exchange
The arithmetic is stark and worth stating plainly. Outdoor air holds a few hundred parts per million of CO2. Exhaled human breath holds roughly 40,000 ppm. Every occupant of a room is a continuous source, and ventilation is the only removal mechanism operating in most spaces.
That means CO2 concentration functions as a direct readout of how much of the air in a room has already passed through somebody's lungs. The higher the number, the larger the share of each breath that is secondhand. Nothing else routinely measured in a hospital gives you that figure.
Ventilation standards encode the same logic. Airborne infection isolation rooms require twelve air changes per hour against six for general patient rooms, and the difference exists because higher exchange rates reduce the concentration of anything suspended in the air, pathogens included alongside odors and particulates.
The gap between that design intent and daily performance is where the operational risk actually lives. A room specified at twelve and delivering nine is still a good room by most measures. It is not the room on the drawing, it is not the room the infection control policy assumes, and nothing except continuous measurement will tell you which one you currently have.
Rebreathed Air Fraction, Explained Simply
The concept that formalizes all of this is the rebreathed air fraction: the proportion of each inhaled breath that another occupant has already exhaled.
Rudnick and Milton established the framework in 2003, showing that the risk of indoor airborne infection transmission could be estimated directly from carbon dioxide concentration without needing to measure outdoor air supply rates, which are notoriously difficult to establish in a working building and vary continuously anyway. Because CO2 and exhaled aerosols are produced together by the same people and removed together by the same ventilation, the quantity you can easily measure stands in for the quantity you cannot.
In practice this means a facilities team can reason about shared air using a sensor rather than commissioning a ventilation survey. The widely applied public health benchmark of 800 ppm is a rough expression of the same underlying idea, marking the point at which the rebreathed share has grown large enough to warrant a look at the system.
None of this requires modeling infection probability, and hospitals should not attempt to. The useful output is comparative rather than absolute: which spaces in your building hold more shared air than others, at which times, and under what occupancy.
What CO2 Cannot Tell You
This is the section that keeps everything above it defensible.
Carbon dioxide describes dilution and nothing else. It says nothing about filtration efficiency, so a room with HEPA filtration and modest outdoor air supply can show elevated readings while performing well against airborne particles. It says nothing about whether anyone present is actually infectious. It is a ventilation signal, never a health verdict, and presenting it as the latter will cost you credibility with infection control colleagues who know the difference.
The measurement carries physical limits too. The framework assumes a reasonably well-mixed space, and researchers examining uncertainties arising from point measurements have shown that a single sensor in a poorly mixed room can misrepresent conditions several meters away from it. Placement matters enormously, and one reading is not a room.
Timing introduces another limit. A reading describes the room at that moment, and a space can sit comfortably below threshold for six hours and climb sharply during a single busy clinic. Any conclusion drawn from a snapshot inherits whatever conditions happened to exist when it was taken.
Treated as one input among several, alongside filtration, pressure relationships, and occupancy limits, CO2 earns its place in the program. Treated as a risk score, it will eventually embarrass whoever presented it that way.
A Ventilation Signal You Can Act On
The practical use is narrow and genuinely valuable. Continuous readings tell you which spaces hold the most shared air, under what conditions they do it, and whether a mechanical change improved matters or quietly made them worse somewhere else.
That comparative picture is the foundation any serious air quality monitoring for hospitals program is built on, and it costs a fraction of the ventilation survey it partly replaces. Alertify monitors conditions continuously through its indoor air quality sensors, with no cameras, no microphones, and a timestamped record retained for 180 days or more.
Book a demo and we will walk through placement for the spaces where shared air matters most in your building.
Frequently Asked Questions
What is the rebreathed air fraction?
It is the proportion of each breath that another occupant has already exhaled. Rudnick and Milton showed in 2003 that it can be estimated directly from CO₂ concentration, because CO₂ and exhaled aerosols are produced together by the same people and removed together by the same ventilation.
Can a CO₂ monitor detect viruses or infection risk?
No. CO₂ is a ventilation signal, never a health verdict. It shows how much shared air a space holds but says nothing about filtration efficiency or whether anyone present is infectious. Use it comparatively, alongside filtration, pressure relationships and occupancy limits, not as a risk score.
Why does sensor placement matter for CO₂ readings?
The underlying framework assumes a reasonably well-mixed room, and research on point measurements shows a single sensor in a poorly mixed space can misrepresent conditions several meters away. Place sensors deliberately in the occupied zone, and remember that one reading is not a room.