Why We Redesigned America’s Only National Database of School Carbon Monoxide Incidents – And What’s Changing

When we first began tracking carbon monoxide (CO) incidents in educational settings, our goal was simple: understand the problem so we could be informed enough to talk to people about it.

We didn’t know it then, but we were setting out to create the most comprehensive, transparent record of CO exposures in daycares, schools, and college campuses across the United States. But as the work of CO Safe Schools grew, and as the number of preventable CO incidents continued to surface, we realized that our database needed to evolve. 

Tracking “what happened” was no longer enough. To drive real policy change, improve emergency response, and help journalists and parents understand the risks in their communities, we needed better definitions, cleaner categories, and clearer markers of impact. We also needed to include school-adjacent locations like stadiums, day or overnight specialty camps, church preschools, and recreational centers used for afterschool learning or play.

Our goal now? Keep occupants safe, wherever they gather to learn.

A National Snapshot of the Crisis: What the Data Shows So Far

As of today, CO Safe Schools has identified and verified 426 carbon monoxide incidents in U.S. educational settings, including daycares, K–12 schools, vocational centers, camps, and college campuses. These are not theoretical risks; they are real events affecting real people in communities across the country.

Across these incidents, the patterns are unmistakable:

• CO incidents occur in every region of the United States, across all building ages and school types.

• A large portion of events happen during normal school hours, exposing students and staff at their most vulnerable.

• Detection failures remain a persistent theme — many buildings had no CO detection at all, the wrong type of detection, or alarms that never activated.

• Thousands of children and educators have been medically treated, many on-site and many taken to hospitals; there have been 15 total fatalities, 13 of them students.

• CO sources are varied, and many are missed by building codes, ranging from malfunctioning boilers and furnaces to idling buses, contractor equipment, ventilation failures, and portable engines.

Even more striking, most of these incidents would never have been publicly known without media reports. There is no national database for tracking school CO exposures (other than ours), no requirement to report CO incidents, and no federal requirement for CO detection in schools.

This lack of transparency is exactly why our database needed to evolve, because without clear, standardized, accurate reporting, we cannot prevent the next incident.

What are the updates to the CO Safe Schools Incident Database?

Below is a walkthrough of how we redesigned our fields and why each one matters for public health, safety, and accountability.

1. REGION: Understanding Nationwide Patterns

We continue to use the U.S. Census Bureau regions (Northeast, South, Midwest, West). Why? Because this allows us to identify:

• geographic hotspots

• states that outperform (or underperform) in detection and prevention

• regional gaps in code adoption or enforcement

Regional tracking tells a story that raw state numbers alone cannot.

2. STATE: Where Incidents Are Occurring

State-level reporting might seem obvious, but it's essential. Each state has:

• different building codes

• different CO detection requirements

• different enforcement practices

• different school construction timelines

• different climates that impact equipment use

Because CO requirements are set state-by-state (there is no federal CO requirement for schools), this field helps us understand where policy gaps directly correlate with real harm.

3. DATE: Why Timing Matters

Logging the date of the incident allows us to track:

• Academic year vs. calendar year trends

• winter heating-season spikes

• equipment startup failures at the first cold snap

• portable equipment use during warm months

• construction-season increases

• patterns around holidays or breaks

While CO incidents do tend to occur in the fall/winter months during cold weather, our data shows that CO is a year-round issue and should be treated as such.

4. CITY & SCHOOL NAME: Localizing the Impact

When CO exposures happen, they affect communities: real students, real teachers, and real families.

Recording the city and school name allows:

• verification of incidents

• cross-checking with official reports

• journalists to follow up

• districts to understand systemic patterns

• comparison across similar facility types

Transparency matters. Naming matters.

5. CAMPUS CATEGORY: A New, Cleaner Classification

One of the most important updates we made was simplifying “what kind of campus is this?”

Instead of focusing on ownership (public, private, charter), our categories now reflect who the building is serving and the activities that occur there, not who owns it.

This allows us to more accurately compare:

• Daycare & Early Childhood

• K–12 Schools

• Higher Education (Colleges, Universities, and Professional/Trade Training Sites)

• Campus Transportation or Private On-Campus Vehicles

• Camps or Recreational Spaces for Learning or Active Play

This shift reduces noise in the data and centers the actual population affected.

6. CAMPUS SUBTYPE: What Is the Building Used For?

Campus “Subtype” digs deeper into how the property is being used, not its governance.

Examples include:

• Daycare/Preschool

• Elementary

• Middle

• High

• Professional/Trades

• Transportation (bus line, bus garage, vehicle)

• Outdoor/External Use (like stadiums)

• Camps (Day or Overnight, for recreational, worship, or other purposes)

Higher education will also have its own subtypes:

• Residential (off-site or on-site dorms, housing, or apartments where occupants sleep)

• Educational/Administrative buildings (where occupants do work, learn, use labs)

• Mixed Use (a blend of retail, dining, and general use)

• Outdoor Use (parking lots, stadium, etc., often used for outdoor activities)

This matters because risk profiles vary dramatically between a boiler room below an elementary school and a student housing unit with gas stoves. Subtype helps us track those nuances.

7. CO DETECTION ON SITE: Rebuilt for Accuracy

This is one of the most important fields we track.

Here’s what we can track currently:

• Yes, alerted (provided a signal to occupants or emergency services)

• Yes, but did not alert (should have, but didn’t)

• No (no detection on site)

• Personal alarm activated (someone’s private alarm alerted to danger)

• Not reported or blank – no information provided

When possible, we include additional information provided.

How we’d love to categorize? The way is listed below. Unfortunately, the press (and those releasing statements) aren’t reporting this way:

• Hardwired CO system

• Portable meter (staff/contractor)

• Plug-in or battery-operated alarm

• Low-level CO monitors

• IAQ monitoring (if it includes CO)

• No detection installed

• Unknown

Why?  Because the vast majority of school CO incidents, particularly those with injury, involve absent, insufficient, or malfunctioning detection. Understanding what existed (or didn’t) is critical for policy change.

8. PPM LEVEL REPORTING: A Crucial Missing Piece in Most Incidents

Very few school CO incidents include reported ppm levels, but when they do, this field allows us to capture:

• severity of and level of accumulation

• whether ppm was recorded pre- or post-evacuation

This field helps contextualize the true danger level of each incident.

9. INJURY FIELDS:  The Human Element and Most Important Data We Track

Previously, we used a single “injury notes” field. Now we’ve broken this out into:

• # of Students treated on-site

• # of Adults treated on-site

• # of Undisclosed (age not reported) treated on site

• General # of Occupants seeking care on their own/after the fact (if reported)

• # of Students taken to hospital (ER or admitted)

• # of Adults taken to hospital (ER or admitted)

• # Undisclosed hospitalized (ER or admitted)

• Student fatalities

• Adult fatalities

This clarity matters because:

• harm to students vs. staff can indicate different exposure patterns

• injured counts matter for resource allocation

• hospitals often treat children more aggressively than adults

• undisclosed categories help track incomplete reporting and still account for injuries

This restructuring gives a clearer picture of populations being affected.

10. SOURCE CATEGORY: Standardizing the “What Happened”

We completely overhauled this field to establish five standardized, easy-to-understand CO source categories that reflect what created the CO in the first place (and how it impacted occupants)

1. Mechanical – Stationary: Boilers, furnaces, HVAC units, water heaters, generators

2. Mechanical – Portable: Construction tools, landscape equipment, maintenance equipment

3. Ventilation Issue: Improper ventilation, disconnected flues, blocked vents; sometimes this can be in combination with one of the other categories

4. Vehicular: Buses, cars, forklifts, delivery trucks, idling engines near intakes

5. Environmental/Fire: Fire-related CO inside or near the building (nature, structure, rubbish, or equipment)

These categories now allow national comparison of root causes.

11. SOURCE NOTES: Recording Human Details Behind Each Event

Source Notes allow us to capture context:

• Was a vent intentionally propped open?

• Was a contractor using a gas-powered saw indoors?

• Did a bus idle outside an air intake?

• Was the boiler overdue for service?

These details help identify systemic patterns and repeated preventable mistakes.

Why Tracking School Carbon Monoxide Incidents Matters

This database is more than numbers; it’s a national record of preventable harm. By refining our fields, we are:

• improving national CO awareness

• supporting media accuracy

• helping districts identify risks

• equipping policymakers with evidence

• influencing future codes and standards

• building a consistent, nationwide language around CO incidents

These updates bring us one step closer to making CO safety mainstream, and ensuring every child and educator is protected.