The noninvasive method is capnography — specifically, waveform capnography using a nasal cannula or a nasal-oral (oral-nasal) sampling cannula connected to a side-stream or main-stream CO2 sensor. The carbon dioxide waveform is displayed continuously, so the rescuer sees a value in millimeters of mercury (mmHg) and a shape, not just a number. This works only while the patient has enough circulation to carry CO2 from the lungs to the sensor, which is the boundary that matters most during a resuscitation.

Capnography measures the partial pressure of carbon dioxide at the end of exhalation, called end-tidal CO2 or EtCO2. In a resuscitation, you'll hear it called "waveform capnography," "sidestream capnography," or "end-tidal CO2 monitoring." A colorimetric detector is a related but different tool: it changes color to confirm tube placement, but it does not give a continuous numeric reading or a waveform. That distinction trips up a lot of people, and it's worth holding onto as you read.

What EtCO2 Actually Measures and Why It Matters in a Resuscitation

EtCO2 is the concentration of carbon dioxide in the gas you exhale at the very end of a breath. Because CO2 travels from tissue to blood to lungs, the number reflects three things at once: how well the body is making CO2, how well blood is moving it to the lungs, and how well the lungs are venting it. During a cardiac arrest, the first two collapse. That's exactly why the number becomes a live readout of circulation.

A normal EtCO2 in a healthy, spontaneously breathing adult sits around 35 to 45 mmHg. During CPR with a well-placed advanced airway, a reading above roughly 10 mmHg is the commonly cited threshold for adequate compressions, and values persistently below that are associated with poor outcomes. The American Heart Association's 2020 guidelines on cardiopulmonary resuscitation and emergency cardiovascular care treat waveform capnography as the preferred method for confirming and monitoring advanced airway placement, and it's also used to gauge CPR quality and detect return of spontaneous circulation.

Here's the part the standard explainers skip: the number is only as good as the gas sampling. If the sensor sits in a circuit with a lot of dead space, or the sample line fills with condensation, or the cannula shifts off the nose, the trace flattens or reads falsely low. During a resuscitation, those artifacts look exactly like a dying patient. You learn to check the cannula and the line before you trust a falling number.

The output is a waveform, and the shape carries information the number doesn't. A square, plateau-topped wave with a sharp upstroke suggests a patent airway and reasonable perfusion. A shark-fin or sloping upstroke suggests obstruction or a leak. A sudden drop to near zero suggests the tube came out or the circuit disconnected. Reading the shape is a different skill from reading the number, and both matter.

The Oral-Nasal Sampling Cannula: The Default Tool for Spontaneous-Breathing Resuscitation

For a patient who's still breathing — a respiratory arrest, an overdose, a seizure, a shock state — the standard noninvasive setup is a nasal-oral sampling cannula that sits under the nose and in front of the mouth, connected by tubing to a sidestream capnometer. The cannula splits the sample between the nostrils and the mouth so you get a consistent trace whether the patient breathes through the nose or the mouth. That's the practical reason the oral-nasal version exists rather than a plain nasal cannula.

Sidestream devices pull a small sample of gas — typically in the range of 50 milliliters per minute, though it varies by manufacturer — through a thin tube to a sensor in the monitor. The advantage during a resuscitation is that the sampling line is light and the patient can wear it while you work. The disadvantage is that the long, thin line is prone to clogging with secretions or condensation, and it can kink. When the trace goes flat, check the line before you conclude the patient has no circulation.

Mainstream capnography places the sensor directly in the breathing circuit or airway adapter instead of pulling gas away. It responds faster and doesn't suffer from sample-line clogging, but the sensor is heavier and adds dead space, which makes it less attractive for a spontaneously breathing patient wearing a cannula. For the cannula-based scenario in this question, sidestream is the usual choice.

One real limitation: an oral-nasal cannula doesn't seal the mouth. If the patient is gasping or has an obstructed airway, the trace can be erratic, and the number can lag behind the actual change in the patient's condition. Treat a flat trace on a spontaneously breathing patient as a cannula problem first, a perfusion problem second, and a monitor problem third — in that order.

Sidestream Versus Mainstream Capnography: A Side-by-Side Comparison

Both are noninvasive. The choice hinges on what you're monitoring and how mobile the patient is.

Feature Sidestream capnography Mainstream capnography
Sensor location In the monitor; gas pulled through a sampling line In the breathing circuit or airway adapter
Typical use during resuscitation Oral-nasal cannula for spontaneously breathing patients Intubated or advanced-airway patients in the circuit
Sample flow Roughly 50 mL/min, varies by manufacturer (confirm on the maker's site) None — no gas is diverted
Response speed Slightly slower due to transport time through the line Faster, sensor reads gas in place
Common failure mode Clogged or kinked sampling line, water trap saturation Sensor fouling, added dead space, heavier adapter
Added dead space Minimal at the cannula Higher at the adapter
Best fit Mobile, spontaneously breathing patient worn with a cannula Ventilated patient with a secured advanced airway

The verdict splits cleanly. If the patient is breathing on their own and you need a portable trace, sidestream with an oral-nasal cannula wins because it's light and wearable. If the patient is intubated and on a ventilator, mainstream wins on response time and line reliability. Neither is a substitute for a colorimetric detector when the sole question is tube placement.

Capnography Versus Colorimetric Detectors: They Are Not the Same Test

A colorimetric end-tidal CO2 detector is a small plastic device that clips between the bag-valve mask or tube and the patient and changes color — purple to yellow — when CO2 is present. It's cheap, disposable, and useful for a single question: is the tube in the airway? It does not produce a continuous number, and it does not show a waveform. Calling it "capnography" is a common misuse of the term.

The color change also fades. Once the detector is exposed to CO2 for a while, the color shift reverses or degrades, so it can't track trends across a long resuscitation. That's the structural reason capnography replaced it for monitoring: you can't see whether the EtCO2 is climbing back up after a compression cycle with a device that only tells you "yellow."

Cost matters here too. A colorimetric detector runs a few dollars per unit and is discarded after one use. Capnography requires a monitor and either a reusable sensor or a disposable sampling line, so the capital cost is higher, but you get continuous data across the whole event. The rough cost-per-use calculation is: monitor cost divided by device lifespan, plus the per-patient cannula or sampling line cost. The exact figures depend on the device, so confirm pricing with the manufacturer or your supply chain.

Where the colorimetric device still earns its place: in a cramped space with no monitor, or as a backup confirmation that the tube is in the trachea before you connect the main capnometer. It's a placement check, not a monitoring tool. Say that out loud in your training and the confusion disappears.

Reading the Waveform When the Number Stalls

The waveform is where experienced clinicians get information the figure alone can't give. Four basic shapes cover most resuscitation scenarios:

Normal, plateau-topped wave: patent airway, adequate perfusion, good compressions. Shark-fin or sloping upstroke: obstruction, bronchospasm, or a leak around the airway. Sudden drop to near zero with a flat line: tube dislodged, circuit disconnected, or cardiac arrest with no output. Gradual rise over minutes: improving perfusion, warming of the patient, or return of spontaneous circulation.

The stall case is the one to drill. Picture a patient in respiratory arrest who's being bagged through a mask. The EtCO2 reads 30 mmHg and holds. Then the number freezes on 30 while the waveform looks normal. Is the patient getting worse? Not necessarily — sidestream monitors often hold the last value when sampling fails, and a sampling-line clog produces exactly this frozen display. Check the line, the water trap, and the cannula position before you act on a number that won't move.

A rising EtCO2 is generally good news, but not always. During CPR, an abrupt jump of 10 mmHg or more often signals return of spontaneous circulation, because circulation has resumed and CO2 is being delivered to the lungs again. A malpositioned tube can also cause a false elevation on one side if the sensor reads CO2 that's built up from a different source. Read the trend, not the single reading.

The sensor temperature and the ambient environment matter more than most people realize. Sidestream monitors compensate for water vapor and barometric pressure, but a patient in a cold environment or a sensor in a hot ambulance can drift. Recalibrate or zero the device per the manufacturer's instructions before the event if you can. During the event, trust the waveform shape before you trust an isolated number.

What the Evidence Says About EtCO2 as a CPR Quality Guide

The physiology is the evidence. During CPR, cardiac output falls to a fraction of normal, so less CO2 reaches the lungs, and EtCO2 drops. Better compressions push more blood, more CO2 reaches the lungs, and EtCO2 rises. That relationship is why the AHA guidelines list capnography as a tool for monitoring CPR quality and detecting return of spontaneous circulation, and it's the basis for the 10 mmHg threshold cited earlier.

The limits of that evidence are worth stating plainly. EtCO2 is a surrogate for perfusion, not a direct measure of it, so a low number doesn't prove compressions are shallow and a high number doesn't prove they're good. Factors unrelated to CPR quality — the patient's metabolic rate, body temperature, the presence of a severe acidosis, and the sampling setup — all shift the number. A patient with a very low metabolic rate or severe hypothermia can have a misleadingly low EtCO2 even with adequate compressions.

There's a real edge case with obesity and chest-wall compliance. A patient with a stiff chest wall may need higher compression force to generate the same tidal volume and CO2 clearance, and the EtCO2 may read lower than a leaner patient with identical CPR quality. That's a reason to use the trend within a single patient over time, not to compare absolute numbers between patients.

What this means for practice: use EtCO2 to watch for change, not to grade a single snapshot. A rising trend over several minutes is meaningful; a single reading of 8 mmHg in a cold, acidotic patient is a data point, not a verdict.

Capnography During Pediatric and Neonatal Resuscitation

Pediatric resuscitation follows the same principles, with a smaller cannula and a lower expected EtCO2. A healthy child has an EtCO2 of roughly 35 to 45 mmHg, similar to an adult, but the sampling cannula must fit a small face, and the tidal volumes are much smaller, which makes dead space and leak around the cannula more consequential. A poorly fitting nasal cannula on an infant can give a trace so diluted by room air that it reads falsely low.

The 2020 AHA pediatric guidelines follow the adult recommendations on waveform capnography for confirming advanced airway placement and for monitoring CPR quality. The operating principle is identical. The practical difference is that in small patients, the sampling line flow can exceed the patient's minute ventilation, pulling in room air and flattening the trace. That's a real failure mode, and it's one reason neonatology teams often rely on colorimetric confirmation during delivery-room resuscitation before switching to continuous monitoring.

Neonatal resuscitation adds a wrinkle: the first minutes of life involve a transition from placental to pulmonary circulation, and CO2 readings can be volatile. Treat the early trace as a confirmation of airway placement and a rough guide, not as a stable physiologic value, until the baby's breathing pattern settles.

Common Mistakes and Artifacts That Produce a False Reading

The single most common mistake is trusting a frozen or flat trace without checking the hardware. Sidestream monitors hold their last value when sampling fails, so a "stuck" number is a hardware story until proven otherwise. The second most common is mistaking a colorimetric detector's color change for a continuous EtCO2 measurement — they answer different questions.

Other frequent errors:

Cannula displacement: the oral-nasal cannula slips down the chin during chest compressions, and the trace degrades. Re-seat it. Water trap saturation: condensation from humidified gas fills the trap, and the trace goes flat. Empty or replace it. Room-air dilution: on a small patient, the sample flow pulls in ambient air, and EtCO2 reads low. Confirm the line is seated and the flow rate matches the patient size. Sensor drift: a device that wasn't zeroed per the manufacturer's instructions can read a few mmHg off. Zero it before the event. Misreading a rising number as ROSC: not every rise is return of circulation. Confirm with a pulse check and the clinical picture.

The failure-handling rule is simple: flat waveform, check the cannula, then the line, then the water trap, then the monitor. Only after those four checks do you treat the trace as a physiologic signal. The quiet version of this rule: in a real resuscitation, the monitor is the fifth person in the room, not the first.

How to Set Up and Use Capnography During an Active Resuscitation

Setup takes under a minute if the equipment is staged, which is why the step sequence matters. Run it in this order:

1: Power on the monitor and confirm the CO2 module is active and zeroed per the manufacturer's instructions. 2: Attach the oral-nasal sampling cannula to the sampling line and seat it under the nose and in front of the mouth. 3: Connect the sampling line to the monitor's CO2 port, checking that the water trap is seated and dry. 4: Confirm a waveform appears with the patient's first breaths; verify the trace shows a plateau, not a flat line. 5: Note the EtCO2 value and the waveform shape on the record at the start of the event. 6: During compressions, watch the trend across cycles rather than reacting to single readings; re-check the cannula if the trace degrades. 7: On any sudden change, run the four-step hardware check — cannula, line, trap, monitor — before changing your resuscitation approach.

If the trace doesn't appear at step 4, the likely causes are a kinked sampling line, a saturated water trap, or a cannula that isn't seated. Replace the line or trap and re-seat the cannula before concluding the patient has no CO2 output. A trace that appears and then vanishes mid-event usually means the cannula moved during compressions or the line clogged with secretions.

FAQ

Is capnography invasive?

No. Capnography is noninvasive because the sensor reads exhaled gas at the airway — through a cannula at the nose and mouth or an adapter in the breathing circuit — without entering the body. The only "invasiveness" is the advanced airway itself, which is a separate procedure. The CO2 measurement adds nothing beyond a cannula on the face or a sensor in the circuit.

Can I use a plain nasal cannula instead of an oral-nasal one?

You can, but the trace will be less reliable in a patient who breathes through the mouth. A plain nasal cannula samples only nasal gas, so mouth breathing dilutes or misses the sample. An oral-nasal sampling cannula splits the sample between nose and mouth, which is why it's the standard choice for a spontaneously breathing resuscitation patient who isn't intubated.

Does a colorimetric detector count as capnography?

No. A colorimetric detector changes color in the presence of CO2 and answers one question — is the tube in the airway — but it produces no waveform and no continuous number. Capnography gives a real-time trace and a numeric EtCO2, which is what you need to monitor CPR quality and detect return of spontaneous circulation. The two devices are complementary, not interchangeable.

What EtCO2 value suggests return of spontaneous circulation during CPR?

An abrupt rise of roughly 10 mmHg or more from the baseline during compressions is the commonly cited signal, and it should be confirmed with a pulse check. A single elevated reading without a pulse is not return of circulation. The trend across several cycles carries more weight than any one number, and the waveform shape should be re-checked at the same time.

Why does my EtCO2 reading freeze or go flat?

A frozen or flat trace is usually a hardware problem, not a physiologic one. Sidestream monitors hold the last value when sampling fails, and the usual culprits are a kinked sampling line, a saturated water trap, or a displaced cannula. Check those four items — cannula, line, trap, monitor — before you change your resuscitation approach based on the reading.

Does capnography work the same in children and infants?

The principle is the same, but the practical setup differs. Smaller patients need a smaller cannula, and the sampling flow can exceed a small patient's minute ventilation, pulling in room air and flattening the trace. Pediatric guidelines still recommend waveform capnography for airway confirmation and CPR monitoring, but the trace should be read with more caution in the smallest patients.

Is capnography more expensive than a colorimetric detector?

Yes, in capital terms. A colorimetric detector is a few dollars per use and disposable, while capnography needs a monitor plus a per-patient sampling line or cannula. The rough per-use cost is the monitor cost divided by its lifespan plus the disposable cost. Exact figures vary by device — confirm pricing with the manufacturer or your supply chain before budgeting.

This article is general educational information about resuscitation monitoring and is not a diagnosis, treatment plan, or clinical protocol. Device specifications, guideline thresholds, and recommended practices vary by manufacturer and by jurisdiction, so confirm the current instructions for your equipment and follow your local resuscitation guidelines. For an actual emergency, call your local emergency services immediately and follow the direction of a qualified clinician on scene.