The Normal Waveform
Capnography measures CO2 partial pressure continuously through the respiratory cycle, producing a waveform (capnogram) with four distinct phases per breath. EtCO2 — the value most people quote — is just the number read at the end of the plateau, phase III.
| Phase | What's happening |
|---|---|
| 0 (baseline) | Inspiration — CO2 should read zero on a functioning circuit. |
| I (dead space) | Start of exhalation — CO2-free gas from the anatomic dead space (conducting airways) is expelled first. |
| II (alveolar upstroke) | CO2-rich alveolar gas begins mixing with the dead-space gas; the steep rise. |
| III (alveolar plateau) | Relatively uniform alveolar gas is exhaled; EtCO2 is read at the very end of this phase, the point closest to alveolar (and by extension, arterial) CO2. |
Normal values (adult): EtCO2 35–45 mmHg. PaCO2 typically exceeds EtCO2 by 2–5 mmHg (the a-ET gradient) due to dead-space ventilation — a gradient that widens meaningfully in conditions that increase dead space, like PE or severe COPD.
EtCO2 is a ventilation, perfusion, and metabolism monitor simultaneously — a sudden change can mean any of the three shifted, which is exactly why the waveform shape matters as much as the number.
Abnormal Patterns
| Pattern | Appearance | Cause |
|---|---|---|
| Obstructive ("shark fin") | Upsloping phase II, no clean plateau, "scooped" phase III | Bronchospasm — asthma, COPD, partial ETT obstruction/kink |
| Hypoventilation | Elevated EtCO2 (>45), normal shape | Sedation, opioids, rebreathing, hyperthermia or malignant hyperthermia, deliberately permissive settings |
| Hyperventilation | Decreased EtCO2 (<30), normal shape | Pain, anxiety, sepsis, pregnancy, compensatory response to metabolic acidosis; hypothermia lowers EtCO2 by reducing CO2 production |
| Curare cleft | Small notch/dip mid-phase II | Spontaneous respiratory effort breaking through incomplete paralysis — a real-time paralysis-depth signal |
| Cardiogenic oscillations | Regular small ripples on the plateau | Mechanical transmission of cardiac contraction to the airway — usually a reassuring sign of adequate cardiac output |
| Rebreathing | Baseline doesn't return to zero | Exhausted CO2 absorber, absorber not properly attached, low fresh gas flow, circuit leak |
| Sudden fall to near-zero | Abrupt loss of waveform | Esophageal intubation, complete disconnection or extubation, or — the one to never forget — cardiac arrest |
The dangerous confusion: a flat capnography line does not automatically mean esophageal intubation. If the tube was confirmed correctly placed moments earlier, an abrupt flatline is a circulatory arrest until proven otherwise — chest compressions restore some waveform even in a correctly placed tube, so a flat line during ongoing CPR (rather than at a pulse check) is a different problem than a flat line right after intubation.
Capnography in Cardiac Arrest
ETCO2 during CPR is primarily a marker of pulmonary blood flow, and therefore a real-time, breath-by-breath readout of compression quality and cardiac output — arguably its single highest-yield use in resuscitation.
| Finding | Evidence |
|---|---|
| Low threshold, poor prognosis | EtCO2 <10 mmHg after 20 minutes of CPR has been associated with as low as a 0.5% likelihood of ROSC across pooled observational data. |
| ILCOR 2020 consensus | Suggests against using any EtCO2 cutoff value alone to predict mortality or to decide to stop resuscitation (weak recommendation, low-certainty evidence). Values above about 20 mmHg are associated with better odds, but no single number is decisive. |
| Trend over threshold | Multiple reviews now emphasize that trending EtCO2 (rising, falling, abrupt jump) carries more prognostic weight than any single static value, given how many confounders (epinephrine dosing, airway type, compression quality itself) affect the absolute number. |
| Abrupt rise during CPR | A sudden jump in EtCO2 mid-resuscitation is one of the most reliable non-invasive signals that ROSC has just occurred — often visible before a pulse check confirms it. |
Confirmed CO2 detection after intubation remains the single most specific bedside method to confirm correct tracheal (versus esophageal) tube placement during CPR — that role is settled; the prognostic-threshold role is genuinely still being refined.
Clinical Applications by Setting
Traumatic brain injury: target EtCO2 35–40 mmHg. Avoid hyperventilation outside of brief, deliberate use for acute herniation with Cushing's reflex — cerebral vasoconstriction from over-ventilation can worsen ischemia in an already injured brain.
Post-ROSC: target EtCO2 35–45 mmHg — avoid both hyper- and hypoventilation in the immediate post-arrest period, when cerebral autoregulation is impaired.
Sepsis / DKA / other compensatory tachypnea: match the ventilator to what the patient's own respiratory drive is already doing rather than "normalizing" a compensatory hyperventilation — a metabolic acidosis being compensated by a low PaCO2 needs that low PaCO2 maintained, not corrected.
Procedural sedation: capnography detects hypoventilation before it becomes hypoxemia — often the earliest available warning sign, arriving before the pulse oximeter changes at all, especially with supplemental oxygen running.
If you're bagging a patient without capnography, you're ventilating blind — the waveform (or its absence) is the fastest confirmation of both airway position and adequacy of ventilation available at the bedside.
Quick Reference Card & Primary Sources
| Pattern / value | Bedside meaning |
|---|---|
| Normal EtCO2 35–45 mmHg | Adequate ventilation, perfusion, metabolism |
| Shark fin | Bronchospasm / partial obstruction — bronchodilate, check tube |
| Baseline not returning to zero | Rebreathing — check absorber, flow, circuit |
| Sudden flatline, mid-CPR | Loss of compressions/output, not necessarily esophageal placement |
| EtCO2 <10 at 20 min CPR | Very poor prognosis (~0.5% ROSC in pooled data) |
| EtCO2 ≥20 at 20 min CPR | Associated with better odds; ILCOR advises against any cutoff alone |
| Abrupt EtCO2 rise mid-CPR | Suggests ROSC has just occurred |
Primary sources checked for this guide:
- Capnography during cardiac arrest. Resuscitation, reviewed 2018.
- The prognostic value of end-tidal carbon dioxide during cardiac arrest: a systematic review. Resuscitation, 2013.
- End-tidal CO2 trajectory-based prognostication of out-of-hospital cardiac arrest. Reviewed via PMC, 2024.
- 2020 International Liaison Committee on Resuscitation (ILCOR) consensus on science and treatment recommendations, EtCO2 during CPR.
- Grmec S, Lah K, Tusek-Bunc K. Difference in end-tidal CO2 between asphyxial and VF/VT cardiac arrest in the prehospital setting. Crit Care. 2003;7:R139-44.
- Pantazopoulos C, et al. Capnography in emergency medicine. Eur J Emerg Med. 2018.
- Kodali BS. Capnography outside the operating rooms. Anesthesiology. 2013;118(1):192-201.
- Silvestri S, et al. The effectiveness of out-of-hospital continuous EtCO2 monitoring on the rate of unrecognized misplaced intubation. Ann Emerg Med. 2005;45(5):497-503.
- Link MS, et al. Part 7: Adult Advanced Cardiovascular Life Support. Circulation. 2015;132(18 Suppl 2):S444-64.