Respiratory Monitoring in the ED

The main way to monitor ventilation in the ED is via end-tidal CO₂ (ETCO₂) measurements.

ETCO₂ is measured via capnometers (carbon dioxide meters), which can be either qualitative (“Is there ETCO₂?”) or quantitative (“What is the ETCO₂?”).

To answer the question “Is CO₂ present?”, we use colourimetric capnometers. These are qualitative or semi-quantitative. Colourimetric capnographers change colour depending on the ETCO₂ concentration, using a pH sensitive strip. Unfortunately they are only 80% sensitive in cardiac arrest situations.

To answer the question “What is the CO₂?”, we use continuous infrared spectroscopy (CIS) capnometers, which are quantitative and measure the partial pressure (or %) of CO₂ in expired air (rather than pH). The advantage is that CIS capnometers can give more than a “yes or no” answer. They provide a number as well as a waveform (“capnography”), and are quickly becoming the standard of care in emergency intubation situations. They are already the standard of care for elective intubations in theatre.

A typical capnograph looks like this:

On the Y-axis is the CO_2, and the X-axis is time. Notice that the first part of expiration the CO₂ level is low, this is due to the expired dead space. Also note that the CO₂ does not start dropping until inspiration starts.

The answer is no, except in specific circumstances.

In most people, the ETCO₂ is lower than the PaCO₂. This is known as the “ETCO₂ – PaCO₂ gradient”. In healthy people, the normal gradient is 4-5 mmHg. However, this difference is unpredictable and is unreliable to be used as a measure.

The gradient may be increased by:

1. Reduced pulmonary blood flow — e.g following cardiac arrest, hypovolaemia, and shock.
2. Anatomical dead space
3. COPD

Sadly, no, the gradient is unstable. In studies with anaesthetised patients, the gradient remains stable in only 60-80% of patients.

A high ETCO₂ invariably means there is a high PaCO₂, which generally means impending respiratory failure! As a side note, the PvCO₂ is also unreliable as a marker of PaCO₂.

1.   To help verify correct placement of the endotracheal tube (ETT).

• Particularly during the peri-intubation period, as the mortality of unrecognised oesophageal intubation approaches 100%…
• It can also provide rapid recognition of extubation, for example during ongoing CPR or with transport of the ventilated patient.

2.   As an adjunct to procedural sedation.

• It provides an early warning of bradypneoa and apnoea predicting the development of hypoxia due to respiratory depression up to a minute prior to pulse oximetry.
• In real time, therefore, it is superior to pulse oximetry in identifying oversedation and apnoea.

3.   In Traumatic Brain Injury to monitor for hypocapneoa

• Hyperventilation drives hypocapnoea, which MAY worsen the outcome in traumatic brain injury, by decreasing coronary blood flow resulting in areas of brain ischaemia.
• Although the arterial CO₂ (PaCO₂ ) is the gold standard measurement, the ETCO₂ may guide ventilation, with the caveat regarding unstable ETCO₂–PaCO₂ gradients.

4.   To help predict the prognosis of ongoing CPR.

• An ETCO₂ of less than 10mmHg following 20 minutes of CPR is a very poor prognostic sign, suggesting resuscitation attempts should be ceased.

5.   To detect Return of Spontaneous Circulation (ROSC).

• A sudden rise in ETCO₂ may indicate ROSC. Rhythm and pulse checks should be considered.

Pulse oximetry needs a good waveform which is a measure of the arterio-venous (AV) difference. Without a good AV difference, and hence waveform, it is inaccurate. Things that affect the AV difference include hypotension, hypothermia, and peripheral vasoconstriction.

Accuracy:

• Between 70-100% the accuracy is good (+/- 2% for each standard deviation)
• Between 50-70% the accuracy decreases somewhat  (+/- 3%)
• Below 50% pulse oximetry is inaccurate

… although we would hope that once saturations reached 50% there would be other clinical markers of hypoxia…

Pulse oximetry (using modern devices) is NOT affected by…

• Anaemia (it is still accurate down to 2.3 g/dL Hb)
• Acidemia
• Ambient Light
• Dark Skin
• Nail Polish (red is ok, although darker colours may affect the reading to a mild degree)

Pulse oximetry does not show current oxygenation – it shows PAST oxygenation.

This lag may be up to 2-3 minutes in critically ill patients, due to reduced blood flow from vasoconstriction or hypothermia. In this respect, it is not a real time monitor, although forehead pulse oximetry probes have less lag.

Although newer pulse oximetry monitors can differentiate between different haemoglobin states, most hospitals do not have these monitors. Met-haemoglobinaemia and carboxy-haemoglobinaemia both result in PaO2 overestimation, as these haemoglobins absorb a similar light wavelength to oxy-haemoglobin.

To reduce hypoxemic episodes, especially whenever definitive airway management is being obtained. It may also be used to titrate FiO₂ – to prevent hyperoxaemia.

Pulse oximetry DOES NOT confirm ETT placement.

In the words of the authors, (which is true of all monitors, tests and investigations), never rely on monitoring over clinical judgement!