A Leak in a Capnography Sampling Line Induced a Difference between Arterial and End-tidal CO²

To the Editor:—

Capnography is widely used to monitor respiration during anesthesia. Herein, we report how a pinhole in a sampling line caused a problem with capnography during anesthesia.

A 59-yr-old man was scheduled for sinus surgery. Anesthesia was induced with propofol (50 mg) and vecuronium bromide (6 mg) intravenously, and the trachea was intubated with an endotracheal tube uneventfully. Immediately after induction, the capnography monitor (BP508, sidestream type; Nippon Colin Co., Komaki, Japan) showed an end-tidal carbon dioxide (ETco²) value of 35–37 mmHg. After disinfecting around his nose, the patient’s face was covered with sterile drapes. ETco²decreased suddenly to approximately 20 mmHg, but the patient’s vital signs did not change. We checked the connection of the endotracheal tube, checked for leaks in the endotracheal tube cuff, and checked the respirator circuit and auscultation of the chest but did not find any abnormality. Next, we checked the blood gases, and the arterial carbon dioxide tension (Paco²) was 50 mmHg. We reconnected the capnography plug and recalibrated the capnograph, but ETco²was unchanged. Next, we replaced the capnograph with another capnograph without exchanging the sampling line, but ETco²remained low. Then, we exchanged the capnograph including the sampling line. ETco²was then 48 mmHg.

After the operation, we examined the original sampling line. Although it seemed normal, when it was submerged in water, we found a pinhole in the sampling line near the endotracheal tube connector.

In a later experiment, we attempted to reproduce this observation. For this test, we used a test lung (Lung simulator SMS; Datex-Ohmeda, Helsinki, Finland) supplied with carbon dioxide and ventilated with an oxygen–air mixture. A capnograph similar to that used in the operating room was tested. Respiration was controlled with a respirator at a rate of 6 breaths/min, a tidal volume of 500 ml, and a fraction of inspired oxygen (Fio²) of 0.4. We made a pinhole in the sampling line with a 23-gauge needle, and ETco²immediately decreased from 35 mmHg to 24 mmHg (fig. 1). We then enlarged the pinhole with an 18-gauge needle, and ETco²decreased from 24 mmHg to 14 mmHg (fig. 1).

Factors that increase the difference between arterial and end-tidal carbon dioxide (P^(a-ET)co²) include increased pulmonary ventilation/perfusion mismatch and gas sampling problems. 1The situation described is an example of a gas sampling problem caused by enhancement of room air via  the unrecognized pinhole.

We did not detect the pinhole before anesthesia, and no pinhole was found when we tested for a leak after the operation, which suggests that a pinhole is difficult to find. Therefore, we recommend changing the sampling line of the capnograph regularly and considering a small leak in the sampling line when seeking the cause of an increase in P^(a-ET)co².