Transcutaneous monitoring: back to the future - An important adjunct to care during high frequency oscillatory ventilation

Effective use of HFOV requires close attention to lung volume, with use of an “optimal volume” strategy to open the lung and maintain it open [5]. Mean airway pressure is adjusted to minimize FO₂(I) requirement without evidence of under- or overdistention on chest X-ray. Continuous pulse oximetry assists with adjustments of mean airway pressure and FO₂(I).

Continuous assessment of CO₂ is also very important during HFOV. The oscillator is a powerful machine that can quickly drive arterial CO₂ to unsafe levels. Evidence is accumulating that suggests cerebral damage may result from hypocarbia [6,7]. 

Many infants on HFOV have indwelling arterial lines; however, frequent blood draws may be necessary to appropriately monitor CO₂ changes, leading to increased infection risk and/or anemia. A non-invasive, continuous estimate of pCO₂ during HFOV would be safer and more effective. Transcutaneous monitoring can provide this estimate.

Transcutaneous (tc) monitoring is not new; it has been available for well over twenty years [8]. Early machines were cumbersome and difficult to use. Accurate tcpO₂ assessment necessitated heating the skin to 43 °C, which often led to skin burns in small preterm infants. 

After the advent of pulse oximetry, use of tc monitoring faded in most NICUs. Unfortunately, this led to “throwing the baby out with the bathwater”, as tcpCO₂ monitoring also dramatically decreased despite the lack of a replacement for CO₂ monitoring such as pulse oximetry.

Currently available tc monitors are small and easy to use. Importantly, they can be used to monitor both tcpO₂ and tcpCO₂, or either one separately. 

Even more importantly, use of tcpCO₂ alone can accurately be done at a monitor temperature of 40 °C, thus not causing skin burns [9], and site changes can be done as infrequently as every six to eight hours. The machine must simply be calibrated at the appropriate temperature.

The tcpCO₂ will correlate with the pCO₂(aB) – that is, as one goes up the other goes up; as one goes down the other goes down. The “closeness” of the numbers will depend on the thickness of the skin and the perfusion of the site. 

The numbers are seldom identical, as they measure different things: one measures the pCO₂(aB) of arterial blood and the other the CO₂ diffusing from the cutaneous tissue. The numbers, however, will correlate (trend together).

It is important to check tcpCO₂ values with arterial blood gas samples or well-done capillary samples with each tc site change. Perfusion will vary somewhat from site to site, and thus the “closeness” of the numbers may also change. 

A rising tcpCO₂ should always be considered a patient problem until proven otherwise. Something often forgotten is that an increasing tcpCO₂ may of and by itself indicate decreasing perfusion in the patient – perhaps sepsis or impending shock. 

Though the tcpCO₂ will still trend correctly, the tcpCO₂ will be considerably higher than the pCO₂(aB) in a patient with significant circulatory compromise. In these cases the underlying cause of the problem must be treated.

“Something is wrong with the machine” is unfortunately often heard before evaluation of the patient has been done. A recent article, for example, documented the value of a rising tcpCO₂ in alerting staff to a pneumothorax well before acute decompensation of the patient [10]. 

A steadily rising or falling tcpCO₂ should prompt careful attention to reasons for under- or overventilation, not an immediate recalibration of the monitor or, worse, turning a blind eye to the readouts because “the machine is not working”.

Troubleshooting the tc monitor is relatively easy. The calibration cylinder must contain sufficient gas and must be turned on during calibration. The cable must be intact. The sensor must be remembraned as per the manufacturer’s recommendations. 

Sufficient contact fluid must be placed between the skin and the sensor. Recalibration should be done every six to eight hours if only tcpCO₂ is being used. We have found every six hours to be best in this circumstance; towards eight hours the contact fluid tends to evaporate, leading to spurious values. 

The sensor site should be changed every three to four hours if both tcpO₂ and tcpCO₂ are utilized. Heating of the sensor to 43 °C is needed if the tcpO₂ is employed, and the site must be changed more frequently to avoid skin burns. 

A tcpCO₂ value of 0 or tcpO₂ of about 150 means the sensor has dislodged or an air bubble is under the sensor. These are the values expected for room air. tcpCO₂ values that jump about wildly indicate need for recalibration/remembraning. Steadily rising or falling values reflect patient status.

Though pulse oximetry has largely replaced the need for tcpO₂ monitoring, tcpO₂ monitoring can provide useful and complimentary information should the practitioner choose to use it. High pO₂(aB) should be avoided in most cases [11,12]. 

Because of the shape of the oxygen-hemoglobin dissociation curve, an oxygen saturation in an acceptable range could be associated with a pO₂(aB) that is unnecessarily high. By the same token, a low or borderline saturation might be associated with an acceptable pO₂(aB) because of shifts in the oxygen-hemoglobin dissociation curve and varying amounts of fetal hemoglobin. 

tcpO₂ monitoring can be very useful in titrating the FO₂(I). Use of tcpO₂ monitoring requires more frequent site changes and close attention to the baby’s skin to avoid burns.

In most cases, however, the “burn” is a reddened area just under the sensor that heals without residua. Occasionally, tiny scars can result if the sensor temperature is too high or the sensor is left on the skin for too long.

For a patient being started on HFOV, the tcpCO₂ monitor should be placed on the patient prior to instituting HFOV. Once the tcpCO₂ is stable and a correlating ABG has been obtained, HFOV can be started and the amplitude adjusted using the tc monitor.

Severe hypocapnia, such as can occur with an inadvertently high amplitude or postsurfactant can thus be entirely avoided. Hypercapnia from tube secretions, tube malposition, accidental extubation, pneumothorax or insufficient amplitude can also be quickly noted and appropriate interventions given.

The tcpCO₂ monitor is also very valuable as the patient begins to wean, avoiding hypocarbia and allowing the staff to pace the wean appropriately.

Optimal use of HFOV should include concurrent use of tcpCO₂ monitoring to ensure prevention of hypo- and hypercarbia and timely interventions for both complications of therapy and patient weaning.