By continuously monitoring blood gases through a signal transmitted via fiber-optic catheter from a sensor threaded into an artery, clinicians can obtain updated readings every second.

By India Smith

Continuous in vivo blood gas monitoring is not new. Abstracts reporting initial experience with the technology date back more than 10 years. The first device, designed for adult and pediatric use, received approval from the Food and Drug Administration in 1991. A neonatal version was approved in 1998, yet only a handful of institutions currently use the monitors routinely. Although they have proven to be highly accurate and reliable, they are also costly, technically challenging, and suitable only for the most critically ill patients. Most RCPs have never seen one outside the animal laboratory.

This situation, however, may be about to change. With the growth of high frequency oscillatory ventilation and inhaled nitric oxide therapy, and the introduction of new strategies for optimizing conventional ventilation, there is also a need for better, faster ways to implement these strategies. Thus, there is increasing interest in a monitor that can provide instantaneous blood gas information.

So contends Irwin Weiss, MD, director of pediatric critical care medicine at UCLA Medical Center in Los Angeles. According to Weiss, the continuous monitor uses a signal transmitted via fiber-optic catheter from a sensor threaded into an artery—either the radial, femoral, or, in the case of neonates, umbilical artery. A console displays real-time readings of body temperature, Po2, pH, and Pco2, and the readings update every second.1

For practitioners who are accustomed to intermittent blood gases from specimens that are drawn and sent to a laboratory, it is a radically different way to look at the patient. “It is surprising the first time you see it,” says Weiss, who admits he was somewhat shocked to discover how rapidly blood gases can fluctuate in critically ill patients. “We saw things we had never known about—even temperature changes. We saw fever spikes we had never known were there,” he says.2

He anticipates many applications for the technology, including acute respiratory distress syndrome, asthma crises, and cardiac surgery—particularly in patients with poor cardiac function where acidosis is a constant concern. “Any time you need to know the blood gas values immediately, can’t wait, or don’t want to wait, this could give you the information,” Weiss says.

According to Patricia Meyers, RRT, development leader in the Infant Pulmonary Research Center of the Children’s Hospitals and Clinics in St Paul, continuous blood gas monitoring allows settings on the high frequency oscillator to be optimized almost immediately. “With continuous monitoring, you can see second-to-second changes, whereas with conventional blood gases, you first have to get a specimen drawn, run it to the laboratory, and wait for it to be tested. By then, everything may have changed in the child,” Meyers says. Even with a point-of-care instrument, Meyers estimates that conventional blood gases require 4-5 minutes to obtain results. “With continuous monitoring, I can have the baby optimized in that time,” she says.

For clinicians who are not accustomed to the pace, however, the continuous updating of blood gas information can take some getting used to. “The first time we used the inline blood gas monitor with the oscillator, rounds were taking place in the unit. The readout on the monitor showed that we needed to reduce the mean airway pressure (MAP) on the oscillator, so I interrupted the nurse practitioner who was rounding to get an order to do that. A minute later I had to interrupt her again because the blood gas monitor was updating every second and it showed that we needed to reduce the MAP even further. Finally she broke off rounds and stood watching the monitor,” Meyers recounts.

Although most of her patients who receive continuous blood gas monitoring are intubated and on ventilators, the device can also be used to avoid ventilation and the potential complications associated with it. Meyers describes a case of a neonate with an atypical diaphragmatic hernia who received an inline blood gas monitor that revealed he was maintaining adequate values breathing on his own. “Normally, he would have been intubated right away, but this child was able to make it right up to the time of surgery without mechanical ventilation. It was only because we were able to monitor him continuously that we felt secure in letting him do that,” Meyers says.

She describes another case of a newborn with meconium aspiration who initially received an inline blood gas monitor and was started on the high frequency oscillator, but later had to be transferred to another hospital for extracorporeal membrane oxygenation (ECMO). “The ECMO team decided to leave the sensor in place and found that the monitoring was very useful,” Meyers says. Although the ECMO circuit permits the use of an inline Po2 monitor, full blood gases are usually tested intermittently. Furthermore, blood gas levels in the baby may be different from the levels in the ECMO circuit. “Ideally, you should be able to monitor blood gases continuously in the baby and on both the arterial and venous sides of the circuit,” Meyers says.

“It is potentially very insightful,” says Gordon Lassen, BS, RRT, director of respiratory care at Utah Valley Regional Medical Center, Provo, Utah. Lassen has been using the device in the animal laboratory as part of a training course on the high frequency oscillator, where the strategy for optimizing lung volume is quite different from conventional ventilation.

“Optimization is really reflected in only two ways: Po2 and x-rays. That is about all we have right now to tell us that we are optimizing lung volume and stabilizing the lung,” Lassen says. With continuous inline blood gas monitoring, he is able to obtain real-time Po2. “We plot it on the computer so students can see that as you adjust lung volumes, Po2 adjusts accordingly,” he continues. “Of course, it is a little more complicated than that because you have to balance the ventilation-perfusion ratio to manage the vascular space, but the inline blood gas gives an instantaneous look at what your adjustments have done to the Po2 within the animal,” Lassen says.3

He also intends to use continuous blood gases in research on interventions in the neonatal intensive care unit (NICU) based on oximetry. “We are so ingrained in our use of oximetry, and in what a “normal” range is, that I wonder what is happening in the patient,” Lassen says. To illustrate, he describes a recent incident in which he was monitoring a newborn with continuous blood gases showing a Po2 in the 40s. “At the same time, the oximeter was showing a saturation of 91%-92%. I watched a nurse go over and turn down the oxygen a little bit because the saturation was reading a little high,” he recounts. Admittedly, the nurse was not accustomed to using the real-time blood gas monitor.

“But the neonate is very different from the adult,” Lassen says. Because critically ill newborns display such rapid instability in their hemodynamics and blood gases, he wonders whether clinicians are making too few interventions and allowing too many delays before making ventilator changes. “With the micro-preemie, we are concerned about the Po2 level in the blood and its effect on the eyes. So, I am interested in whether practitioners are becoming desensitized to oximetry and not attuned to the upper limits of oxygenation,” Lassen says.

“If the stats go up to 98%-99% and they mute the alarms, or if they are busy elsewhere and the alarms just become white noise in the unit, they don’t realize that the baby has an extreme Po2 and it just sits there for 10 or 15 minutes before somebody acts on it: that raises a lot of questions,” Lassen says. These are questions he believes can be addressed with studies using continuous blood gas monitoring.

Lassen also points out that some bigger babies in the NICU are able to be somewhat active. This activity in conjunction with lung disease, and the fact that the infant is on a ventilator, can cause frequent changes in oxygen saturation. “But instead of adjusting Fio2 or ventilator parameters, maybe the answer is simply to go over, comfort the babies, and let them restabilize on their own,” Lassen says. Again, a study using continuous blood gas monitoring could answer the question.

According to Mark Heulitt, MD, associate professor of pediatrics, critical care medicine, neonatology, and ECMO at Arkansas Children’s Hospital in Little Rock, the greatest interest in adopting in vivo blood gas monitoring clinically is likely to be in the NICU not only because many practitioners have had a chance to use it during training on the high frequency oscillator, but also because it could potentially reduce the cost of intermittent blood gases. “Normally, we don’t test as many blood gases in the pediatric ICU, but we do test a lot of blood gases in the NICU,” Heulitt says. It is not unusual to do eight or 10 sets of blood gases in the first 2 hours of treatment in the NICU. There, the technology might conceivably pay for itself. It could also reduce blood sampling—a major concern when patients are very small. Transfusion requirements might likewise be lessened, which are both a clinical benefit and a further cost saving.

“However, if we decide to commit to something like this, we want to be able to commit to it across the whole hospital—PICU, NICU, and the burn unit,” Heulitt says. It is a complex decision in which the advantages must be weighed against cost, technical sophistication required, and the investment the hospital has already made in blood gas instruments and staffing.4 “But it would be ideal for patients who are critically ill who need their blood gases looked at over a very short interval of time,” he says.

Heulitt also considers it a good trending tool. According to Meyers, the clinician can go back as far as 24 hours to see what changes have been made and what has been accomplished. If the trending time is put back to 24 hours, the user can scroll across and see past values displayed next to current values. “Or, if a nurse arrives in the morning and wants to see what happened during the night, she can put up a trending screen,” Meyers says.

In her studies on the accuracy of the device,5 Meyers has found it to last for as long as 5 days. Normally, it is inserted immediately after an umbilical artery catheter is placed, but has been successfully inserted up to 36 hours later. Meyers has attempted to place one 5 days out, but was not able to get it to work. “When we took it out, it had a very fine thread of fibrin hanging on the end,” she reports. The arterial catheters develop fibrin deposits over time, so sensors that are inserted later are believed to pick up debris that interfere with their functioning.

The devices also function better in high position, advanced to the level of T-8 to T-10, rather than in the original low position at L-3 to L-4. “It is in a larger vessel, there is more turbulence around it, and I think things move better,” Meyers says.

According to Mark Mammel, MD, director of neonatal research and education at the Children’s Hospitals and Clinics in St Paul, because the device is somewhat different from a regular umbilical artery catheter, it needs to be secured differently. “And like all fiber-optic devices, if you clamp or kink it in the wrong place, it breaks. So you have to be more careful how you place babies when you move them into the prone position,” Mammel warns. The actual insertion is not difficult. In Mammel’s program, a respiratory therapist inserts the catheter while a nurse practitioner, neonatologist, or fellow who also knows how to insert it is present, providing a second pair of eyes; placement is confirmed with an x-ray. Mammel now uses it as a routine monitoring tool. Unlike the Swan-Ganz catheter, he does not believe it presents any issues of interpretation nor does it complicate decision-making. “You run into problems with new technology when it gives you information that you have never had before, or information that you have had but so infrequently that you don’t have a sense of how to integrate it into management. We are so accustomed to using blood gases that anyone looking at this monitor has an immediate sense of what they ought to do,” he says.

Especially in dealing with very sick patients, babies on oscillators with pulmonary hypertension, Mammel finds the monitor is a tremendous help because such patients change so quickly. “And you can alter your treatment before you are dealing with a complication. Maybe you can anticipate something that’s about to happen, as opposed to waiting to get a chest x-ray to show you have an air leak, or that the patient’s lungs have changed and he is now acidotic and hypoxic. It enables you to act more in concert with the changing illness as opposed to reacting after the fact.”

According to Mammel, because outcome is multifactorial, a trial will be needed to determine whether the technology is starting to improve outcomes.6 “But when we have this information available, I always feel that we are able to provide care in the most rapid fashion possible. It gets rid of all the issues of lag time between taking a blood gas specimen to the laboratory and wondering when you get the results back if things are really the same.”

Weiss agrees. He is increasingly interested in applying continuous blood gas monitoring in situations in which blood gas values can change significantly within a short time and without much warning, or without triggering other forms of monitoring.7 “I am thinking of a patient with head trauma. Do you want to do a blood gas every 3 hours or do you want to be able to monitor them continuously?”

Weiss has used the sensor several times to monitor extubation, and finds it an excellent tool for showing whether patients are holding their own. Weiss uses several parameters, but with the continuous blood gas monitoring, he is not as worried about whether the patient is ventilating adequately. “While patients may look OK, their CO2 can rise to high levels pretty fast. We can avoid that with continuous-line blood gases,” Weiss says.

He describes a case of an adolescent in respiratory failure who received continuous blood gas monitoring, allowing the team to institute multiple interventions over a span of 5 1/2 hours. “We were able to try multiple interventions one after the other because we were never waiting for a blood gas. As soon as we saw that one thing was not working, we were able to try something else. Unfortunately, that girl did not make it. But we decided to write up the case to demonstrate the efficacy of the monitoring in extreme situations,” he says.

Like Mammel, Weiss does not believe that the technology overloads clinicians or makes management more complex. “I don’t see how it could,” he says. “The essence of life is your pH, Pco2, and Po2. It is not the systemic vascular resistance or theoretical oxygen supply-demand curves. Those are the kinds of numbers you get from the Swan-Ganz catheter and that’s why there is a controversy over the Swan. Do those kinds of numbers actually mean anything?

“But here we are talking about your Po2 and your pH—the essence of life. We know exactly what to do with that information.”


India Smith is a contributing writer for RT Magazine.


1. Meruva RK, Meyerhoff ME. Catheter-type electrode for potentiometric monitoring of oxygen, pH, and carbon dioxide. Biosens Bioelectron. 1998;13:201-212.
2. Weiss IK, Fink S, Harrison R, Feldman JD, Brill JE. Clinical assessment in continuous arterial blood gas monitoring in the pediatric intensive care unit. Pediatrics. 1999;103:440-445.
3. Morgan C, Dear PR, Newell SJ. Relationship of oscillatory amplitude and ABG changes in premature infants on oscillator using MPIAS. Arch Dis Child Fetal Neonatal Ed. 2000;82:237-242.
4. Mahutte CK. Intravascular versus extravascular “on demand” blood gas systems. Clin Biochem. 1998;31:119-130.
5. Meyers P, Worwa C, Trusty R, Mammel M. Continuous arterial blood gas monitoring accuracy and reliability—this time through an umbilical artery catheter (UAC). Respir Care. 2000;45:1006.
6. Trowbridge CC. Effects of continuous blood gas monitoring during cardiopulmonary bypass: a prospective randomized trial, Part III. J Extracorporeal Technol. 2000;32:129-137.
7. Zaugg M, Luchinetti E, Zalunardo MP, et al. Substantial changes in arterial blood gases during thorascopic surgery can be missed by conventional intermittent laboratory blood gas analyses. Anesth Analg. 1998;87:647-653.