Pressure-Limited Ventilation of Infants with Low-Compliance Lungs: The Efficacy of an Adult Circle System Versus Two Free-Standing Intensive Care Unit Ventilator Systems Using an In Vitro Model

Stevenson, G. W. MD*; Horn, Babette MD*; Tobin, Michael MD*; Chen, Edwin H. PhD‡; Sautel, Michael RRT†; Hall, Steven C. MD*; Coté, Charles J. MD*

Anesthesia & Analgesia:
doi: 10.1213/00000539-199909000-00017
Pediatric Anesthesia

We compared the efficacy of a Drager Narkomed GS (North American Drager, Telford, PA) equipped with an adult circle system with two free-standing infant ventilator systems (Servo 300; Siemens Medical Systems, Danvers, MA and Babylog 8000; North American Drager) to deliver minute ventilation (VE) using pressure-limited ventilation to a test lung set to low compliance. To simulate a wide variety of potential patterns of ventilation, VE was measured at peak inspiratory pressures (PIP) of 20, 30, 40, and 50 cm H2O and at respiratory rates (RR) of 20, 30, 40, and 50 breaths/min. Each measurement was made three times; the average was used for data analysis using the multiple regression technique. Delivered VE was positively correlated with both PIP (

P = 0.001) and RR (P = 0.001). Only minimal differences in VE were observed between the circle and the two free-standing systems. At lower RR and PIP, the Babylog 8000 system delivered slightly higher VE than the circle system, whereas at higher RR and PIP, the Babylog 8000 delivered slightly lower VE than the circle system; these differences in VE were not statistically significant (P = 0.45). The Servo 300 delivered slightly higher VE than the circle system in all test conditions, but these differences were not statistically significant (P = 0.09). None of the differences in delivered VE between the Servo 300 and the circle system are of clinical importance.

Implications: Our laboratory investigation suggests that pressure-limited ventilation delivered by a standard adult circle system compares favorably with that of freestanding infant ventilators used in pressure-limited mode. Changing from an adult circle system to a free-standing pressure-limited ventilator may not substantially improve ventilation of a low-compliance infant lung; the efficacy of such a practice should be investigated.

Author Information

*Department of Anesthesiology, Children’s Memorial Hospital and †Division of Respiratory Care, Northwestern University Medical School; and ‡Department of Biostatistics, School of Public Health, University of Illinois at Chicago, Chicago, Illinois

April 22, 1999.

Address correspondence and reprint requests to G. W. Stevenson, MD, Department of Anesthesiology, Children’s Memorial Hospital, 2300 Children’s Plaza, Chicago, IL 60614.

Article Outline

Using standard adult circle systems (adult bellows, standard CO2 canister) for all children, including infants, allows for the use of lower fresh gas flows, which reduces costs and simplifies equipment requirements (1–4). Routine delivery of pediatric anesthesia via adult circle systems may also result in improved patient safety and reduce operating room (OR) turnover time by standardizing equipment for all patients (4). However, some practitioners believe that infants with low lung compliance are best ventilated in the operative setting by using the free-standing pressure-limited ventilators often used in the intensive care unit (ICU), rather than standard anesthesia machine ventilators (4). There are little data to support or refute this practice.

The purpose of this study was to compare delivered minute ventilation (VE) with a test lung set in low-compliance infant mode, using a Drager system (Narkomed GS; North American Drager, Telford, PA) equipped with a standard adult circle versus two free-standing pressure-limited ventilators often used in neonatal ICUs during pressure-limited ventilation. VE measurements were made over a range of target peak inspiratory pressures (PIP) and respiratory rates (RR) to simulate a wide variety of potential ventilation patterns used for ventilation of infants with low lung compliance.

Back to Top | Article Outline


The ventilator and test lung model in this study have been previously described (5–7). VE delivered by all ventilation systems tested was measured using a Bio-Tek test lung (Bio-Tek Instruments, Inc., Winooski, VT). The Bio-Tek test lung consists of wire wool-filled metal canisters capable of compliance adjustment to simulate both healthy (normal compliance) and abnormal (low compliance) infant lungs, as defined by the American National Standards Institute (8). The Bio-Tek test lung measures VE delivered to the lung with a computer-driven analysis of expired flow volume curves (accuracy ±4%). The Bio-Tek test lung was connected to the ventilation systems via a 3.5-mm inner diameter endotracheal tube (ETT) cut distally (removing the Murphy eye to prevent system leakage), with 15-mm connectors on both ends (Fig. 1). The Bio-Tek test lung was adjusted for ambient barometric pressure, temperature, and humidity before testing. At the recommendation of the Bio-Tek test lung manufacturer, the least restrictive adapter (Rp20) connected the ETT to the test lung. The Bio-Tek test lung was set in the low-compliance infant mode for all testing.

VE delivered to the test lung via a circle system was compared with VE delivered by two free-standing ventilation systems. The circle system tested was a Drager Narkomed GS circle system equipped with an adult bellows and a disposable circuit. The two freestanding ventilation systems were a Drager Babylog 8000 (North American Drager) and a Servo 300 (Siemens Medical Systems, Inc., Danvers, MA). Both the Babylog 8000 and the Servo 300 ventilation systems were equipped with an infant respiratory circuit and humidifier. The humidifier was filled to the appropriate level with a polyurethane elastomer, rather than water, to prevent adverse effects on the test lung assembly.

Regardless of the system being tested, an inspiratory to expiratory flow ratio (I:E) of 1:2 was maintained. No significant leaks in any ventilation system were detected before experimental testing. Three trials were conducted to measure each VE data point. To replicate a variety of ventilator settings that might clinically occur during ventilation of an infant with low-compliance lungs, VE was measured using all test systems with RR set at 20, 30, 40, or 50 breaths/min with target PIP of 20, 30, 40, and 50 cm H2O. For all three ventilation systems, target PIP were achieved by adjusting the inspiratory pressure limit (pressure-limited ventilation). The multiple regression technique was used to analyze the data. The dependent variable was VE; independent variables were the ventilator delivery system being used (circle, Babylog 8000, and Servo 300 coded into two dummy variables) and RR, PIP, and their interaction terms.

Back to Top | Article Outline


Three measurements of VE were made at each combination of RR and PIP for each ventilation system; the average of these measurements was used for data analysis. VE measurements at given combinations of RR and PIP were very consistent. The largest coefficient of variation (the standard deviation expressed as a percentage of the mean) was 1.6% for the circle system, 7.6% for the Babylog 8000, and 4.9% for the Servo 300.

The three systems produced similar VE (Figs. 2 and 3). When both RR and PIP were low (RR 20 or 30 breaths/min; PIP 20 or 30 cm H2O), the Babylog 8000 produced slightly higher VE than the circle system in three of four test conditions, with an average increase in VE of 2.5% with respect to VE produced by the circle system (−2.0% to 7.7%). For the 12 remaining test conditions (RR = 40 or 50 breaths/min or PIP 40 or 50 cm H2O), the Babylog 8000 produced slightly lower VE than the circle system in 9 of 12 test conditions (mean −2.8%, range −12.4% to 4.3%). These differences were not statistically significant (P = 0.45). The Servo 300 produced slightly higher VE than the circle system in 16 of 16 combinations of PIP and RR, with an average increase in VE of 8.1% (2.1%–12.9%). These differences did not reach statistical significance (P = 0.09). Detailed results of data analysis are shown in Appendix 1.

Back to Top | Article Outline


When an infant with very low-compliance lungs requires surgery and anesthesia, some clinicians believe that free-standing pressure-limited ventilation systems used in the neonatal ICU may offer an advantage over standard adult circle ventilation systems (9). For this reason, such infants are often transported to the OR along with their freestanding ICU ventilator; alternatively, an ICU ventilator is substituted intraoperatively should high PIP be required for ventilation. This practice may present a danger to the patient should the anesthesiologist not be completely familiar with use of these non-OR ventilators. There are no studies comparing the efficacy of the OR use of free-standing ICU ventilators with that of standard adult circle systems to justify this practice.

In previous studies, using the same infant lung model set both to low and normal compliance, we examined the effect of circuit compliance and ventilator mode set-up on VE using a standard adult circle system. We found that neither the compliance of the circuit tubing, the precise mode of ventilator set-up, nor the endotracheal tube size connecting the test lung to the circle system influenced VE delivered to the low-compliance infant lung model; VE was determined by PIP and RR (5,6). In a separate study using the same lung model, we determined that a Bain system does not offer any advantage over a standard adult circle system in VE delivered during ventilation of low-compliance infant lungs; VE was determined by PIP and RR (7).

In the current study, we compared VE delivered to a low-compliance infant test lung using a commonly available standard adult circle system with that delivered by two free-standing infant ICU ventilator systems all set in pressure-limited mode. We compared VE delivered by the three ventilation systems with the test lung set in the low-compliance mode. We assumed that the intraoperative use of a free-standing ICU ventilator would be considered only in infants with low lung compliance. Our results show that there is little difference in VE delivered by the adult circle system versus the two free-standing systems tested. The Babylog 8000 system tended to deliver slightly higher VE than the circle system at lower RR and PIP and slightly lower VE at higher RR and PIP. These differences did not reach statistical significance, and would certainly not be of clinical importance. The Servo 300 system consistently delivered slightly higher VE than the circle system (an average increase in VE of 8.1%), but it is unlikely that this difference is of clinical importance.

Our study did not compare VE between systems during volume-limited ventilation because most currently available anesthesia adult circle systems are equipped with ventilators that are not easily adjusted to infant tidal volumes during volume-limited ventilation. In addition, if volume-limited ventilation is attempted in infants, it is very difficult to predetermine the appropriate tidal volume per kilogram that will result in adequate chest expansion (25–300 mL/kg, depending on the size of the infant) (9). More typically, pressure-limited ventilation is used in infants with low-compliance lungs: ventilation is achieved by upward adjustment of the PIP until satisfactory endpoints of ventilation are achieved. Our earlier in vitro studies with the Drager GS circle system showed equivalence of VE at a given PIP regardless of the precise mode of ventilation (pressure-limited versus volume-limited) used to achieve that PIP. The results of our current study would probably have been similar had we chosen to compare the three test systems during volume-limited ventilation rather than pressure-limited ventilation. We used an I:E. ratio of 1:2 for all VE comparisons in the present study. We cannot rule out the possibility that more significant variations in VE would have been observed between systems if other I:E ratios had been tested. We also cannot rule out the possibility that other free-standing ventilator systems would have performed better than the Servo 300 or the Drager 8000 systems.

If our in vitro study can be extrapolated to the clinical setting, there seems to be limited clinical advantage to switching from a standard adult circle system to a free-standing pressure-limited infant ventilator should high PIP or RR be required for pressure-limited ventilation of an infant with low-compliance lungs in the operative setting. Our results, coupled with the potential intrinsic harm of using an unfamiliar ventilator (that often is not compatible with anesthetic vapor delivery), suggest that such a practice be investigated. Our study does not rule out the possibility that free-standing ICU ventilators offer other ventilation modalities that might be of benefit to some subset of infants in the operative setting.

Back to Top | Article Outline


1. Badgwell JM. Delivery and monitoring of alveolar ventilation. In: Badgwell JM, ed. Clinical pediatric anesthesia. New York: Lippincott-Raven, 1997: 113–43.
2. Coté CJ. Pediatric breathing circuits and anesthesia machines. Int Anesthesiol Clin 1992; 30: 51–61.
3. Fisher DM. Anesthesia equipment for pediatrics. In: Gregory GA, ed. Pediatric anesthesia. New York: Churchill Livingstone, 1994: 197–225.
4. Steven JM, Cohen DE, Sclabassi RJ. Anesthesia equipment and monitoring. In: Motoyama EK, Davis PJ, eds. Smith’s anesthesia for infants and children. St. Louis, MO: Mosby, 1996: 229–79.
5. Stevenson GW, Tobin MJ, Horn BJ, et al. The effect of circuit compliance on delivered ventilation with use of an adult circle system for time cycled volume controlled ventilation using an infant lung model. Paediatr Anaesth 1998; 8: 139–44.
6. Tobin MJ, Stevenson GW, Horn BJ, et al. A comparison of three modes of ventilation with the use of an adult circle system in an infant lung model. Anesth Analg 1998; 87: 766–71.
7. Stevenson GW, Tobin MJ, Horn BJ, et al. An adult system versus a Bain system: comparative ability to deliver minute ventilation to an infant lung model with pressure limited ventilation. Anesth Analg 1998; 88: 527–30.
8. American National Standard for Breathing Machines for Medical Use. New York: American National Standards Institute, 1976: 7–32.
9. Badgwell JM, Swan J, Foster AC. Volume-controlled ventilation is made possible in infants by using compliant breathing circuits with large compression volume. Anesth Analg 1996; 82: 719–23.
Back to Top | Article Outline


© 1999 International Anesthesia Research Society