A comparison of three neonatal resuscitation devices

Transcription

1 Resuscitation 67 (2005) A comparison of three neonatal resuscitation devices Stacie Bennett, Neil N. Finer, Wade Rich, Yvonne Vaucher Department of Pediatrics, Division of Neonatology, UCSD Medical Center, 401 Dickenson, MPF 1-140, San Diego, CA , USA Received 31 October 2004; received in revised form 25 February 2005; accepted 25 February 2005 Abstract Background: Ventilation during neonatal resuscitation involves the use of self-inflating bags, flow-inflating bags, and T-piece resuscitators. The ability of operators to deliver desired peak inspiratory pressures (PIP), positive end expiratory pressures (PEEP), prolonged inflations and the length of time to transition between different pressures has not been compared for all three of these devices. Objective: To compare the ability of neonatal resuscitation personnel to deliver predetermined ventilation interventions using these devices in advance of a clinical trial of neonatal resuscitation. Design/methods: We studied 31 operators (neomatologists, neonatal respiratory therapists, neonatal fellows, a pediatrician, pediatric residents, neonatal nurse practitioners, and neonatal nurses) using a T-piece resuscitator (Neopuff, Fisher & Paykel Healthcare, Auckland, New Zealand), a self-inflating bag (Baby Blue II, Vital Signs, Totowa, NJ), and a flow-inflating bag (Model E191 Anesthesia Associates, San Marcos, CA). The self-inflating bag was tested with and without the manufacturer s PEEP valve. Using a continuous pressure recording system and a neonatal manikin, we evaluated the ability to deliver a consistent PIP of 20 or 40 cmh 2 O and a PEEP of 5 cmh 2 O during 30 s of ventilation, the ability to maintain a 5 s inflation at a PIP of 20 cmh 2 O and the time to transition from a PIP of 20 to 40 cmh 2 O. Each device was evaluated with and without a qualitative CO 2 detector (Pedicap Nellcor Pleasanton, CA). Results: The T-piece resuscitator delivered the desired PIP more precisely and consistently compared with the self-inflating bag at a target of 20 cmh 2 O (maximum PIP 20.7 cmh 2 O, S.D. = 0.8 versus 24.7 cmh 2 O, S.D. = 2.8; p < 0.001). At a target of 40 cmh 2 O, the maximum pressure delivered with the T-piece resuscitator was significantly less than both the flow-inflating bag and the self-inflating bag (39.7 cmh 2 O, S.D. = 2.1 versus 44 cmh 2 O, S.D. = 3.3 versus 45.3 cmh 2 O, S.D. = 4.7; p < 0.001). It took significantly longer to increase the PIP from 20 to 40 cmh 2 O using the T-piece resuscitator compared to the self-inflating bag or the flow-inflating bag (5.7 s versus 2.2 s versus 1.8 s; p < 0.001), and three operators could not make the transition in the allotted 15 s time limit. During the 5 s prolonged inflation, the T-piece resuscitator and the flow-inflating bag maintained a pressure greater than 18 cmh 2 O for a longer time than the self-inflating bag (4 s versus 3.7 s versus 2.2 s; p < 0.001). The self-inflating bag with the PEEP valve in place provided significantly less PEEP than both the T-piece resuscitator and the flow-inflating bag (3.6 cmh 2 O versus 4.4 cmh 2 O versus 4.4 cmh 2 O; p < 0.005). The Pedicap did not significantly affect any of the observed results, and there were no consistent operator differences between different disciplines or years of experience. Conclusions: The T-piece resuscitator delivered the desired pressures more accurately, but required greater time to increase the PIP from 20 to 40 cmh 2 O. It was difficult to maintain a prolonged inflation and deliver the desired PEEP with the self-inflating bag even with the PEEP valve in place. There is a need for improvement in the design and function of current manual resuscitation devices and for prospective trials to evaluate the optimal method of bag and mask ventilation during resuscitation of the newborn infant Elsevier Ireland Ltd. All rights reserved. Keywords: Bag and mask; Ventilation; Resuscitation; Manikin; Neopuff; Newborn infant; T-piece resuscitator; Self-inflating bag; Flow-inflating bag; Prolonged inflation; Positive end-expiratory pressure; Pedicap 1. Introduction A Spanish translated version of the Abstract and Keywords of this article appears as an Appendix at /j.resuscitation Corresponding author. Tel.: ; fax: address: scbennett@ucsd.edu (S. Bennett). Approximately 5 16% of neonates require some type of resuscitation at the time of birth [1,2]. Bag and mask ventilation is the most common intervention occurring in up to 10.6% of neonatal resuscitations [1], and is perhaps the most /$ see front matter 2005 Elsevier Ireland Ltd. All rights reserved. doi: /j.resuscitation

2 114 S. Bennett et al. / Resuscitation 67 (2005) problematic area of neonatal resuscitation. Currently there are three different types of bag and mask devices used in neonatal resuscitation. These devices include self-inflating bags, flow-inflating bags (also called anesthesia bags), and the T-piece resuscitator (such as the Neopuff ). At this time there is little evidence as to which device is the most effective for neonatal resuscitation. The most recent Neonatal Resuscitation Program (NRP) Textbook [2] describes the flow-inflating bag and the self-inflating bag and reviews the advantages and disadvantages of each device, but does not recommend one device over the other. This lack of recommendation is likely to reflect the limited evidence on the efficacy of each device. There is no mention of T-piece resuscitators although they are used in clinical practice. In a survey by O Donnell et al. [3], a questionnaire was sent to 46 NICUs in 23 countries. They reported that the Neopuff was used in 30% of centers. The use of all three bag mask ventilation systems including the T-piece resuscitator was discussed by Zideman et al. in their recommendations on resuscitation but no recommendations was given as to the superiority of any device [4]. Hussey et al. recently compared the Neopuff to the selfinflating and anesthesia bag with an intubated neonatal manikin [5]. They found that the self-inflating bag delivered a significantly higher peak inspiratory pressure (PIP) and lower positive end expiratory pressure (PEEP) then the other two devices. There was no manometer or PEEP valve used with the self-inflating bag and potentially useful interventions including 3 5 s prolonged inflations [4,6] were not evaluated. An in vitro comparison with bag and mask of the self-inflating bag, the flow-inflating bag, and the T-piece resuscitator evaluating the operators ability to consistently deliver targeted PIP, PEEP, prolonged inflations, and an efficient transition from a lower to higher PIP with each device is essential before initiating a clinical comparative trial during actual neonatal resuscitations. During our review of video taped resuscitations, we found that a colorimetric CO 2 detector placed between the mask and the ventilation device changed color when an airway was established. Although this doesn t confirm adequate ventilation, it is useful for the operator to know that there was a patent airway. There currently is no literature on the use of a colorimetric CO 2 detector during bag mask ventilation. We believe that the use of CO 2 detector can provide important information for the operator, and we wanted to evaluate whether the use of such a device would compromise the ability of operators to use any of the currently available bag and mask devices. 2. Methods/subjects We asked a variety of operators to participate in this study to reflect operators from different disciplines and years of experience who participate in neonatal resuscitation. The operators included neonatal respiratory therapists, neonatal nurses, neonatologists, neonatal fellows, neonatal nurse practitioners, and pediatric residents. The years of experience in neonatal resuscitation and job discipline were recorded for each participant prior to the start of the study for later comparison. The study was performed on a neonatal manikin (Laerdal Armonk, NY). When we tested the self-inflating device with the manikins any leak made it impossible to deliver a prolonged inflation or PEEP. Due to the variability of air leaks with different manikins in different head positions, we occluded the airway below the larynx so we could assess the operator s ability to create a seal over the airway and deliver desired pressures with each device. Three neonatal resuscitation devices were compared in this study. These included a T-piece Resuscitator (Neopuff, Fisher & Paykel Healthcare), a flow-inflating bag with a flow control tailpiece (Model E191 Anesthesia Associates, San Marcos, CA), and a self-inflating bag with a volume of 450 ml (Baby Blue II, Vital Signs, Totowa, NJ) with a set pressure relief of cmh 2 O. We chose the Baby Blue II self-inflating bag for this study because there is a manufacturer designed PEEP valve available. It was used both with and without the PEEP valve in place. The operator could adjust the PEEP valve to deliver the goal PEEP for the intervention. All devices were used with a round clear cushioned face-mask (Owens-BriGam Morganton, NC). Ten liters per minute of flow was used for both the T-piece resuscitator and the flow-inflating bag. Because the self-inflating bag does not need an air source none was used during this study. A pressure manometer was used for all interventions. For consistency, the same pressure manometer was used with all three devices connected via a T-adapter placed between the mask and bag device. The manometer was calibrated to a pressure transducer between participants. The pressures were recorded continuously via a pressure transducer (Celesco, Canoga Park, CA) and retained on a laptop computer (airway pressure monitor system and software were provided by Fisher & Paykel Healthcare). Each operator performed the following interventions: 1. Thirty seconds of manual breathes with a target PIP and PEEP of 20 and 5 cmh 2 O, respectively, and a breath rate of breathes per minute (bpm). 2. Thirty seconds of manual breathes with a target PIP and PEEP of 40 and 5 cmh 2 O, respectively, again with a rate of bpm. 3. A 30 s intervention in which the operator would start giving breaths at a PIP of 20 cmh 2 O and a PEEP of 5 cmh 2 O, respectively, and was then instructed to give a 5 s prolonged inflation with a target PIP of 20 cmh 2 O followed by a return to routine breaths at a PIP of 20 cmh 2 O and a PEEP of 5 cmh 2 O. After a brief period the operator was instructed to increase the PIP to 40 cmh 2 O. The transition time was recorded and a limit of 15 s was given to the operator to make this transition. After practicing each intervention with all three devices, each operator performed all of the interventions with every device in a random order. For the self-inflating bag the inter-

3 S. Bennett et al. / Resuscitation 67 (2005) ventions were also performed with and without the PEEP valve in place to determine if the use of this valve affected function. All participants were shown the pressure relief valve on the self-inflating bag and were comfortable with its use. They were instructed to adjust this valve as they saw fit to deliver the requested target pressures. The current practice in our nursery and resuscitation areas is to perform all resuscitations requiring bag mask ventilation with the addition of a colorimetric CO 2 detector (Pedicap Nellcor Pleasanton, CA) in place to confirm effective gas exchange. This device creates a 90 angle between the mask and the resuscitation device, which could potentially affect the ability to deliver target pressures and maintain an adequate seal. All operators were asked to repeat each intervention with the Pedicap in place for all three devices. This project did not involve human subjects and all participants gave their permission to use the data. Data analysis was performed using a computerized statistics package (SPSS Inc., Version #11, Chicago, IL). When comparing all bags, different operator disciplines and years of experience a one-way analysis of variance was used with Bonferroni post hoc comparison. The two-tailed t-test was used when comparing individual devices. p values less than 0.05 were considered statistically significant. 3. Results Thirty-one individuals participated in the study. There were 6 neonatologists, 5 neonatology fellows, 1 pediatrician, 2 pediatric residents, 1 neonatal nurse practitioner, 13 neonatal respiratory therapists, and 3 neonatal nurses. The years of experience varied from 1 year to greater than 20 years. All participants completed all aspects of the evaluation, which required approximately 40 min for each participant. The detailed results are presented in Table 1. At a target PIP of 20 cmh 2 O, there was a significant difference between the mean PIP of the T-piece resuscitator and the selfinflating bag (20.1 cmh 2 O, S.D. = 0.8 versus 21.0 cmh 2 O, S.D. = 1.7; p = 0.003). The mean PIP at a target of 40 cmh 2 O was significantly lower with the T-piece resuscitator compared to both the self-inflating bag and the flow-inflating bag (38.2 cmh 2 O, S.D. = 2.1 versus 39.6 cmh 2 O, S.D. = 2.7 versus 39.5 cmh 2 O, S.D. = 2.0; p < 0.006). The mean maximum PIP, the average maximum pressure delivered during an intervention for each participant, was analyzed to assess the peak pressures delivered by the operator with each of the devices. The maximum pressure was lower for the T-piece resuscitator compared to both the flowinflating bag and the self-inflating bag at a target of 20 cmh 2 O (20.7 cmh 2 O, S.D. = 0.8 versus 24.0 cmh 2 O, S.D. = 3.6 versus 24.7 cmh 2 O, S.D. = 2.8; p < 0.001). No operator delivered a breath with a pressure higher than 23 cmh 2 O using the T-piece resuscitator compared to 36.7% and 65% of operators using the flow-inflating bag and the self-inflating bag. Similarly, at a target PIP of 40 cmh 2 O, the maximum pressure for the T-piece resuscitator was significantly less than both the self-inflating bag and the flow-inflating bag (39.7 cmh 2 O, S.D. = 2.1 versus 45.3 cmh 2 O, S.D. = 4.7 versus 44.0 cmh 2 O, S.D. = 3.3; p < 0.001, Fig. 1). Only one operator delivered a breath greater than 41 cmh 2 O using the T-piece resuscitator, compared to 81.7% and 96.7% of operators using the flow-inflating bag and the self-inflating bag. The highest pressure delivered was 62 cmh 2 O with the selfinflating bag. Twenty of the 31 operators did not adjust the pressure relief valve when attempting to deliver pressures of 40 cmh 2 O. Transition from a PIP of 20 to 40 cmh 2 O took significantly longer using the T-piece resuscitator than both the flow-inflating and self-inflating bags (5.7 s, S.D. = 2.2 versus 1.8 s, S.D. = 0.8 versus 2.2 s, S.D. = 1.5; p < 0.001). Three participants could not make the transition using the T-piece resuscitator within the allotted 15 s. During the 5 s prolonged inflation, a pressure greater than 18 cmh 2 O was maintained for significantly longer with the T-piece resuscitator and the flow-inflating bag than the selfinflating bag (4 s, S.D. = 1.1 versus 3.7 s, S.D. = 1.7 versus 1.5 s, S.D. = 1.6; p < 0.001). Table 1 Results from the three neonatal resuscitation devices Intervention SIB (S.D.) FIB (S.D.) TPR (S.D.) Significance Mean PIP target = 20 cmh 2 O 21.0 * (1.7) 20.5 (1.5) 20.1 * (0.8) * p = Mean PIP target = 40 cmh 2 O 39.6 * (2.7) 39.5 ** (2.0) 38.2 *,** (2.1) * p = 0.002, ** p = Mean PEEP target = 5 cmh 2 O 3.6 *,** (1.9) 4.4 ** (1.2) 4.4 * (0.6) * p = 0.005, ** p = Maximum PIP (mean) target = 20 cmh 2 O 24.7 * (2.8) 24.0 ** (3.6) 20.7 *,** (0.8) * p < 0.001, ** p < Maximum PIP (mean) target = 40 cmh 2 O 45.3 * (4.7) 44 ** (3.3) 39.7 *,** (2.1) * p < 0.001, ** p < Highest PIP delivered target = 20 cmh 2 O N/A Highest PIP delivered target = 40 cmh 2 O a N/A Prolonged inflation >18 cmh 2 O (s) 1.5 *,** (1.6) 3.7 * (1.7) 4.0 ** (1.1) * p < 0.001, ** p < Transition time from 20 to 40 cmh 2 O (s) 2.2 * (1.5) 1.8 ** (0.8) 5.7 *,**,b (2.2) * p < 0.001, ** p < Breath rate target = bpm 51 * (18.8) 46.3 (12.5) 41.5 * (9.1) * p < SIB: self-inflating bag, FIB: flow-inflating bag, TPR: T-piece resuscitator, S.D.: standard deviation, bpm: breaths per minute. * and ** indicate the p value for the device comparison. a All but one of the operators delivered pressures less than 41 cmh 2 O. b Three individuals did not make the transition in the allotted 15 s.

4 116 S. Bennett et al. / Resuscitation 67 (2005) Fig. 1. Comparison of pressures between the three devices. SIB: selfinflating bag, FIB: flow-inflating bag, TPR: T-piece resuscitator. ( ): The maximum pressure ±2 standard deviations at a goal of 40 cmh 2 O. ( ): The maximum pressure ±2 standard deviations at a goal of 20 cmh 2 O. ( ): The average PEEP ±2 standard deviations at a goal of 5 cmh 2 O. The selfinflating bag was used with the manufacturer s PEEP valve. The self-inflating bag provides significantly less PEEP with the manufacturer s PEEP valve in place than both the T-piece resuscitator and the flow-inflating bag (3.6 cmh 2 O, S.D. = 1.9 versus 4.4 cmh 2 O, S.D. = 0.6 versus 4.4 cmh 2 O, S.D. = 1.2; p < 0.005). The self-inflating bag without the manufacturer s PEEP valve in place provided no PEEP (p < 0.001). No other significant differences were noted with the self-inflating bag in performing any other interventions with or without the PEEP valve in place. Although operators delivered breath rates within the target range of bpm with all the devices, a lower breath rate was seen with the T-piece resuscitator and the flow-inflating bag than the self-inflating bag (41.5 bpm versus 46.3 bpm versus 51 bpm; p < 0.001). 4. Discussion This study compares all available types of resuscitation devices and the operators ability to deliver PEEP and a prolonged inflation, two potentially beneficial resuscitation interventions. In our current prospective model, we chose to evaluate the ability of operators to create and maintain a seal with each of the devices and to deliver goal pressures. We chose pressures of 20 and 40 cmh 2 O to represent our PIP targets and 5 cmh 2 O for our PEEP target. As discussed by both NRP [2] and Milner [7], we felt 20 cmh 2 O is a commonly recommended pressure to ventilate normal neonatal lungs and therefore would be a good initial pressure for our in vitro resuscitation interventions. We chose 40 cmh 2 O as our high target pressure, as previous references have determined that such pressures may be required for adequate resuscitation in infants with diseased or immature lungs [2,7]. We did not measure tidal volumes in this study because in current clinical practice tidal volumes are not measured during resuscitation. We also chose to set inflation pressures as targets as opposed to chest rise and fall, improved heart rate and color because these measurements cannot adequately be recorded on a neonatal manikin and we wanted to evaluate the ability of operators to obtain and maintain adequate mask seals and pressures with each device prior to initiating a clinical trial. We used a manikin with a ligated airway as we noted excessive leaks on 4 of 5 tested manikins, which would not occur in an infant. These leaks occurred as air passed from the airway to the outside without inflating the manikin lung. In addition, our intention was to test the ability of resuscitators to achieve an adequate seal over the manikin s face and airway and to evaluate their ability to create an adequate airway-mask pressure. During actual resuscitation, the first goal when delivering bag and mask ventilation, is to establish an adequate pressure to ensure gas transfer to the lungs. While an adequate airway inflation pressure does not ensure adequate tidal ventilation, there can be no ventilation if there is an inadequate pressure. While our ligation would have decreased the needed delivered volumes, this did not obviate the need to establish a seal over the manikin airway. Our observations during actual resuscitations are that resuscitation teams do monitor actual airway pressures, and will begin to evaluate chest wall motion only after they have an adequate seal with measurable airway pressures. The T-piece resuscitator more consistently provides desired pressures, with smaller standard deviations in agreement with our previous observations, Fig. 1 [8]. Operators had more difficulty in increasing the PIP with the T-piece resuscitator, possibly accounting for the lower PIP achieved with this device. Even with the self-inflating bag s pressure-limiting valve in place, set between 35 and 45 cmh 2 O, pressures in excess of 45 cmh 2 O were delivered by 45% of the 20 operators who did not attempt to defeat the pressure relief valve. This failure of the pressure relief valve to prevent the delivery of pressures above the preset pressure limit has been previously noted by ourselves and others [9 11]. Prior to this study we were aware that a number of turns on the inspiratory pressure dial are required to increase the pressure on the Neopuff, and have noted that such changes are occasionally necessary during actual video recorded resuscitations. Many operators in such a situation may choose another device because of the inability to increase the Neopuff to the desired target pressure in a timely fashion, which requires almost three complete 360 turns of the pressure relief knob. Our observations are especially relevant as our staff is very familiar with this device, and the Neopuff is the most commonly used resuscitation device in our center. An initial prolonged inflation of up to 5 s duration has been shown to assist in the formation of a functional residual capacity (FRC) during delivery room resuscitations [6]. We teach the use of this technique for infants who do not respond to initial bag and mask ventilation. In the current study, we did not find a difference between the ability of the operators to use a T-piece resuscitator and the flowinflating bag to generate and maintain a prolonged inflation.

5 S. Bennett et al. / Resuscitation 67 (2005) However, even with practice the operators were unable to maintain a prolonged inflation with the self-inflating bag. This represents an inherent problem with the device design in that there is no airflow to maintain a prolonged breath, and any loss of airway seal will result in an immediate loss in pressure, which cannot be restored without another bag compression. It has been previously demonstrated that the flow-inflating bag is more difficult for inexperienced operators [12,13]. Many of our operators had significant resuscitation experience and each operator was able to practice each maneuver until they were comfortable before the study intervention was started. This may explain why we did not find any operator differences between these devices. CPAP is beneficial in maintaining FRC and improves gas exchange during cardiopulmonary resuscitation [14]. Early PEEP improves the response to surfactant, improves lung mechanics, increases surfactant pools, and reduces lung injury [15,16]. The use of PEEP and/or CPAP has been demonstrated to reduce the number of infants requiring intubation in the delivery room [17]. Early CPAP is used for premature infants with respiratory distress and PEEP is used for all neonates requiring mechanical ventilation, yet no recommendation exists from any current neonatal resuscitation guidelines for the use of CPAP or PEEP. A survey by Graham et al. demonstrated that 59% of neonatologists in the United States used CPAP or PEEP during neonatal resuscitation [18]. We felt that it was important to evaluate whether these three different device systems could effectively deliver CPAP/PEEP in a resuscitation situation. We used a manometer for all of our interventions. When using the self-inflating bag with the PEEP valve in place operators using the device were able to generate some PEEP although significantly less than both the flow-inflating bag and the T-piece resuscitator. This is important in that 83% of centers recently surveyed by O Donnell et al. use self-inflating bags during at least some of their neonatal resuscitations [3]. An inherent problem with the self-inflating bag generating PEEP is that once the mask seal is broken, PEEP cannot be delivered until another manual breath is given as there is no continuous air-flow. Both the T-piece resuscitator and the flow-inflating bag provide a continuous flow of gas to the airway, which is operator controlled, and the adjustment of this flow is used to maintain CPAP or PEEP. We currently teach our resuscitators that if there is no detectable carbon dioxide seen using a colormetric detector, during bag and mask ventilation that they need to reassess the airway. Because the use of such a detector has not been described during bag mask ventilation, we wanted to assess its use with each tested device. We were encouraged that the presence of the CO 2 detector did not affect the ability of the operators to use any of the tested devices significantly. Further research is required to confirm the potential benefit of these devices during neonatal resuscitation. Limitations to our study were the need to use a neonatal manikin so that the operator could not evaluate the response of the patient and adjust ventilation accordingly, and the lack of measurement of tidal volume, which in not considered part of current resuscitation practice. We chose to use a neonatal manikin to provide a more ideal model to assess ability to make a seal and deliver pressures prior to starting a clinical trial. 5. Conclusions Our results demonstrate that all current types of bag and mask devices for neonatal resuscitation can deliver target PIPs. Self-inflating bags cannot deliver a sustained prolonged inflation or provide adequate PEEP, even when equipped with a manufacturer s PEEP valve. Although the Neopuff delivers more consistent pressures with less variability, it requires more time to increase peak inspiratory pressures during resuscitation interventions. There is a need for improvement in the design and function of current manual resuscitation devices and for prospective trials to evaluate the optimal method of bag and mask ventilation for the newborn infant. Conflict of interest Dr. N Finer has received research support from Fisher & Paykel. Neither Fisher & Paykel nor any other manufacturer was involved in the design, conduct, analysis or writing of the manuscript for this study, which is the sole responsibility of the authors. Acknowledgement We gratefully acknowledge the provision of the pressure monitoring apparatus by Fisher & Paykel, Auckland, New Zealand. References [1] Singhal N, McMillan DD, Yee WH, Akierman AR, Yee YJ. Evaluation of the effectiveness of the standardized neonatal resuscitation program. J Perinatal 2001;21: [2] Kattwinkel J, editor. Textbook of neonatal resuscitation. 4th ed. American Academy of Pediatrics and American Heart Association; [3] O Donnell CPF, Davis PG, Morley CJ. Positive pressure ventilation at neonatal resuscitation: review of equipment and international survey of practice. Acta Paediatr 2004;93(5): [4] Ziderman DA, Bingham R, Beattie T, et al. Recommendations on resuscitation of babies at birth. Resuscitation 1998;37: [5] Hussey SD, Ryan CA, Murphy BP. Comparison of three manual ventilation devices using an intubated mannequin. Arch Dis Child Fetal Neonatal Ed 2004;89:F [6] Vyas H, Milner AD, Hopkin IE, Boon AW. Physiologic responses to prolonged and slow-rise inflation in the resuscitation of the asphyxiated newborn infant. J Pediatr 1981;99(4): [7] Milner AD. Resuscitation of the newborn. Arch Dis Child 1991;66:66 9.

6 118 S. Bennett et al. / Resuscitation 67 (2005) [8] Finer NN, Rich W, Craft A, Henderson C. Comparison of methods of bag and mask ventilation for neonatal resuscitation. Resuscitation 2001;49: [9] Kain ZN, Berde CB, Benjamin PK, Thompson JE. Performance of pediatric resuscitation bags assessed with and infant lung simulator. Anesth Analg 1993;77: [10] Kissoon N, Connors R, Tiffin N, Frewen TC. An evaluation of the physical and functional characteristics of resuscitators for use in pediatrics. Crit Care Med 1992;20(2): [11] Finer NN, Barrington KJ, Al-Fadley F, Peters KL. Limitations of self-inflating resuscitators. Pediatrics 1986;77(3): [12] Kanter RK. Evaluation of mask-bag ventilation in resuscitation of infants. AJDC 1987;141: [13] Mondolfi AA, Grenier BM, Thompson JE, Bachur R. Comparison of self-inflating bags with anesthesia bags for bag mask ventilation in the pediatric emergency department. Pediatr Emerg Care 1997;13(4): [14] Hevesi ZG, Thrush DN, Downs JB, Smith RA. Cardiopulmonary resuscitation effect of CPAP on gas exchange during chest compressions. Anesthesiology 1999;90(4): [15] Hartog A, Gommers D, Haitsma JJ, Lachmann B. Improvement of lung mechanics by exogenous surfactant: effect of prior application of high positive end-expiratory pressure. Br J Anaesth 2000;85(5): [16] Michna J, Jobe AH, Ikegami M. Positive end-expiratory pressure preserves surfactant function in preterm lambs. Am J Respir Crit Care Med 1999;160(2): [17] Gittermann MK, Fusch C, Gittermann AR, Regazzoni BM, Moessinger AC. Early nasal continuous positive airway pressure treatment reduces the need for intubation in very low birth weight infants. Eur J Pediatr 1997;156: [18] Graham AN, Finer NN. The use of continuous positive airway pressure and positive end expiratory pressure in the delivery room. Pediatr Res 2001;49:400A.