OBJECTIVES. The objectives of this study were to determine the impact of different volume-targeted levels on the work of breathing and to investigate whether a level that reduced the work of breathing below that experienced during ventilatory support without volume targeting could be determined.
METHODS. The transdiaphragmatic pressure-time product, as an estimate of the work of breathing, was measured for 20 infants (median gestational age: 28 weeks) who were being weaned from respiratory support by using patient-triggered ventilation (either assist-control ventilation or synchronous intermittent mandatory ventilation). The transdiaphragmatic pressure-time product was measured first without volume targeting (baseline) and then at volume-targeted levels of 4, 5, and 6 mL/kg, delivered in random order. After each volume-targeted level, the infants were returned to baseline. Each step was maintained for 20 minutes.
RESULTS. The mean transdiaphragmatic pressure-time product was higher with volume targeting at 4 mL/kg in comparison with baseline, regardless of the patient-triggered mode. The transdiaphragmatic pressure-time product was higher at a volume-targeted level of 4 mL/kg in comparison with 5 mL/kg and at 5 mL/kg in comparison with 6 mL/kg. The mean work of breathing was below that at baseline only at a volume-targeted level of 6 mL/kg.
CONCLUSIONS. Low volume-targeted levels increase the work of breathing during volume-targeted ventilation. Our results suggest that, during weaning, a volume-targeted level of 6 mL/kg, rather than a lower level, could be used to avoid an increase in the work of breathing.
During volume-targeted ventilation (VTV), a nearly constant tidal volume is delivered. This minimizes excessive tidal volume delivery, reducing volutrauma and the likelihood of hypocarbia. In one study, use of VTV was associated with avoidance of severe hypocarbia or hypercarbia in >90% of the first 48 hours for infants at <33 weeks of gestation.1 VTV may be useful in situations in which there are changes in lung function, such as when infants are recovering from respiratory distress and being weaned from a ventilator. During VTV, compared with pressure-limited ventilation, blood gas values were maintained at lower airway pressures.2 Those results suggested that infants make a greater contribution to minute ventilation during VTV than during pressure-limited ventilation and hence their work of breathing (WOB) may be higher with VTV, which may affect weaning and extubation adversely.3 It seems likely, however, that higher volume-targeted (VT) levels would reduce the WOB and, with a sufficiently large targeted volume, the WOB would be lower during ventilatory support with versus without volume targeting. These hypotheses have not been tested, however, and a variety of VT levels have been used in VTV studies.4–8 The aims of this study were to determine the impact of different VT levels on the WOB and to investigate whether a level that reduced the WOB below that experienced during ventilatory support without volume targeting could be determined.
Study Group and Measurements
Prematurely born infants <1 week of age who were being weaned from mechanical ventilation by using a triggered mode (synchronous intermittent mandatory ventilation [SIMV] or assist-control ventilation [ACV]) and were thought to be likely to be extubated within the next 24 hours were eligible for entry into the study. Infants were entered into the study if their parents gave written informed consent. The study was approved by the King's College Hospital research ethics committee.
The infants were supported with SLE 5000 ventilators (software versions 3.2 and 4.1; SLE Ltd, South Croydon, UK). During VTV with a SLE 5000 ventilator, the maximal (set) peak inspiratory pressure was delivered to the infant only if the VT level was not achieved. In addition, inflation was terminated once the VT level was achieved, which meant that the delivered inflation time might be shorter than the preset inflation time. In this study, if the delivered inflation time was <0.2 seconds, then the waveform was altered to give a shallower upstroke to the inflating pressure, prolonging the inflation time.
The WOB was assessed through measurement of the transdiaphragmatic pressure-time product (PTPdi). Measurements were made first in the triggered mode without volume targeting (baseline) and then at VT levels of 4, 5, and 6 mL/kg. The order in which the infants received different VT levels was randomized. After each period of volume targeting, the infants were returned to baseline. Each step was maintained for 20 minutes. The PTPdi was recorded over the last 5 minutes at each step. The baseline PTPdi was calculated by averaging the PTPdi results for the 4 baseline periods.
For measurement of PTPdi, esophageal and gastric pressures were measured with a dual-pressure tipped transducer (Gaeltec, Dunvegan, Scotland), and the signals were amplified. The 2 pressure transducers were 5 cm apart, with the lower transducer 0.3 cm from the catheter tip. Airflow was measured with a pneumotachograph (Mercury F10L; GM Instruments, Kilwinning, Scotland) connected to a differential pressure transducer (±2 cm H2O; MP45; Validyne, Northridge, CA) and inserted between the endotracheal tube and the ventilator manifold. Tidal volume was assessed through digital integration of the flow signal. Mouth pressure was measured through a side port on the pneumotachograph with a differential pressure transducer (±100 cm H2O; MP45; Validyne). The signals from the pressure transducers were amplified by using a carrier amplifier (CD280; Validyne). The pressure and flow signals were recorded and displayed in real time on a laptop computer (Satellite 230 CX; Toshiba, Tokyo, Japan) running an application written with Labview software (National Instruments, Austin, TX), with 100-Hz analog-to-digital sampling (DAQ 16XE-50; National Instruments). Transdiaphragmatic pressure was calculated through digital subtraction of the esophageal pressure from the gastric pressure by the data acquisition software. The PTPdi was calculated after acquisition from the transdiaphragmatic pressure signal integrated against time for each breath and expressed per minute. The PTPdi was calculated from 20 consecutive breaths.
Differences in the demographic features of the infants with the 2 triggered modes were assessed for statistical significance by using the Mann-Whitney U test. The PTPdi data were shown to be normally distributed by using the Kolmogorov-Smirnov test, and differences were assessed for statistical significance by using repeated-measures one-way analysis of variance. Correction was made for planned multiple comparisons by using Bonferroni's test. SPSS 16 (SPSS, Chicago, IL) was used.
Recruitment of 20 infants provided the ability to detect, with 90% power at the 5% level, a 1-SD difference in the PTPdi results.
Twenty infants were studied at a median postnatal age of 3.5 days (range: 1–8 days). The infants were supported with ACV or SIMV delivered by SLE 5000 ventilators via shouldered endotracheal tubes; there is minimal or no leakage around such tubes.9 Ten infants were supported with SIMV (median rate: 35 breaths per minute [range: 20–45 breaths per minute]) and did not differ significantly, with respect to their gestational age (P = .25) or birth weight (P = .14), from the 10 infants supported with ACV (Table 1). The infants supported with SIMV, however, were studied at a significantly lower postnatal age than were those supported with ACV (median: 4 days [range: 1–6 days] vs 6 days [range: 3–8 days]; P = .017). There were no other statistically significant differences in the demographic features of the SIMV and ACV groups (Table 1). All of the infants followed the standard weaning policy of the neonatal unit, that is, infants were weaned from mechanical ventilation by being transferred to patient-triggered ventilation, and sedation was then stopped. Peak pressure was gradually reduced according to the results of blood gas monitoring. During SIMV, the ventilator rate also was reduced, but never below 20 breaths per minute. The back-up rate during ACV and the SIMV rate remained constant throughout the study period. All of the study infants were receiving caffeine, which had been administered in a loading dose of 20 mg/kg once the peak pressure was reduced below 20 cm H2O.
Overall, the differences in PTPdi according to VT level were statistically significant (F = 25.75; P < .001). The mean PTPdi was significantly higher at a VT level of 4 mL/kg, compared with baseline (P < .001) (Table 2). A VT level of 4 mL/kg was associated with a higher mean PTPdi, compared with 5 mL/kg (P = .003) and 6 mL/kg (P < .001), and a VT level of 5 mL/kg was associated with a higher mean PTPdi, compared with 6 mL/kg (P = .003) (Table 2). The mean WOB at a VT level of 5 mL/kg tended to be higher than that at baseline (P = .08); the mean WOB at a VT level of 6 mL/kg was lower than that at baseline, but this was not statistically significant (P = .3) (Table 2). For the infants supported with ACV, the mean PTPdi at a VT level of 4 mL/kg was significantly higher than that at baseline (P = .01) and that at 6 mL/kg (P = .005) (Fig 1). For the infants supported with SIMV, the mean PTPdi at a VT level of 4 mL/kg was significantly higher than that at baseline (P = .038) and that at 6 mL/kg (P = .005) (Fig 2). Comparison of the PTPdi values for the infants supported with SIMV and those supported with ACV revealed no significant differences at baseline or at any VT level.
PTPdi values did not significantly increase progressively throughout the duration of the study. There were no significant differences in inflation time at the 3 VT levels. Overall, the mean expired tidal volume (Vte) differed significantly between the 3 VT levels (F = 8.8; P = .02), and the Vte at a VT level of 4 mL/kg was significantly lower than that at a VT level of 6 mL/kg (P = .025). The Vte at all 3 levels exceeded the set VT level (Table 3). Overall, the infant's respiratory rate decreased with increasing VT levels (F = 5.5; P = .014). As expected, the mean peak inspiratory pressure increased as the VT level was increased (F = 5.96; P = .009), but there were no significant differences in the total minute volumes at the different VT levels (F = 1.14; P = .36) (Table 3).
We demonstrated that the WOB, as assessed with PTPdi measurements, was higher with a VT level of 4 mL/kg, compared with triggered ventilation without volume targeting. In addition, we determined that the mean WOB was below that for triggered ventilation without volume targeting only when a VT level of 6 mL/kg was used. If we had used levels of >6 mL/kg, then the WOB might have been significantly lower than that experienced with triggered ventilation without volume targeting. We did not study larger volumes because they are more likely to cause endothelial and peripheral airway injury10 and to result in hypocarbia, which increases the likelihood of periventricular leukomalacia in low birth weight infants.11 The mean Vte at each VT level, however, exceeded the set value, which suggests that all of the VT settings we used were too low, with respect to at least some of the infants' physiologic requirements.
We examined VT levels of 4, 5, and 6 mL/kg because these values are within the tidal volume range for spontaneously breathing, prematurely born infants. In addition, although VT levels between 3 and 8 mL/kg were used in previous studies of VTV, VT levels ranged from 4 to 6 mL/kg in the majority of studies.4–8 Within that range, we determined that the mean PTPdi value was greater at lower VT levels, compared with higher levels. Other evidence also suggests that higher rather than lower VT levels should be used. Increased levels of proinflammatory cytokines and a longer duration of ventilation were demonstrated for infants ventilated during the acute phase of respiratory distress syndrome when a tidal volume of 3 mL/kg, rather than 5 mL/kg, was used.12 In addition, a reduction in the duration of hypoxemic episodes during SIMV was demonstrated when a targeted volume of 6 mL/kg, but not 4.5 mL/kg, was used.4
Addition of volume targeting at 3.0 mL/kg, rather than 4.5 mL/kg, to SIMV was associated with a greater spontaneous minute volume in one series,13 which suggests a greater WOB at the lower VT level. We demonstrated similar total (infant and ventilator) minute volumes at the 3 VT levels. There were no significant differences in the inflation times at the 3 VT levels but, not surprisingly, the Vte values did differ significantly, being lowest at a VT level of 4 mL/kg. The reduction in Vte as the VT level was reduced was associated with a significant increase in the infants' spontaneous respiratory rates (Table 3), which likely explains why the minute volumes were similar at the 3 VT levels but the WOB was significantly greater at the lower VT levels. During weaning, gradual transfer of the WOB from the ventilator to the infant must occur, but clearly the WOB must not be excessive. We demonstrated that the PTPdi was 50% higher at a VT level of 4 mL/kg, compared with 6 mL/kg. For infants with a high oxygen cost of breathing, oxygen consumption (which is an indicator of the WOB) was ∼30% lower with ACV, compared with SIMV (14 breaths per minute),14 and weaning duration was prolonged when slow-rate SIMV rather than ACV was used,15 which suggests that even a 30% increase in oxygen consumption is adverse. In another study, a 30% difference in the WOB was noted for prematurely born infants for whom extubation failed or succeeded.3 Therefore, our results demonstrating a 50% increase in the WOB at a VT level of 4 mL/kg, compared with 6 mL/kg, suggest that the former is disadvantageous.
The WOB was assessed through measurement of PTPdi. PTPdi reflects the energy expenditure of the diaphragm, in that it assesses expenditure during isometric and nonisometric contractions, unlike classical pressure-volume calculations. It also allows assessment of efforts that do not result in triggered ventilator breaths. Infants in this study underwent ventilation with the SLE 5000, which delivers targeted tidal volumes or volume limited ventilation. We demonstrated previously that the airway pressure waveform during VTV differs according to the type of ventilator used.16 During VTV with the SLE 5000, inflation is terminated short of the preset inflation time if the volume has already been delivered; this also occurs with the VIP Bird (VIP Bird, Palm Springs, CA), but that ventilator has a slower upstroke.16 In contrast, the airway pressure waveform with the Draeger Babylog 8000 (Draeger Lubeck, Germany) has a positive pressure plateau, and that with the Stephanie (F Stephan GmbH, Medizintechnik, Kirchstrasse, Germany) has a slow upstroke, with the peak pressure increasing until the end of the preset inflation time.16 The results of the comparison of triggered ventilation with and without volume targeting might have been different if we had used a different ventilator type. We used the same ventilator type to compare the different VT levels however, and ensured a minimal ventilator inflation time, and demonstrated that lower VT levels, compared with higher levels, were associated with greater PTPdi, which indicates a greater WOB. Although the patient's WOB is influenced by the gas flow,17 this did not bias our results, because the SLE 5000 ventilator delivers a constant flow with or without volume targeting.
Infants who were being weaned from mechanical ventilation by using triggered ventilation were assessed because this is likely to be a situation in which VTV would be useful, that is, during the recovery phase, when lung function is changing. The studied infants undergoing ACV and SIMV were of similar gestational ages and birth weights, but those undergoing SIMV were studied at a slightly but significantly lower postnatal age. Similar results were obtained for the 2 triggered modes (Table 1), that is, the mean PTPdi was significantly greater when volume targeting was added at 4 mL/kg, compared with baseline, and when lower VT levels, compared with higher levels, were used. We did not demonstrate any significant differences in PTPdi values between the infants supported with SIMV and those supported with ACV, either at baseline or at any VT level. The minimal SIMV rate used was 20 breaths per minute, however, and it is possible that the difference in PTPdi might have been greater if we had used a lower SIMV rate, because reduction of the SIMV rate below 20 breaths per minute prolongs the duration of weaning.15 In addition, we emphasize that we did not power our study to detect significant differences between the groups supported with SIMV versus ACV. We had hypothesized that the PTPdi would be greater with lower rather than higher VT levels. Therefore, to avoid any bias from a cumulative effect of an increased WOB, we randomized the order in which the VT levels were used, and we did not observe any cumulative increase in WOB with time. In addition, the baseline PTPdi was calculated from the mean of the 4 periods at baseline settings.
We demonstrated that the WOB was significantly greater with VTV with a VT level of 4 mL/kg, compared with ventilation without volume targeting, and only when the VT level was increased to 6 mL/kg was the mean WOB below that found with triggered ventilation without volume targeting. For prematurely born infants, excessive WOB may predispose them to fatigue18 and adversely affect their ability to trigger the ventilator, which may prolong the duration of weaning. It has been demonstrated that infants with a high WOB, compared with a lower WOB, are more likely to fail extubation,3 and increased WOB also may contribute to a failure to grow.18 Our results suggest that, if VTV is used during weaning, then the VT level should be 6 mL/kg rather than a lower level, to avoid an increase in the WOB. Although general guidelines can be developed from this work, individualized care, with assessment of each infant's particular need on the basis of respiratory rate and clinical indications of the WOB, is important. Our study indicated that the WOB depends on the target VT setting, and it should increase awareness of the drawbacks of excessively low VT levels; however, it does not define the optimal weaning strategy with VTV, which could be achieved only with a prospective comparison of different VT levels, with the duration of ventilation as the primary end point.
Dr Patel was supported by the Charles Wolfson Charitable Trust. Dr Sharma was supported by the WellChild Trust.
- Accepted December 15, 2008.
- Address correspondence to Anne Greenough, MD, Newborn Unit, Fourth Floor, Golden Jubilee Wing, King's College Hospital, Denmark Hill, London, SE5 9RS, United Kingdom. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
What's Known on This Subject
During VTV, a nearly constant tidal volume is delivered. A variety of VT levels have been used in VTV.
What This Study Adds
Addition of low VT levels to patient-triggered modes increases the WOB during weaning. This can be avoided by using a VT level of 6 mL/kg.
- ↵Cheema IU, Ahluwalia JS. Feasibility of tidal volume-guided ventilation in newborn infants: a randomized, crossover trial using the volume guarantee modality. Pediatrics.2001;107 (6):1323– 1328
- ↵Herrera CM, Gerhardt T, Claure N, et al. Effects of volume-guaranteed synchronized intermittent mandatory ventilation in preterm infants recovering from respiratory failure. Pediatrics.2002;110 (3):529– 533
- ↵Rozé JC, Liet JM, Gournay V, Debillon T, Gaultier C. Oxygen cost of breathing and weaning process in newborn infants. Eur Respir J.1997;10 (11):2583– 2585
- ↵Dimitriou G, Greenough A, Griffin F, Chan V. Synchronous intermittent mandatory ventilation modes compared with patient triggered ventilation during weaning. Arch Dis Child Fetal Neonatal Ed.1995;72 (3):F188– FF190
- ↵Kallet RH, Campbell AR, Dicker RA, Katz JA, Mackersie RC. Work of breathing during lung-protective ventilation in patients with acute lung injury and acute respiratory distress syndrome: a comparison between volume and pressure-regulated breathing modes. Respir Care.2005;50 (12):1623– 1631
- ↵Guslits BG, Gaston SE, Bryan MH, England SJ, Bryan AC. Diaphragmatic work of breathing in premature human infants. J Appl Physiol.1987;62 (4):1410– 1415
- Copyright © 2009 by the American Academy of Pediatrics