Randomized, Controlled Trial Comparing Synchronized Intermittent Mandatory Ventilation and Synchronized Intermittent Mandatory Ventilation Plus Pressure Support in Preterm Infants
BACKGROUND. Prolonged mechanical ventilation is associated with lung injury in preterm infants. In these infants, weaning from synchronized intermittent mandatory ventilation may be delayed by their inability to cope with increased respiratory loads. The addition of pressure support to synchronized intermittent mandatory ventilation can offset these loads and may facilitate weaning.
OBJECTIVE. The purpose of this work was to compare synchronized intermittent mandatory ventilation and synchronized intermittent mandatory ventilation plus pressure support in weaning from mechanical ventilation and the duration of supplemental oxygen dependency in preterm infants with respiratory failure.
METHODS. Preterm infants weighing 500 to 1000 g at birth who required mechanical ventilation during the first postnatal week were randomly assigned to synchronized intermittent mandatory ventilation or synchronized intermittent mandatory ventilation plus pressure support. In both groups, weaning followed a set protocol during the first 28 days. Outcomes were assessed during the first 28 days and until discharge or death.
RESULTS. There were 107 infants enrolled (53 synchronized intermittent mandatory ventilation plus pressure support and 54 synchronized intermittent mandatory ventilation). Demographic and perinatal data, mortality, and morbidity did not differ between groups. During the first 28 days, infants in the synchronized intermittent mandatory ventilation plus pressure support group reached minimal ventilator settings and were extubated earlier than infants in the synchronized intermittent mandatory ventilation group. Total duration of mechanical ventilation, duration of oxygen dependency, and oxygen need at 36 weeks' postmenstrual age alone or combined with death did not differ between groups. However, infants in synchronized intermittent mandatory ventilation plus pressure support within the 700- to 1000-g birth weight strata had a shorter oxygen dependency.
CONCLUSIONS. The results of this study suggest that the addition of pressure support as a supplement to synchronized intermittent mandatory ventilation during the first 28 days may play a role in reducing the duration of mechanical ventilation in extremely low birth-weight infants, and it may lead to a reduced oxygen dependency in the 700- to 1000-g birth weight strata.
- mechanical ventilation
- synchronized intermittent mandatory ventilation
- pressure support
- preterm infant
- supplemental oxygen
Despite the emergence of new modes of mechanical ventilation, the optimal ventilatory strategy that can provide adequate ventilation and minimize the risk of lung injury in premature infants with respiratory failure has not been identified. Synchronized intermittent mandatory ventilation (SIMV) is one of the most widely used modalities of respiratory support and has been found to reduce the duration of mechanical ventilation compared with conventional intermittent mandatory ventilation.1,2 SIMV weaning in preterm infants is achieved by a gradual reduction of the SIMV rate and peak inspiratory pressure (PIP) while the infants increase their spontaneous ventilation. This weaning process can be slowed by the inability of these infants to cope with the increased work of breathing because of the resistance of small endotracheal tubes, elevations in lung elastance and airway resistance induced by disease, increased dead space, and poor chest wall stability.3–10
Pressure support (PS) assists every spontaneous breath by providing an inspiratory pressure boost aimed at reducing work of breathing.11–13 PS has been shown to augment spontaneous ventilation in neonates with congenital heart disease14 and decrease the work of breathing in children during cardiac surgery.15
The short-term effects of combining SIMV and PS (SIMV+PS) in preterm infants were investigated previously.16 The combined SIMV+PS maintained adequate ventilation and gas exchange while simultaneously preventing an increase in inspiratory effort during an acute reduction in SIMV rate. These findings support the use of PS as a supplement to SIMV to facilitate weaning from ventilatory support. This may be important in preterm infants in whom accelerated weaning and an exposure to lower SIMV rates may reduce the risk of lung injury. Therefore, it was hypothesized that SIMV+PS, compared with SIMV alone, will accelerate weaning from mechanical ventilation and minimize lung injury as indicated by a reduced oxygen dependency. The objective of the present study was to compare SIMV+PS to SIMV alone in a group of preterm infants with regard to ventilator weaning and need for supplemental oxygen.
The study was conducted in the Newborn Intensive Care Unit of the University of Miami-Jackson Memorial Medical Center between September 2002 and November 2004. The study was reviewed and approved by the University of Miami Institutional Review Board for the Protection of Human Subjects in Research.
All of the preterm infants with birth weights (BWs) between 500 and 1000 g who were appropriate for gestational age and required mechanical ventilation for >12 hours from the first day after birth until day 7 were considered eligible for the study. The exclusion criteria included the presence of congenital anomalies, neuromuscular disease, lung hypoplasia, congenital heart disease, and hypotension requiring intravenous medications. Infants were also excluded if they had pulmonary interstitial emphysema (PIE) or pneumothorax, if they required a previous course of >24 hours of high-frequency oscillatory ventilation (HFOV), or if they received sedation or muscle relaxation. Infants were enrolled after written informed consent was obtained from the parents or legal guardians.
Design and Randomization
It was estimated that 96 neonates surviving to day 28 needed to be enrolled to show a 30% difference in the duration of oxygen dependency between groups at an α of .05 with a power of 90%. After adjustment for center mortality rate in this population of preterm infants, it was estimated that enrollment of 108 infants was necessary.
Block randomization within 2 BW strata (500–699 and 700–1000 g) was conducted to assign eligible infants to 1 of 2 modes of mechanical ventilation, SIMV or SIMV+PS, using sequential sealed opaque envelopes from a computer-generated randomization list. Caregivers were not blinded to the assigned modality. Stratification was decided based on the differences in incidence of bronchopulmonary dysplasia and duration of mechanical ventilation at different BWs.
Mechanical ventilation for all of the infants was provided using time-cycled, pressure-limited, flow-triggered neonatal ventilators (VIP Gold, Viasys Healthcare, Yorba Linda, CA). On randomization, ventilator settings for infants assigned to the SIMV+PS group were initially modified as follows: SIMV rate was decreased by 10 breaths per minute, and PS was added at 30% to 50% of the difference between PIP and positive end-expiratory pressure (PEEP) of the SIMV breaths. The initial decrease in rate in the SIMV+PS group was used to compensate for the added PS to minimize the increase in mean airway pressure and the risk of hyperventilation. Infants randomly assigned to the SIMV group remained on settings as determined by the clinical team.
Subsequently, mechanical support was managed in both groups during the first 28 postnatal days according to a protocol accepted previously by consensus of the center clinical staff as follows. The fraction of inspired oxygen (Fio2) was adjusted to keep the arterial oxygen saturation by pulse oximetry (SPo2) between 88% and 95%, PIP was decreased to a minimum needed to provide a tidal volume (Vt) between 3 and 6 mL/kg, PEEP was decreased to 4 to 5 cm H2O, SIMV rate was adjusted to keep the arterial carbon dioxide tension (Paco2) between 40 and 65 mmHg and pH ≥7.25. In the SIMV+PS group, the level of PS was started at 50% of the PIP-PEEP difference and thereafter maintained at or decreased gradually below this level to a minimum of 30% as tolerated. Surfactant was administered following the center guidelines in both groups of infants as needed.
The decision to remove from mechanical ventilation was left to the discretion of the clinical team. All of the infants received caffeine before extubation and were maintained on nasal continuous positive airway pressure for ≥48 hours after extubation. Extubated infants who showed signs of respiratory deterioration including an increased oxygen requirement with Fio2 > 0.5 to maintain SPo2 ≥ 88%, or a Paco2 ≥ 65 mmHg and pH ≤ 7.25, or severe apneic episodes requiring bag-mask ventilation were returned to mechanical ventilation. Ventilatory support was resumed using the originally assigned mode, and ventilator management followed the above-mentioned protocol. Infants remained in the assigned mode until day 28. Thereafter, ventilatory management was left to the discretion of the clinical team, which was allowed to use either mode of ventilation. HFOV was used as rescue intervention only for infants who had Paco2 >70 mmHg on 2 consecutive arterial blood gases despite an SIMV rate of 60 breaths per min, SPo2 < 88% on 1.0 Fio2, or for infants who developed PIE. General clinical management, as well as nursing care and monitoring were done according to the center routine protocol throughout the infant's hospitalization.
Data Collection, Outcomes, and Data Analysis
Daily ventilatory parameters of PIP, PEEP, PS, SIMV rate, mean airway pressure, total respiratory rate, Vt of SIMV breaths, SPo2, and Fio2 were documented. Data documentation included arterial or capillary blood gases obtained as per unit protocol or as dictated by the clinical condition. Infants were followed until hospital discharge or death.
Respiratory outcomes were assessed as those occurring from birth until day 28 while the study protocol was actively followed and those occurring from birth until discharge or death. Respiratory outcome parameters included: time to reach a level of minimal ventilator settings consisting of PIP ≤16 cm H2O, Fio2 ≤0.30, PEEP ≤5 cm H2O, and SIMV rate ≤15 breaths per minute and remained on these settings for ≥48 hours (the clinical teams were not aware of the existence of this criterion or when an infant had met such criterion); time to first successful extubation (planned or inadvertent) defined as ≥48 hours without mechanical ventilation; age of final removal from mechanical ventilation; total cumulative duration of mechanical ventilation; ventilator requirement at day 28; total cumulative duration of supplemental oxygen; and oxygen requirement at day 28 and at 36 weeks' postmenstrual age (PMA). A ventilator day was defined as any daily period of >12 hours on conventional mechanical ventilation or HFOV. A supplemental oxygen day was defined as a daily requirement of Fio2 >0.21 for >12 hours. In addition, infants were monitored for the occurrence of respiratory complications including pneumothorax and PIE. Morbidity assessment included intraventricular hemorrhage (IVH), patent ductus arteriosus (PDA), sepsis, necrotizing enterocolitis (NEC), and retinopathy of prematurity (ROP).
Data were analyzed on an intention-to-treat basis. Between-group comparisons of continuous variables were conducted using Student's t test or the nonparametric Mann-Whitney rank-sum test for nonnormally distributed data. Categorical variables were compared by χ2 test or Fisher's exact test as appropriate. Kaplan-Meier log-rank survival analysis was used to compare between-group data involving the time required to reach the above-mentioned milestones. An overall significance level of 0.05 was used for all of the statistical comparisons.
A total of 265 infants weighing between 500 and 1000 g at birth were admitted to the newborn intensive care unit during the trial. Of these, 125 infants were eligible for the study. A total of 107 infants were enrolled; 54 were randomly assigned to the SIMV group and 53 to SIMV+PS. Eighteen eligible infants were not enrolled: 16 for parental refusal and 2 for inability to get consent on time. A total of 140 infants were excluded from the study. Of these, 60 did not require mechanical ventilation during the screening period, 13 infants had congenital anomalies, 26 were intrauterine growth retarded, 9 had a recent history of pneumothorax or PIE, and 32 infants were hemodynamically unstable. Among the 140 infants excluded, there were 5 infants for whom consent was obtained but were recognized as ineligible shortly after randomization: 1 was diagnosed with PIE, 1 was inadvertently extubated, 1 developed intractable hypotension and died, and 2 infants were diagnosed with large ventricular septal defect with left heart overload. These 5 enrolled infants were randomly assigned, although they had met exclusion criteria during the interval between enrollment and randomization. They were not included in the analysis. The diagram of Fig 1 illustrates the flow of participants from screening to study completion.
Population characteristics including BW, gestational age, gender, exposure to antenatal steroids and maternal chorioamnionitis, and surfactant treatment were not different between the SIMV and SIMV+PS groups. At the time of random assignment, infants in both groups had comparable ages and were on similar ventilator settings (Table 1). Surfactant was administered therapeutically. Only 2 infants (1 per group) received surfactant after random assignment.
Assessment of respiratory parameters during the first 28 days after birth while the infants were managed according to the study protocol indicated that infants in the SIMV+PS group reached the criteria of minimal ventilator settings earlier (Fig 2) and were permanently removed from mechanical ventilation sooner (Fig 3) than infants in the SIMV group.
When comparing the SIMV rates used to support infants during the first 28 days, the SIMV rates for infants in SIMV+PS were significantly lower than those for the SIMV group on days 4, 7, and 10 (Fig 4), whereas Paco2 levels did not differ between groups. Median PIP and Vt of SIMV breaths during the first 28 days did not differ between groups. PIP of the SIMV breaths were maintained within 15 and 18 cmH2O and Vt within 4.6 and 7.0 mL/Kg (25th and 75th percentile) in both groups.
By design, ventilator rate was reduced immediately after random assignment by 10 breaths per minute in all of the infants in the SIMV+PS group. Weaning also occurred in the SIMV group, where 44% of the infants in the SIMV group had a reduction in rate right after random assignment, and 55% were on a lower rate 12 hours later. Not every infant in the SIMV+PS group tolerated the initial weaning, and 24% of them were with a higher ventilator rate 12 hours after random assignment.
Table 2 shows respiratory outcomes assessed during the first 28 days after birth. The proportion of infants requiring mechanical ventilation at day 28 was lower in the SIMV+PS group (47% vs 69%; P < .04). The duration of mechanical ventilation and supplemental oxygen during the first 28 days and oxygen need at day 28 did not differ significantly between groups. Respiratory outcomes assessed during the first 28 days within each BW strata did not differ significantly between groups.
Table 3 shows the respiratory outcomes assessed from birth until discharge or death. The age when infants were permanently removed from mechanical ventilation and the total duration of mechanical ventilation did not differ significantly between groups. Although infants in SIMV+PS had a shorter median duration of supplemental oxygen, and a smaller proportion of them required oxygen at 36 weeks' PMA, these differences were not statistically significant. Similarly, the combined outcome of death or oxygen at 36 weeks' PMA was not different. Most outcomes assessed from birth to discharge or death within each BW strata did not differ between groups. However, the SIMV+PS group had a shorter duration of oxygen dependency compared with the SIMV group within the 700- to 1000-g BW strata.
As the study protocol was followed until day 28, the clinical team subsequently determined the ventilatory modality. Among infants alive at day 28, 19 (39%) in the SIMV group and 12 (24%) in the SIMV+PS group were subsequently switched to the other modality. After day 28, infants required mechanical ventilation for a median (25th–75th percentile) duration of 26 (15–44) and 13 (5–51) days in the SIMV and SIMV+PS group, respectively. The median (25th–75th percentile) ventilator rate at extubation was 15 (15–15) in the SIMV group compared with 10 (10–15) breaths per minute in the SIMV+PS group (P < .001).
Overall mortality and death before day 28 did not differ between groups. No infant developed pneumothorax, and the proportion of infants who developed PIE was not different between groups. HFOV was used as a rescue intervention in 23% of all infants. This was mainly indicated for persistent hypercapnia. The proportion of infants requiring HFOV did not differ between groups. The rates of other morbidities including sepsis, PDA, IVH grade III or higher, NEC, and ROP stage III or greater were not different between groups (Table 4).
The rate of endotracheal colonization was 76% and 81% in the SIMV and SIMV+PS groups, respectively. Systemic infections were diagnosed in 66% and 67% of the infants in the SIMV and SIMV+PS groups, respectively. Postnatal steroids were given to 11% and 6% of the infants in the SIMV and SIMV+PS groups, respectively.
The results of this randomized, controlled trial indicate that the use of PS in addition to SIMV during the first 28 days allowed for an earlier reduction in ventilatory support and a shorter ventilator dependency compared with SIMV alone in a group of extremely low BW infants. This combined strategy also allowed maintenance of a lower SIMV frequency compared with SIMV alone during the first 10 days.
In a previous pilot study, a similar range of PS as the one used in the present trial was shown to unload the respiratory pump and maintain spontaneous Vt within a reference range during an acute reduction in SIMV rate in preterm infants with different stages of lung disease.16 In this study, PS was purposely maintained at a fraction of the SIMV pressure amplitude. This was done to avert exposure to higher peak pressures in the range of those of the SIMV breaths. Thus, in this trial, the combined use of SIMV and PS was not only intended to accelerate weaning but also to provide a more lung-protective ventilatory support by reducing exposure to the larger SIMV breaths.
The mechanisms by which PS can facilitate weaning are multiple. In addition to facilitating maintenance of a consistent spontaneous Vt and better alveolar ventilation than during SIMV alone without the need for an increase in breathing effort, PS may attenuate the thoracoabdominal asynchrony often observed in preterm infants. Because spontaneous minute ventilation depends on proper diaphragmatic function, PS may also avert diaphragm overexertion and fatigue because of the increased work of breathing that occurs during SIMV weaning. Also, a more consistent spontaneous breathing and reduced reliance on SIMV breaths can preserve diaphragm conditioning and avert diaphragmatic dysfunction as observed during prolonged ventilator dependency.17
With the underlying assumption that the combined SIMV+PS would result in a more lung-protective ventilatory support, it was expected that the potential beneficial effects would be reflected in a reduced oxygen dependency. Although the duration of oxygen dependency and oxygen need at 36 weeks' PMA were not lower in the SIMV+PS group, there was a significant reduction in oxygen dependency in SIMV+PS infants within the 700- to 1000-g BW strata. The observed benefit of SIMV+PS in the larger infants may have resulted from a more consistent respiratory drive and diaphragmatic function compared with the smaller and more immature infants. This may have allowed PS, a patient-triggered modality, to function more effectively in assisting spontaneous inspiration.
In a randomized, controlled trial that compared volume-targeted to pressure-limited ventilation in preterm infants,18 PS was used in the weaning phase, which may have contributed to the faster weaning. Although those infants were more mature than the infants in the present study, those findings are in agreement with the present results.
Although PS can be used independently of SIMV, in this study the use of the combined SIMV+PS strategy was intended to provide a background level of minimum ventilation. This may be particularly important when used in preterm infants with unstable respiratory drive or insufficient breathing effort to trigger the ventilator adequately to maintain ventilation and gas exchange. Newer infant ventilators can provide a backup ventilatory rate during PS ventilation, but the effects of such a strategy remain to be explored.
Assist/control (A/C), a widely used ventilatory modality in preterm infants, provides mechanical unloading to every spontaneous inspiration similarly to PS and a controlled backup ventilatory rate in the absence of spontaneous breathing. A/C weaning, like in PS, consists of a gradual reduction in PIP, which applies to both assist and control type of breaths. PIP levels during A/C could reach similar low levels as those of PS used in the present study. However, in the absence of spontaneous inspiratory effort, the backup control breaths may not provide sufficient ventilation unless the backup rate is set at levels near the spontaneous breathing frequency. This differs from SIMV+PS as used in the present study where the PIP of SIMV breaths maintained sufficient Vt in the absence of the spontaneous contribution at relatively low SIMV rates. Although this does not necessarily support the use of higher PIP for SIMV breaths, it suggests further investigation on the relevance of maintaining a minimum SIMV rate.
A trial of this nature is limited by the inability to blind the investigators and the clinical team; therefore, the existence of bias cannot be entirely ruled out. To ameliorate this, a weaning protocol was formulated and agreed by consensus between the investigators and clinical team. It can be argued that the initial reduction in ventilator rate in the SIMV+PS group after random assignment could have lead to faster weaning. However, weaning also occurred in the SIMV group, suggesting absence of bias in the management of the ventilator rate during the postrandomization period. To detect bias, specific data were collected. More aggressive rate weaning in the SIMV+PS group or slower rate weaning in the SIMV group would have resulted in differences in ventilator rate or Paco2 levels. However, Paco2 levels were strikingly similar between groups, and ventilator rate differed only at days 4, 7, and 10. Most importantly, although the clinical team weaned both groups of infants using the same criteria, they were not aware of the criterion used to define minimal ventilatory support or when an infant had reached those settings.
In this study, the decision to extubate was left to the clinical team to minimize the influence of any existing bias in the research team, but the existence of bias in the timing of extubation cannot be entirely ruled out. A bias toward SIMV+PS is likely to have resulted in extubation from higher ventilator rates. However, the ventilator rate at the time of extubation was lower in the SIMV+PS group. Bias in the ventilator management is unlikely to have played a role in the total duration and final removal from supplemental oxygen, because for most infants the duration of oxygen exceeded the duration of ventilation, and the removal from oxygen occurred at a later time.
The study protocol was followed until day 28, and subsequently the clinical team determined the modality of ventilation. This study design was aimed at evaluating the use of a potentially gentler ventilatory strategy during the early postnatal period when lung protection may play a more important role. This is perhaps a limitation of the study, because one could speculate that more striking differences could have been obtained, particularly in the smaller BW strata, if the intervention had lasted longer. As a greater proportion of infants in the SIMV group were switched to the alternate experimental modality, this may have also attenuated the differences in respiratory outcome between the 2 groups.
Although the rate of surfactant administration did not differ between groups, it was higher in the SIMV+PS group, particularly within the 700- to 1000-g strata. In this center, surfactant is used only therapeutically, and the greater proportion of infants who required surfactant in the SIMV+PS group correlates with the higher mean Fio2 at random assignment.
Infants who were extremely ill during the first postnatal week and those who did not require mechanical ventilation were not included in the study. Thus, the findings cannot be generalized to the entire population of infants with BW <1000 g. However, infants enrolled in this study represent most infants in that BW group who survive for at least the first 28 days and require mechanical ventilation and oxygen for a significant period of time. This is the group of infants who are at higher risk for lung injury and bronchopulmonary dysplasia and could benefit most from prompt weaning of the ventilatory support.
The results of this study suggest that the addition of pressure support as a supplement to SIMV during the first 28 days may play a role in reducing the duration of mechanical ventilation in extremely low BW infants, and it may reduce oxygen dependency among infants in the 700- to 1000-g BW strata. These data underline the relevance of lung protective ventilatory strategies during the early course of mechanical ventilation in preterm infants. Given the multifactorial nature of lung injury in premature infants and its sequelae, these results need to be further confirmed.
Support for this work was provided by the University of Miami Project NewBorn. Equipment was provided by Viasys Healthcare.
We thank the neonatologists, fellows, nurses, and respiratory therapists of the University of Miami/Jackson Memorial Medical Center NICU for their support. We thank Dr Tilo Gerhardt for thoughtful review of the manuscript. We are particularly grateful to the infants who participated in this trial and their parents or legal guardians for their trust.
- Accepted May 23, 2006.
- Address correspondence to Eduardo Bancalari, MD, Division of Newborn Medicine, Department of Pediatrics, University of Miami Miller School of Medicine, PO Box 016960 R-131, Miami, FL 33101. E-mail:
Financial Disclosure: Viasys Healthcare provided the ventilatory equipment used in this trial. The University of Miami and Drs Bancalari and Claure have a licensing agreement with Viasys Healthcare of a product not related to the ventilator used in this study. Viasys Healthcare provides financial support to an annual educational activity of the University of Miama Division of Newborn Medicine.
This work was presented in part at the meeting of the Society of Pediatric Research; May 1–4, 2004; San Francisco, CA.
- ↵Chen JY, Ling UP, Chen JH. Comparison of synchronized and conventional intermittent mandatory ventilation in neonates. Acta Pediatr Jpn.1997;39 :578– 583
- ↵Roze JC, Liet JM, Gournay V, Debillon T, Gaultier C. Oxygen cost of breathing and weaning process in newborn infants. Eur Respir J.1997;10 :2583– 2585
- Heldt GP, Mcllroy MB. Distorsion of chest wall and work of diaphragm in preterm infant. J Appl Physiol.1987;62 :164– 169
- ↵Numa AH, Newth CJ. Anatomic dead space in infants and children. J Appl Physiol.1996;80 :1485– 1489
- ↵Sinha SK, Donn SM, Gavey J, et al. Randomised trial of volume controlled versus time cycled, pressure limited ventilation in preterm infants with respiratory distress syndrome. Arch Dis Child Fetal Neonatal Ed.1997;77 :F202– F205
- Copyright © 2006 by the American Academy of Pediatrics