Effect of high-frequency oscillatory ventilation with intermittent sigh breaths on carbon dioxide levels in neonates
Article information
Abstract
Background
High-frequency oscillatory ventilation (HFOV) minimizes ventilator-induced lung injuries. Spontaneous sigh breathing may augment the functional residual capacity, increase lung compliance, and recruit atelectatic alveoli.
Purpose
To evaluate the difference in the partial pressure of carbon dioxide (PaCO2) in neonates receiving invasive HFOV as the primary mode of respiratory support before versus after sigh breaths (Sighs).
Methods
This prospective study was conducted between January and December 2023. Intubated preterm and term neonates who underwent HFOV with an available arterial line were enrolled in this study after informed parental consent was obtained. Sighs were set at a frequency of 3 breaths/min and pressure of 5 cmH2O above the mean airway pressure for 2 hours. Arterial blood gas was collected before and after Sighs and analyzed using 2 dependent tests.
Results
Thirty neonates with a mean gestational age of 33.6±4.1 weeks and median date of intervention of 1.88 days (interquartile range, 0.87–3.79 days) were enrolled. The mean PaCO2 level was significantly lower in the HFOV with Sighs group (45.2±6.6 mmHg) versus the HFOV alone group (48.8±3.1 mmHg) with a mean difference (MD) of -3.6 mmHg (95% confidence interval [CI], -6.3 to -0.9; P=0.01). Subgroup analyses indicated the ability of Sighs to reduce the PaCO2 level in neonates with respiratory distress syndrome (n=15; MD [95% CI]=-4.2 [-8.2 to -0.2] mmHg; P= 0.04).
Conclusion
Sighing can reduce PaCO2 levels in neonates ventilated with HFOV, particularly those with respiratory distress syndrome.
Key message
Question: Can sigh breaths (Sighs) application during high-frequency oscillatory ventilation (HFOV) decrease partial pressure of carbon dioxide (PaCO2) levels?
Finding: The mean PaCO2 level after Sighs during HFOV was significantly decreased compared to that after HFOV alone (mean difference, -3.6 mmHg).
Meaning: HFOV plus Sighs functionality can reduce PaCO2 levels. However, further studies are required to conclusively determine the effects of Sighs.
Graphical abstract. HFOV, high-frequency oscillatory ventilation; Sigh, sigh breath; FRC, functional residual capacity; CI, confidence interval.
Introduction
High-frequency oscillatory ventilation (HFOV) can be used as either a primary (failing noninvasive ventilation [NIV]) or rescue (failing conventional mechanical ventilation [CMV]) ventilation mode in neonates [1]. Physiologically, HFOV has been considered a lung-protective ventilation method that minimizes ventilator-induced lung injury due to the very small delivery tidal volumes (less than the anatomical dead space) [1]. During HFOV, oxygenation is controlled by the mean airway pressure (MAP) and fraction of inspired oxygen (FiO2), while ventilation is controlled by the oscillatory pressure amplitude (delta pressure, dP) and frequency. The inspiratory-to-expiratory ratio (I:E) affects both oxygenation and ventilation. According to one prior meta-analysis, the application of elective HFOV results in a small reduction in the risk of bronchopulmonary dysplasia (BPD), death, and severe retinopathy of prematurity compared with CMV; however, these effects have been inconsistent across trials [2].
Spontaneous periodic sigh breaths, deep and interspersed breaths of larger tidal volumes than the preceding or following breaths, occur more frequently in newborns than in adults [3]. Sigh breaths in preterm and term infants are much larger than those in adults [4]. These large augmented breaths in infants are often followed by apnea and hypoventilation, likely occurring secondary to the increased activity of peripheral chemoreceptors in neonates [4]. Sigh frequency in preterm infants increases with the degree of prematurity at birth and the severity of BPD [5]. Sigh breaths more commonly precede periodic epochs during non-rapid eye movement (59%) compared to rapid eye movement (18%) during sleep (P<0.005) [6]. Spontaneous sigh breaths may augment functional residual capacity, increase lung compliance, and recruit atelectatic alveoli [3,7]. In animal studies, sigh increases the release of active surfactants [8], and transiently decreases the partial pressure of carbon dioxide (PaCO2) [9].
In mechanical ventilation with spontaneous breathing, automated sustained inflations or sigh breaths (Sighs) are brief cyclic recruitment maneuvers. In intubated adults with acute hypoxemic respiratory failure or acute respiratory distress syndrome (RDS) undergoing pressure support ventilation (PSV), a randomized clinical trial (RCT) have shown that Sighs may improve lung function by improving the PaO2/FiO2 (P/F) ratio [10], while non-RCTs have shown that Sighs function by decreasing regional heterogeneity [11], improving gas exchange and lung volume, and decreasing respiratory drive [12]. A recent multicenter RCT among traumatic adults with Sighs showed increased ventilator-free days (VFDs; 18.4 days vs. 16.1 days) and reduced 28-day mortality rate (11.6% vs. 17.6%) in patients with Sighs compared to those without, with no differences in nonfatal adverse events [13].
In one before-after study, 19 children undergoing PSV after a major surgery received PSV alone or with Sighs. The Sighs group showed an improved P/F ratio (394.2±127.0 vs. 312.6±137.4, P<0.01) and indexed to body weight compliance (1.01±0.30 mL/kg/cmH2O vs. 0.85±0.35 mL/kg/cmH2O, P<0.01), and a decreased PaCO2 level (34.3±4.6 mmHg vs. 39.3±3.3 mmHg, P<0.001) from baseline [14].
In a neonatal non-RCT, 30 infants with very low birth weight (VLBW) were divided into 2 NIV groups (continuous positive airway pressure [CPAP] and high-flow nasal cannula therapy). This study found that spontaneously occurring Sighs on noninvasive respiratory support due to RDS do not increase end-expiratory lung impedance, or alter the delta Z [15]. However, data regarding neonatal respiratory outcomes in the invasive mode, particularly in the primary HFOV mode, are still lacking. Although the HFOV may reduce ventilator-induced lung injury and improve gas exchange, it does not induce neonatal breathing. Thus, Sighs may improve ventilation, avoid atelectasis, and facilitate the transition into neonatal breathing.
This study aimed to evaluate the differences in PaCO2 levels before and after Sighs during invasive HFOV in preterm and term neonates.
Methods
1. Study design and participants
Data were obtained from a prospective intervention (before-and-after and nonrandomized study) of 30 intubated neonates admitted to a neonatal intensive care unit. The study was approved by the Ethical Committee Board of the Faculty of Medicine, Prince of Songkla University (REC 65–407–1–1), and prospectively registered in the ClinicalTrials.gov database (NCT05682937). Informed parental consent was obtained prior to enrollment in the study.
Preterm and term neonates (gestational age [GA], 24–41 weeks), with a postnatal age of <28 days who had already been ventilated with HFOV for at least 1 hour, and had available umbilical or peripheral arterial lines during the study, were enrolled in the present study. The exclusion criteria were as follows: previous or current pulmonary air leaks (pulmonary interstitial emphysema, pneumothorax, pneumomediastinum, and pneumopericardium); heterogeneous lung diseases (including meconium aspiration syndrome and congenital diaphragmatic hernia); suspected lung hypoplasia; suspected or confirmed intraventricular hemorrhage grade III–IV; suspected or confirmed hypoxic-ischemic encephalopathy or 5-minute Apgar score <3 (severe birth asphyxia); hemodynamic instability despite using inotrope(s); PaCO2 level<45 mmHg or >70 mmHg before intervention; requirement for a new arterial puncture for samples of interventions; moribund status; and parents’ decision not to participate.
The withdrawal criteria included the development of air leak syndrome during the intervention, worsening respiratory distress with increasing HFOV setting (changed frequency, MAP, dP, increased FiO2>0.1, and need for suction or positive-pressure ventilation via a self-inflating bag or T-piece resuscitator) during the study, hemodynamic instability and need to increase the dose or add a new inotrope during the intervention, and parents’ decision not to participate in the study.
2. Intervention
After informed consent was obtained, the first arterial blood gas was collected on the HFOV for at least 1 hour. Subsequently, participants were switched to the HFOV-Sighs mode (the same setting as HFOV mode with only sigh breaths added) for 2 hours. A second arterial blood gas sample was collected 2 hours after starting HFOV-Sighs. Subsequently, the sigh breaths were switched off only in the HFOV mode.
To date, only 3 studies have been conducted on the HFOV-Sighs mode [16-18]. Two unpublished studies from the same principal investigator that compared HFOV-Sighs and HFOV alone were identified in the ClinicalTrial. gov registry [17,18]. In this study, patients were treated with a Sighs frequency of 3 breaths/min, a sigh inspiratory time of 1 second (similar to previous studies [16-18]), and a sigh peak inspiratory pressure of MAP+5 cmH2O [16] (maximum 30 cmH2O [17,18]). The study center had 4 ventilator brands (SLE6000, Dräger, Fabian, and Sensor Medic); however, the Sighs mode was only available in the SLE6000 infant ventilators (SLE, London, UK) and Dräger Babylog VN500 (Dräger, Lübeck, Germany). The HFOV-Sighs setting for both the SLE6000 and Dräger Babylog VN500 (frequency, MAP, dP, and I:E) was the same as that of HFOV.
None of the participants required any interruption of ventilation (e.g., suction requirement, disconnection of ventilator, change in frequency, MAP, dP), nor was an increased FiO2<0.1 observed between the interventions for 2 hours.
Monitoring during HFOV and HFOV-Sighs included vital signs (systolic, diastolic, mean arterial pressure, heart rate, and pulse-oxygen saturation, collected every 15 minutes during intervention for 2 hours and after intervention for 1 hour) and chest x-ray within 24 hours after intervention, or immediately if clinical conditions deteriorated.
3. Ventilatory care
Transient tachypnea of the newborn (TTN) is generally diagnosed based on the observation of early onset of remarkable symptom of tachypnea (>60 breaths/min), and prominent perihilar pulmonary vascular markings, fluid in the intralobar fissures, and small pleural effusions, based on chest radiography. Severe TTN is defined as cases of TTN in infants requiring endotracheal intubation and mechanical ventilation [19]. Conversely, RDS is primarily a disease of prematurity, which presents as tachypnea, retraction, grunting, nasal flaring, and cyanosis. Radiographic findings of RDS include low lung volumes, diffuse reticulogranular pattern, and air bronchograms.
Initial NIV support, including oxygen hood, low or high-flow nasal cannula, CPAP, and bilevel CPAP in neonates is supplemented depending on the severity of respiratory distress. Preterm neonates with RDS are provided with early rescue surfactant care if the FiO2 supplementation in either the NIV (Less Invasive Surfactant Administration method) or invasive (INtubation-SURfactant-Extubation method) modes is >0.30. Only neonates with respiratory failure (for example, PaO2<50 mmHg or SpO2<85% in spite of FiO2>0.6 or maximal pressures 8–10 H2O of CPAP or upper level 10–12 cmH2O of bilevel CPAP; or PaCO2>65–70 mmHg and pH<7.2) are considered for endotracheal intubation.
In routine practice in our unit, HFOV is generally used as the primary treatment mode in most intubated neonates with respiratory failure. The conventional mode is considered for use in intubated neonates with normal lungs (e.g., pre- or postoperation, and apnea owing to prostaglandin E1 in cyanotic heart disease). The initial settings of HFOV mode are: frequency 10–12 Hz, MAP 7–13 cmH2O, dP 15–30 cmH2O, and I:E = 1:1. The targeted pulse-oxygen saturation level ranged from 90%–94%. Intravenous caffeine was not available during the study period. Intravenous aminophylline was routinely administered to very preterm, intubated preterm, and apneic infants who were switched to oral caffeine when they were able to tolerate enteral feeding.
4. Outcome
The primary outcome was the PaCO2 level after 2 hours of HFOV-Sighs compared with the HFOV modes. ABL800 BASIC (Radiometer Medical ApS, København, Denmark) analyzed all blood gas samples within 1 minute after collection. The blood gas machine was auto-calibrated every 4 hours by trained specialists every day. Post hoc subgroup analyses were performed on preterm (GA<37 weeks) and very preterm (GA<32 weeks) neonates, with indications for intubation.
5. Sample size
The results of comparisons between the HFOV and HFOV-Sighs modes have not yet been reported. In one crossover RCT in infants (<1 year) after cardiac surgery, the mean PaCO2 levels yielded by PSV and PSV-Sighs were found to be 42.3±3.7 and 40.4±5.0 mmHg, respectively [20]. For the 2 dependent mean comparison studies using a significance level of <5% with 80% power, a sample of 42 neonates was calculated as required to detect differences in PaCO2 levels between the 2 modes (delta mean, 1.9 mmHg; standard deviation [SD], 4.35). We further performed a 1-year study (enrolling approximately 30–50 neonates), and the results were preliminarily analyzed. The study was terminated once the results reached significance.
6. Statistical analysis
R software (ver. 4.3.2; R Foundation for Statistical Computing, Vienna, Austria) was used for statistical comparisons. Categorical variables were presented as the percentages and compared using the χ2 or Fisher exact test. The Shapiro-Wilk test was used to determine the normality of continuous variables. Parametric variables were presented as means±SDs, and a paired t test was used to compare paired samples. Nonparametric continuous variables are presented as the median (interquartile range, IQR), and the Wilcoxon signed-rank test with continuity correction was applied to compare paired samples.
Results
Ninety-four infants were assessed for eligibility; of these, 64 were excluded because of previous or current pulmonary air leaks (n=3), heterogeneous lung diseases (n=7), suspected or confirmed grade III–IV intraventricular hemorrhage (n=1), hemodynamic instability despite using an inotrope (n=15), PaCO2 level<45 mmHg or >70 mmHg before the intervention (n=30), moribund status (n=7), and parents’ decision not to participate (n=1). Thirty participants (neonates) were finally enrolled.
The mean GA and birth weight (BW) were 33.6±4.1 weeks and 2,305±853 g, respectively. The numbers of neonates with GA at 24–31, 32–36, and 37–41 weeks were 9 (30%), 13 (43%), and 8 (27%), respectively. The median (IQR) 1-minute and 5-minute Apgar scores were 7 (5, 8) and 9 (8, 9), respectively. The numbers of neonates who were male, appropriate for GA, and delivered by cesarean section were 17 (57%), 30 (100%), and 23 (77%), respectively. The causes for intubation included RDS in 15 (50%), severe TTN in 11 (37%), congenital heart disease in 3 (10%), and mild to moderate birth asphyxia in 1 (3%).
The median date of intervention was at 1.88 (IQR, 0.87–3.79; range, 0.20–9.55) days of life. The mean body weight and oxygen index were 2,330±880 g and 3.45±1.37, respectively. Before Sighs, the HFOV was set using the following parameters: (1) frequencies of 10 and 12 Hz in 21 and 9 neonates, respectively; (2) MAP values of 7, 8, 9, 10, and 13 cmH2O in 5, 12, 8, 4, and 1 neonates, respectively; (3) mean dP and FiO2 of 21.3±5.9 cmH2O and 0.31±0.08, respectively; and (4) I:E = 1:1. Sighs pressure and frequency were above 5 cmH2O of MAP and 3 breaths/min, respectively.
Arterial blood gas levels are shown in Table 1. The mean PaCO2 levels after Sighs during HFOV was found to be significantly decreased compared to HFOV alone (45.2±6.6 mmHg vs. 48.8±3.1 mmHg; mean difference [MD] -3.6±7.3 mmHg; 95% confidence interval [CI], -6.3 to -0.9; P=0.01). Individual PaCO2 levels before and after Sighs are shown in Fig. 1. Overall, 70% (21 of 30) of neonates achieved decreased PaCO2 levels after Sighs. In HFOV with Sighs, the range of PaCO2 levels was 31.9–56.9 mmHg, and PaCO2<35 mmHg was found in 3 neonates. In the subgroup analysis, the mean PaCO2 levels after Sighs during HFOV was significantly decreased compared to HFOV alone in neonatal RDS (n=15; 44.8±6.2 mmHg vs. 49.0±3.4 mmHg; MD, -4.2 mmHg; 95% CI, -8.2 to -0.2; P=0.04). No differences were found in pH and PaO2 (regarding GA and indication of intubation in Table 1), including gas transport coefficient, minute volume, and oxygen index, before and after Sighs during HFOV (Table 2).
Subgroup analysis of arterial blood gas measurements before versus after Sighs during high-frequency oscillatory ventilation (HFOV) in the subgroup analysis
Individual partial pressure of carbon dioxide (PaCO2) in each individual neonate during high-frequency oscillatory ventilation (HFOV) versus after 2 hours of Sighs (HFOV+Sighs). Sighs, sigh breaths.
Gas transport coefficient, minute volume, and oxygen index before versus after Sighs use (n=30)
During the Sighs intervention, no participants withdrew (for example, due to developing air leak syndrome, changes in the HFOV setting, or the need for a new or increased dose of inotrope). After the intervention, no participants developed air leak syndrome. All patients survived until discharge.
Discussion
Neonates and premature infants have immature respiratory and central nervous systems, meaning that some require ventilatory support. Based on the concept of neonatal ventilation, gentle invasive ventilation, aggressive extubation, and full NIV support have been implicated in these neonates. However, the exact benefits and disadvantages of CMV and HFOV remain unclear. One prior meta-analysis showed that elective HFOV use resulted in lower BPD and severe retinopathy of prematurity rates than CMV [2]. A small tidal volume from the HFOV may reduce both baro- and volutrauma; however, a knowledge gap exists between the optimal MAP and dP for each GA and BW. Fetal breathing involves a very small diaphragmatic movement, whereas neonatal breathing requires a large movement that is similar to CMV.
Sighs application during HFOV may generate transitional ventilation from a small to large tidal volumes (but at a lower rate than CMV), and has a benefit in recruitable lung diseases and the release of active surfactants [8], particularly RDS. In the present study, Sighs (frequency=3 breaths/min in 2 hours) was found to decrease PaCO2 levels in the overall GA and RDS groups. In the subgroup analysis, the lower GA group showed a higher CO2 clearance; however, MDs were insignificant. Therefore, Sighs may be a choice for reducing PaCO2 levels during HFOV and RDS within a short duration of intervention.
Studies investigating Sighs plus invasive ventilation only in the PSV mode in adults and children, and NIV in neonates are limited. Previous studies have shown that, during the invasive PSV mode, Sighs increased the P/F ratio (in adults with acute RDS [10] and children after major surgery [14,20]), tidal volume (in children [20]), gas exchange and lung volume [12] and VFDs (in adults [13]). Moreover, Sighs have been shown to decrease the PaCO2 level [14] and oxygen index [20] in children, as well as the regional heterogeneity [11], respiratory drive [12], and mortality rate [13] in adults. In contrast to one study of VLBW neonates with RDS, Sighs on NIV support were found not to increase end-expiratory lung impedance or alter delta Z [15].
This was a pioneering study of ventilator-generated Sighs application during HFOV in both term and preterm neonates. Two active, as yet unpublished, studies from the same principal investigator that compared HFOV-Sighs and HFOV alone were identified in the ClinicalTrial. gov registry (effects on blood oxygenation [17] and lung volume [18]). Currently, early- or primary-mode HFOV use and extubation from HFOV are increasingly used in clinical practice. Sighs may be considered in ventilated neonates who have hypercarbia or are prepared before extubation to familiarize themselves with big breaths and succeed in extubation.
However, this study had some limitations. First, the duration of the intervention was only 2 hours to ensure the safety of fragile participants. For the next step, Sighs are considered to require extubation (a longer period for more information on benefits and safety), and investigation in an RCT. Second, the sigh setting is still inclusive, particularly during the neonatal period. Further studies are required to enroll more participants (increase the power) to determine the optimal Sighs setting in each subgroup population (different GA and BW). Third, most modern (hybrid) ventilators provide the option for simultaneous delivery of CMV and HFOV, thus enabling the application of periodic Sighs during HFOV, which may assist in lung volume recruitment. However, Sighs mode is available in only a few special ventilator brands. Fourth, no severe complications, such as air leak syndrome, respiratory deterioration, or hemodynamic instability, were observed in any patients treated with HFOV plus Sighs. Three neonates had low PaCO2 levels (30–35 mmHg); therefore, its application should be cautioned and closely monitored. Fifth, overall Sighs in HFOV reduced the PaCO2 level (70%, 21 of 30); however, respiratory monitoring (gas transport coefficient, minute volume, and P/F ratio in Table 2) did not significantly change after Sighs. As such, it is possible that the difference was only insignificant because the sample size was too small to detect the difference. Similar to Sighs during NIV, end-expiratory lung impedance or altered delta Z did not significantly change after Sighs use [15]. Finally, the wide variation in CO2 levels may have occurred due to the small sample size, wide range of GA (24 to 40 weeks), or underlying disease, and were not interrupted during study period.
In conclusion, the present study showed that HFOV integrating Sighs functionality can reduce PaCO2 levels, particularly in patients with recruitable lung disease, such as RDS. However, further studies are required to conclusively determine the effect of Sighs (recruitment maneuvers in general), to better understand their limits, mechanisms, and indications in individual patients, and to determine the optimal Sighs for different BW and lung pathologies.
Notes
Conflicts of interest
No potential conflict of interest relevant to this article was reported.
Funding
This study was funded by the Faculty of Medicine, Prince of Songkla University (REC 65-407-1-1).
Author contribution
Conceptualization: AT, MP, GM, SD; Data curation: AT, KB; Formal analysis: AT, KB; Fund ing acquisition: AT; Methodology: AT, KB; Project administration: AT; Visualization: AT; Writing-original draft: AT, KB; Writing-review & editing: KB, AT, MP, GM, SD.