The Effect of a Higher Bias Gas Flow on Imposed T-Piece Resistance and Breathing in Preterm Infants at Birth

Objective The resistance created by the PEEP-valve of a T-piece resuscitator is bias gas flow dependent and might affect breathing in preterm infants. In this study we investigated the effect of a higher bias gas flow on the imposed inspiratory and expiratory T-piece resistance and expiratory breaking manoeuvres (EBM) in preterm infants during spontaneous breathing on CPAP at birth. Methods In a retrospective pre-post implementation study of preterm infants <32 weeks gestation, who were stabilised with a T-piece resuscitator, a bias gas flow of 12 L/min was compared to 8 L/min. All spontaneous breaths on CPAP within the first 10 min of starting respiratory support were analysed on a breath-by-breath basis to determine the breathing pattern of each breath and to calculate the imposed inspiratory and expiratory T-piece resistance (Ri, Re), flow rates and tidal volume. Results In total, 54 infants were included (bias gas flow 12 L/min: n = 27, 8 L/min: n = 27) with a median GA of 29+6 (28+4–30+3) and 28+5 (25+6–30+3), respectively (p = 0.182). Ri and Re were significantly lower in the 12 L/min compared to 8 L/min bias flow group [Ri: 29.6 (26.1–33.6) vs. 46.4 (43.0–54.1) cm H2O/L/s, p < 0.001; Re: 32.0 (30.0–35.1) vs. 48.0 (46.3–53.9) cm H2O/L/s, p < 0.001], while the incidence of EBM [77% (53–88) vs. 77% (58–90), p = 0.586] was similar. Conclusion During stabilisation of preterm infants at birth with a T-piece resuscitator, the use of a higher bias gas flow reduced both the imposed inspiratory and expiratory T-piece resistance for the infant, but this did not influence the incidence of EBMs.


INTRODUCTION
Respiratory support, usually in the form of either continuous positive airway pressure (CPAP) or intermittent positive pressure ventilation (iPPV), is often needed to stabilise preterm infants after birth. Currently, this is commonly performed using a T-piece resuscitation device such as the Neopuff TM , which provides positive end expiratory pressure (PEEP) or CPAP when the expiratory outlet is open and positive inflation pressures (PIP) when the expiratory outlet is occluded (1). CPAP or PEEP are generated by providing resistance to flow in the expiratory outlet of the T-piece, also known as the PEEP valve (2). To set the desired PEEP/CPAP level for any given bias gas flow, the outlet that would normally direct flow to the infant is occluded and the resistance of the PEEP valve is regulated by altering its aperture. Higher bias gas flows (e.g., 12-15 L/min) require a larger PEEP valve aperture, because a lower resistance is required to achieve a set PEEP/CPAP level compared to low bias gas flows (e.g., 6-8 L/min). As the infant's expiratory air must pass through the PEEP valve, reductions in the resistance of the valve will result in reduced expiratory resistance imposed on the infant. In theory, reduced imposed resistance should result in reduced breathing effort and energy cost, also known as the imposed work of breathing (iWOB) (2,3).
Recently, a multicentre randomised controlled trial (4) compared a low iWOB respiratory support system with bi-nasal prongs to a high iWOB (5) T-piece system with a face mask as primary respiratory support system in the delivery room. This study reported fewer delivery room intubations when a low iWOB system was used compared to a high iWOB system. Rather than replacing the respiratory support system, we hypothesised that the iWOB may alternatively be reduced by using a higher bias gas flow to reduce the valve resistance of the T-piece.
Although we hypothesise that reducing the imposed T-piece resistance might be beneficial for preterm infants at birth, it is important to evaluate the possible consequences of a reduced expiratory resistance on the infant's breathing pattern. At birth, preterm infants often regulate their own airflow during expiration by increasing laryngeal resistance through expiratory braking manoeuvres (EBM). They close their larynx during expiration to increase pulmonary pressure in order to promote alveolar recruitment and prevent alveolar collapse (2,6). While several studies (7,8) have shown that CPAP reduces EBM in preterm infants at birth, the effect of imposed resistance on EBM is unknown. A lower expiratory resistance could result in higher expiratory airflows and higher deflation rates, which could potentially lead to a loss of FRC due to the lung's higher momentum as it reaches FRC.
Based on previous studies (2,3,5,9), bias gas flows were increased from 8 to 12 L/min in the delivery room at Leiden University Medical Centre (LUMC) in June 2019 to reduce the valve and imposed T-piece resistance during neonatal stabilisation. This was consistent with the operating manual of the Neopuff TM device which recommends a gas flow range from 5 to 15 L/min (10). This retrospective pre-post implementation study aimed to investigate the effect of a higher gas flow on the imposed inspiratory and expiratory T-piece resistance and EBM in preterm infants during spontaneous breathing on CPAP at birth.

MATERIALS AND METHODS
A retrospective pre-post implementation study was performed at the LUMC, a tertiary-level perinatal centre in Netherlands. We selected infants who (i) were born <32 weeks of gestation, (ii) received respiratory support at birth, (iii) had complete respiratory function monitor (RFM) files, and (iv) maintained spontaneous breathing on CPAP for at least 1 min during the first 10 min of respiratory support. All infants were born between March and December 2019, with the higher bias gas flow setting being implemented in June 2019. Due to the transition in June, all infants born in June were excluded. Preterm infants with congenital abnormalities affecting the respiratory transition at birth were excluded. The database was scanned in chronological order and in each group, the first 27 eligible infants were included. Pulmochart software (Applied Biosignals, Weener, Germany) allowed a breath-by-breath analysis to calculate breathing parameters, corrected for birth weight (12).
The use of antenatal corticosteroids and general anaesthetics, mode of delivery, gestational age (GA), birth weight, gender, umbilical cord blood pH and Apgar score at 1, 5, and 10 min after birth were collected from the medical records of all included infants to describe baseline characteristics.
The amount of bias gas flow was measured using the RFM recordings. All spontaneous breaths on CPAP within the first 10 min of respiratory support were analysed on a breath-by-breath basis to calculate peak and mean inspiratory and expiratory flow rates (PIFR, PEFR, MIFR, MEFR), tidal volume (Vt), inspiratory and expiratory time (Ti, Te), minute volume, breathing rate and interbreath variability. When leak or measurement artifacts were present, breaths were excluded from the analysis.
The flow and volume waveforms of the RFM were used to define the breathing pattern with or without expiratory braking, as described by te Pas et al. (6,13). The presence of leak, initial inflations and/or iPPV were noted. In addition, heart rate, oxygen saturation and fraction of inspired oxygen of the first 10 min from birth were noted.
The resistance of the PEEP valve (Rv) as well as the imposed inspiratory and expiratory T-piece resistance (Ri, Re) was calculated using a derivation of Ohm's law (Eq. 1 below and Figure 1).

Rv
P is the pressure difference across the PEEP valve and thus the difference between the pressure within the device (Pd) and atmospheric pressure (Pat; reference to 0 cm H 2 O). Qv is the flow through the PEEP valve, which is the sum of the bias gas flow (Qb) from the Neopuff TM and the expiratory flow through the mask (Qm). During inspiration, Qm has a negative value and so Qv equals the bias flow minus the inspiratory flow through the mask. When CPAP is provided and no flow is entering or leaving the infant (i.e., when Qm = 0, which occurs at end-expiration or end-inspiration), 100% of the bias gas flow (Qb) will exit the T-piece via the PEEP valve and the pressure within the device (Pd) will equal the set CPAP/PEEP level. During expiration, all expired air leaves the system via the PEEP valve, resulting in an increased flow over the valve. As the resistance of the valve and the atmospheric pressure remain the same, the pressure in the device will slightly increase above the set CPAP level. The opposite effect is seen during inspiration; less flow travels through the PEEP valve and the pressure in the device will slightly decrease below the set CPAP level. These fluctuations counteract breathing as the pressure in the lung has to overcome the increased pressure in the device during expiration and the reduced pressure during inspiration. Fluctuations in CPAP were calculated for every breath using Eqs 2, 3.
The local institutional Research Ethics Committee of the LUMC approved the study protocol (G19.129) and issued a statement of no objection for performing this study.
All statistical analyses were performed with IBM SPSS Statistics version 25 (IBM Software, Chicago, IL, United States, 2016). Quantitative data are presented as median (IQR), mean ± SD or n (%) as appropriate. Categorical data were analysed using a Chi-square test, normally distributed continuous variables using the independent t-test and skewed continuous variables using the Mann-Whitney test. Linear regression analysis has been performed to correct for the effect of GA and gender on the percentage of EBMs. A p-value < 0.05 was considered statistically significant. This is the first study to compare the expiratory resistance of a T-piece resuscitator during CPAP with different bias gas flows. Therefore, the sample size was calculated based on the expiratory resistance described by Abbasi et al. (14). This study reported a mean expiratory resistance of 94 ± 12 cm H 2 O/L/s in healthy low birthweight infants <34 weeks 0.5 week postnatally, who didn't receive mechanical ventilation. To detect an absolute decrease of 10% from 94 cm H 2 O/L/s using a standard deviation of 12 cm H 2 O/L/s, with a power of 0.80 and an α-error of 5%, we required 54 infants in total (27 infants per group).

RESULTS
We recorded 107 resuscitations of preterm infants <32 weeks between March and December 2019. Fourteen infants were not eligible as they were born around the implementation date of June 2019. The first 27 infants that met the inclusion criteria were included in each group. The median (IQR) GA was 29 +6 (28 +4 -30 +3 ) for the 12 L/min bias gas flow and 28 +5 (25 +6 -30 +3 ) weeks for the 8 L/min bias gas flow group (p = 0.182), with significantly more males in the 12 L/min bias gas flow group (63.0 vs. 33.3%, p = 0.029). There were no differences in other baseline characteristics between the two groups ( Table 1). The bias gas flow (Qb) measured was 13.2 (12.8-13.9) and 9.1 CPAP exp , PEFR, MEFR and the occurrence of EBM were not different between the groups ( Table 2).
To evaluate the effect of gender and GA on EBMs a regression analysis was performed with covariates bias gas flow group, gender, GA and CPAP level [F(4,49) = 1.429, p = 0.238, R 2 = 0.104], which showed only a significant effect of GA on EBMs (bias gas flow group: p = 0.941, GA: p = 0.048, gender: p = 0.574 and CPAP level p = 0.254).

DISCUSSION
This retrospective study is the first to investigate the effect of a higher bias gas flow on imposed inspiratory and expiratory T-piece resistance and peak airflows in preterm infants during spontaneous breathing while receiving CPAP at birth. We observed that increasing bias gas flow with factor 1.5 reduced the imposed inspiratory and expiratory T-piece resistance with the same factor during stabilisation of preterm infants at birth. The reduced resistance resulted in smaller fluctuations in CPAP FIGURE 2 | Imposed inspiratory and expiratory resistance of the T-piece during spontaneous breathing on CPAP. The median (IQR) resistance (y-axis) for inspiration and expiration of both groups (x-axis) are presented in this boxplot. * p < 0.001. during both in-and expiration, indicating a decreased iWOB. Although a significantly lower PIFR was observed, MIFR, MEFR, tidal volumes, in-and expiration times, minute volumes and the incidence of EBMs remained similar.
Our study confirms the results of a bench test performed by Wald et al. (2), where increasing bias gas flow with a factor 2.5 led to the same factorial reduction in expiratory resistance for the same CPAP level. However, the effect on CPAP fluctuations (peak), inspiratory or expiratory airflows of the infant or the incidence of EBM could not be accounted for in this analysis.
The relationship between resistance and pressure fluctuations as reported in our study was also reported in a cross-over study by Hückstadt et al. (15) who compared CPAP system Infant Flow to the Babylog 8000 in preterm infants admitted to the NICU. In their study, the Infant Flow had smaller CPAP fluctuations, which is attributable to the low expiratory resistance of the device (2,9,15). However, they also reported an increase in PIFR and PEFR when the Infant Flow was used compared to the Babylog 8000 (15). This is in contrast with our results where the reduced resistance did not increase the peak nor the mean airflows. The different outcomes can be explained by Eq. 4 (derived from Ohm's law, Supplementary Material and Figure 1), which shows that, assuming the airway resistance and set CPAP level remain the same, a lower resistance should lead to either higher flows (qm) or lower alveolar pressures (Pal).
The results of Huckstadt (15) can be explained by assuming that both the airway resistance of the infants and the generated alveolar pressure remained the same over the study period, regardless of the device used, resulting in higher airflows when using the device with the lowest resistance.
While these assumptions might be valid in a bench test and in infants who have already successfully transitioned at birth, airway resistance and pulmonary pressures vary during the transition at birth. As the patient characteristics of both groups in our study were comparable, the inspiratory airway resistance of both groups is likely to be similar, suggesting that infants might respond differently to a reduced resistance during their transition by reducing their intrinsic effort. We were not able to measure pulmonary pressures and diaphragm activity, but we hypothesis that the infants detected a lower inspiratory resistance and reacted by expending less effort to produce an inspiration that was equally effective, as indicated by the similar Vt and minute volumes between groups. Although, the significant decrease in PIFR in the post implementation group could imply that the airway resistance was not comparable between our two groups, it is more likely that the reduced resistance led to a different flow curve given the equal mean inspiratory flows.
The fact that the expiratory airflows and the incidence of EBMs were also similar in both group, indicates that the alveolar pressure during expiration must have changed to accommodate the lower resistance. These could include: (i) the elastic recoil of the lung was counteracted by post inspiratory diaphragmatic activity or (ii) active expiration contributed to PEFR to a greater extent in infants receiving a lower bias gas flow. Another explanation could be that we recorded values at 200 Hz and so our measure of PEFR may over-or under-estimate the peak flow during expiration, as it has a steep curve at the start of expiration and a long flattened slope, particularly when expiration is active.
In summary, the different results between the study of Huckstadt (15) and ours could be explained by the differences in methodology (i.e., cross-over vs. retrospective implementation), differences in devices [resistance of Infant Flow is still over a factor 3 less than the resistance of the T-piece with 12 L/min bias gas flow (2)] and/or the timing of our study period (during transition vs. after transition at birth). We hypothesised that reducing the imposed resistance would change the breathing pattern of preterm infants as infants would (i) counteract the passive recoil using diaphragmatic post inspiratory activity and increase the pulmonary resistance, or (ii) increase their breathing rate to compensate for the lower imposed resistance to maintain FRC (6,13). As no differences in breathing were observed in this study, it is plausible to assume that infants benefit from a higher bias gas flow in a T-piece resuscitator as this requires less effort, while still achieving effective breathing. However, as we were not able to measure the effort or airway resistance, this study does not answer the question whether or not decreasing the imposed resistance in preterm infants at birth is clinically relevant. Furthermore, the reported increased airflows by Huckstadt (15) raises the question whether further reducing or eliminating the imposed resistance would have a different impact on the infant's breathing. Further research into the clinical relevance of (reduced) imposed resistance and/or work of breathing is needed before recommendations for neonatal resuscitation devices can be made. Nevertheless, we would suggest that a lower alveolar pressure during expiration potentially may reduce the risk of lung injury, particularly during nonsynchronised iPPV.
This was a retrospective pre-post implementation study and the results should be interpreted with the appropriate caution. Pressure and flow are measured by a flow sensor in-line with the T-piece and it remains unclear if the measured pressure and flow changes are a true representation of the pulmonary pressure and flow changes. Furthermore, the resistance calculations assume laminar flow, but at higher flow rates, it is possible that some flows became turbulent.

CONCLUSION
A higher bias gas flow for generating pressures with a T-piece resuscitator during stabilisation of preterm infants at birth reduced the imposed resistance created by the T-piece resuscitator. This change probably led to less effort but this did not influence EBMs of the infant. This raises the question of whether or not reducing the imposed resistance (and work of breathing) in preterm infants at birth is clinically relevant and whether eliminating the imposed resistance would have an different impact on the infant's breathing.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by Medical Ethics Committee Leiden-Den Haag-Delft. Written informed consent from the participants' legal guardian/next of kin was not required to participate in this study in accordance with the national legislation and the institutional requirements.

AUTHOR CONTRIBUTIONS
KK co-conceived the study (with AP), conducted the study, collected, analysed, and interpreted the data, and wrote the first draft of the manuscript. LW contributed to data collection and analysis, and reviewed and edited the manuscript. SC, AK, and TD contributed to data interpretation, and reviewed and edited the manuscript. SH co-conceived the study, supervised the study, interpreted the data, and reviewed and edited the manuscript. AP co-conceived the study, supervised the study, interpreted the data, and reviewed and edited the first draft of the manuscript. All authors agreed to be accountable for all aspects of the work and approved the final version of the manuscript.