Acquired pediatric subglottic stenosis is reported to have a prevalence of 1–2% (Cantarella et al. 2020). This rate reaches 8.3% in different case series (Percul et al. 2023). Acquired pediatric subglottic stenosis is commonly attributed to prolonged intubation, whereas the etiology also involves a history of traumatic or difficult airway (Powell et al. 2020). Balloon dilatation laryngoplasty is an efficient and safe technique for treating primary and secondary acquired laryngotracheal stenosis (Hautefort et al. 2012). It is successfully applied in these children. Airway surgery typically necessitates a deeper level of anesthesia to control airway reflexes and manage the fluctuations in hemodynamic parameters, which are characteristic of this surgery.

Nevertheless, general anesthesia must be applied several times to patients due to the need for multiple balloon dilatations. At the same time, the procedure also necessitates coping with postoperative complications that may arise. Because alveolar collapse, which is related to general anesthesia, impairs gas exchange by creating a shunt effect, it potentially increases perioperative hypoxemic episodes (Habre et al. 2017; Bonasso et al. 2019), which in turn increases the risk for postoperative pulmonary complications. A recent retrospective analysis found that 40.6% of children who underwent balloon dilatation experienced desaturation. Additionally, tracheotomy was required in 15.6% of cases, with an equal percentage needing tracheal intubation (Tuzuner et al. 2022).

Nasal continuous positive airway pressure (CPAP) acts as a “pressure” bridge between spontaneous breathing and controlled mechanical ventilation. As a result, there is an increasing trend in the prophylactic use of nasal CPAP in pediatric patients following high-risk airway procedures to reduce postoperative airway complications. The procedure has been shown to improve oxygenation by reducing the alveolar-arterial oxygen difference after pediatric laparoscopic surgery (Abdel-Ghaffar et al. 2019). However, it has been proven that CPAP applied after pediatric cardiac surgery has favorable effects on peak expiratory flow (Silva et al. 2016). Still, there is no study published on the prophylactic use of CPAP after balloon dilatation in children with tracheal stenosis. Therefore, the objective of this study was to compare recovery time, the requirement for intensive care, tracheostomy, intubation, the number of desaturation episodes, and the rate of airway complications including bronchospasm and tracheal secretions, in pediatric patients with subglottic stenosis who received CPAP versus those who did not receive CPAP.

MethodsStudy design and participant selection

This prospective, double-blinded, parallel-group, randomized controlled clinical study was conducted between the 1st of January 2022 and the 1st of October 2022 at Health Sciences University Ümraniye Training and Research Hospital in Turkey. The study protocol received ethical approval from the Umraniye Training and Research Hospital Ethics Committee with decision number 35 dated 16th of December 2021 (approval number: B.10.1.TKH.4.34.H.GP.0.01/352). Before the conduct of the study or any study-related procedures, written informed consent was obtained from all legal guardians of the participants. The study was designed following the Declaration of Helsinki defined in 2013 (World Medical Association 2013) and the CONSORT 2010 statement (Schulz et al. 2010; Moher et al. 2010).

The study included pediatric patients who were 0 to 12 years of age, classified as II and III according to the American Society of Anesthesiologists (ASA) physical status classification, and who underwent elective subglottic balloon dilatation under general anesthesia due to acquired or congenital subglottic stenosis.

Patients with congenital or acquired diseases of the primary lung or choanal atresia, those younger than 38 gestational weeks, those older than 12 years, and intubated patients were excluded from the study.

Randomization, allocation concealment, and blinding

Patients were randomized in a 1:1 ratio by the principal investigator through computer-generated simple randomization into two groups: the control non-CPAP group (n = 42) and CPAP group (n = 42). Randomization was performed with the Random Integer Generator software (https://www.random.org/). Group allocations were concealed securely in a password-protected computer and were only disclosed to an attending anesthesiologist who was not directly involved in the study. To eliminate potential bias, this was a double-blinded study, where the investigators, analysts, and participants were not aware of the group assignments.

Study procedures and interventions

Before the operation, the 6–4-3–1 regimen (6 h for solids, 4 h for formula and non-human milk, 3 h for breast milk, and 1 h for clear fluids) was followed for fasting (Frykholm et al. 2022). No patient received premedication. We conducted standard patient monitoring, which involved the use of electrocardiogram (ECG), heart rate (HR) assessment, end-tidal carbon dioxide (ETCO2) measurement, pulse oximetry (SpO2), and temperature monitoring. To achieve anesthesia induction, we administered 100% oxygen via an anesthesia mask to infants and children aged 0 to 2 years while ensuring the preservation of spontaneous breathing at the beginning of anesthesia. Additionally, Sevoflurane inhalation anesthesia was administered at a concentration of 2% to achieve a 1 minimum alveolar concentration (MAC). Patients aged 2 to 12 were intravenously given Propofol 1–2 mg/kg. After a thorough evaluation of vocal cord and respiratory tract anatomy and dynamic airway assessment by the otolaryngologist, intravenous Fentanyl was administered at a dose of 0.5 µg/kg to the patient while maintaining spontaneous breathing. In patients for whom continuous airway control was not desired, a dose of 0.6 mg/kg Rocuronium was administered as necessary.

On the other hand, during the maintenance phase, patients for whom the preservation of spontaneous respiration was desired, intravenous Propofol at a rate of 0.5 to 1 mg/kg/h and Remifentanil at a rate of 0.03 mg/kg/min were administered through titration. In all patients, during the balloon dilatation procedure, a transition to apneic respiration was initiated, and a nasal endotracheal tube (ETT) size 2 (trimmed to the appropriate length based on the patient’s age and weight) was used for oxygen insufflation. An SpO2 level below 94% was considered the controlled hypoxemia threshold, where point active ventilation with a mask was initiated, allowing the surgical team to continue with the procedure, and we could return to apneic ventilation. In patients whose hypoxia persisted during mask ventilation, endotracheal intubation was conducted to restore normoxia and normocapnia. Following this, apneic ventilation was resumed. During the procedure, all patients were administered intravenous Paracetamol at a dose of 10 mg/kg, Prednisolone at 1 mg/kg, and, as needed, Theophylline at 3 mg/kg and/or Magnesium Sulfate at 10 mg/kg.

In cases where desaturation occurred after the procedure, patients were either intubated with an appropriately sized uncuffed endotracheal tube based on their age and the level of tracheal stenosis or in patients with tracheostomy tubes, their secretions were aspirated. Intravenous Sugammadex was administered at 3 mg/kg to reverse the neuromuscular effects at extubation. At the end of the procedure, patients suitable for extubation or decannulation were monitored on spontaneous breathing.

During the postoperative period, the CPAP group received FiO2 of 60% and 8 to 12 mmHg of nasal CPAP, or CPAP was initiated through the tracheostomy cannula. The non-CPAP group received FiO2 of 60%, and oxygen support was provided at a rate of 3 L/min either via mask for those who were extubated or through a T-piece for those with a tracheostomy cannula. In the presence of desaturation, FiO2 was increased to 80% in both groups. Patients who experienced intercostal retractions, persistent dyspnea, and unresolved desaturation within 60 min of observation in the recovery unit were transferred to the pediatric intensive care unit. These patients were either re-intubated or provided invasive/non-invasive mechanical ventilation support through the tracheostomy cannula when needed.

Primary and secondary outcomes

The primary outcome was recovery time, measured in minutes. Secondary outcomes were (1) number of desaturation episodes with oxygen saturation < 90%, (2) development of bronchospasm, (3) development of tracheal secretions, (4) requirement for intensive care, (5) requirement for tracheostomy, and (6) requirement for re-intubation.

Recovery time in this study was defined as the period during which the patient was monitored under full observation in the operating theater (OT) following the procedure, as these patients cannot be directly sent to the post-anesthesia care unit (PACU) due to their condition. Hence, the recovery time was the time from Propofol discontinuation to extubation. Patients were transferred to PACU only after full recovery in the OT, and those requiring further critical care were transferred to the intensive care unit (ICU). Thus, recovery time is a vital measure of postoperative stabilization, and CPAP’s potential in reducing this time is of clinical importance.

Data collection and variables

The demographic data of all patients, including sex, age, weight, subglottic stenosis grade, and previous intubation history, were recorded. The grade of subglottic stenosis was determined according to the Myer and Cotton classification. Throughout the procedure, heart rate, mean arterial pressure and oxygen saturation (SpO2) were recorded at five different time intervals (T0: before the procedure, T1: 5th minute, T2: 10th minute, T3: 15th minute, and T4: 20th minute of the procedure). Following the procedure, data on the study outcomes, including bronchospasm, intubation, tracheostomy, desaturation episodes, duration of recovery time, and the need for intensive care were documented.

Statistical analysis

The power analysis was conducted using the G*power 3.1 program for our study. We used an alpha error of 0.05, a power of 0.80, and an effect size of 0.67, which was calculated based on an expected mean difference showing a 10% reduction in recovery time, which—for an average recovery time of around 30 min, as has been noted for this patient population in our hospital—is equivalent to a reduction by 3 min. The total required sample size was determined to be 72 (72/2 groups = 36 patients per group). However, to account for the possibility of ineligible participants requiring exclusion, the number of recruited participants per group was increased by 15%. Hence, the total number of enrolled participants per group was 42.

In the evaluation of the data, descriptive statistical methods were employed, including mean, standard deviation, median, and interquartile range. The distribution of variables was assessed using the Shapiro–Wilk test. The primary outcome was analyzed using independent samples t test while the secondary outcomes were analyzed using logistic regression models. We reported the mean difference, odds ratios (OR), 95% confidence intervals (95%CIs), and the exact p-values. All logistic regression models were tested for link specification using the linktest and model fit using McFadden’s test. All statistical analyses in this study were conducted according to the per-protocol principle using the NCSS (Number Cruncher Statistical System) 2007 Statistical Software package (UT, USA) and Stata 18.0 (StataCorp, College Station, TX, USA).