This prospective cohort study was conducted in the Department of Otolaryngology, Meir Medical Center, Israel (affiliated with Tel-Aviv University) and approved by the Institutional Review Board (0276-18-MMC).

The study was conducted 2018–2020, and included children 1–12 years-old who underwent powered intracapsular tonsillectomy and adenoidectomy (with or without insertion of ventilation tubes) in our department, due to clinical symptoms of SDB. SDB was diagnosed based on history and physical examination. Children with severe obesity (body mass index for age at the 99th percentile or above), Down syndrome, craniofacial abnormalities, neuromuscular disorders, cyanotic heart diseases or mucopolysaccharidoses were excluded. Parents provided a written informed consent.

Procedures were performed under general anesthesia. Enlarged palatine tonsils were partially removed from within the capsule using a straight blade micro-debrider (Medtronic) and adenoid tissue was removed using a curette. Hemostasis was accomplished using local packing with gauze and suction cautery. Surgeries were performed in the morning by residents under the supervision of an attending otolaryngologist. Patients received 0.15 mg/kg intravenous dexamethasone preoperatively.

All PT procedures performed in our institution are inpatient PT, hence the child is kept for surveillance for at least 24 h post-operatively. Essentially, all children were transferred to PACU for recovery, and then admitted for a 24-h overnight surveillance in the Pediatric Surgery Department.

We built a special follow-up protocol designed for this study which included follow-up at 5 time-points: 2-, 4-, 6-, 8- and 24-h post-surgery. Follow-up at each time-point included: 1. visual analogue pain scale (VAS). Scores were given subjectively by the nurse and the parents. 2. Oral intake: any oral liquid or solid intake was reported by the parents. If a successful intake was recorded in one time-point, the subsequent time-points were considered normal. 3. Oxygen saturation, below 93% was considered abnormal 4. Heart rate between the 10th and 90th age-specific percentiles was considered normal. 5. Physical examination; i.e. to rule out bleeding. Children were examined by a resident for signs of oral or nasal bleeding. 6. Urine output. If the child urinated at any time-point, the subsequent time-points were considered normal. 7. Temperature; Sub-axillary temperature above 37.5 °C was considered elevated. 8. Intake of oral analgesics. 9. Intravenous fluid administration.

From the 8th hour post surgery and onward, a standard inpatient clinical monitoring was performed by the nurses, which included bed-side evaluation every 2–3 h. Since monitoring of SpO2 was not continuous, the team was instructed to closely observe these children and report any change in the clinical status. This monitoring included clinical observation for any apneas, significant snoring, increased work of breathing or bleeding.

After the 24-h follow-up, a clinical decision regarding discharge was made (in consultation with a senior, fellowship-trained, pediatric otolaryngologist (Y.E.)). Children were discharged if they had an optimal oral intake and an optimal respiratory status (normal breathing pattern and without increased work of breathing). Reasons for continued admission were defined a priori as: (1) inadequate oral intake, (2) poor hydration, (3) sub-axillary temperature above 38.5 °C, (4) desaturation, or (5) anuria. Furthermore, any major interventions (such as; re-intubation, oxygen supplementation due to desaturation lower than 92%, continuous positive airway pressure or nasopharyngeal airway use) were recorded.

The study population was divided into two groups; younger than 3 years old and older than 3 years old. All above mentioned variables, interventions and complications were compared between the two groups to assess the risks and safety of PT.

Data were analyzed using SPSS v.28 (IBM Corp., Armonk, NY). Where appropriate, data at each time point were recoded into binary variables (0 = absence, 1 = presence): Patients took either no painkillers or at least one; oxygen saturation was either normal or ≤ 92%; heart rate was either normal or tachycardic; temperature was either normal or indicated a high-grade fever (38.5 degrees Celsius or higher); patients either experienced no bleeding or bleeding; intravenous fluids were either not administered or administered. The first instance of either oral intake or urination were similarly coded, except that once a patient had either eaten or urinated, no new information was coded for these variables across the subsequent time points. Patients were also coded as being discharged or having complications 24 h post-surgery.

To facilitate analyses, data across the five time points were collapsed into single variables. VAS scores at each time point were averaged across all five time points to create a grand mean and were analyzed using independent samples t-test. For the remaining variables, if, at least once during the five time points, patients used a painkiller, had oxygen saturation ≤ 92%, were tachycardic, had a high-grade fever, bled, were administered intravenous fluids, had oral intake, or urinated, these variables were coded as 1. Chi-square tests were used to compare these variables between the younger and older groups. We explored whether we had sufficient power to find a large difference between the two groups. For an independent samples t-test with two-tailed alpha = . 05, an allocation ratio of 2:1, d = 0.80 (Cohen, 1992), and power = 0.80, N = 58 is required. For a chi-square test two-tailed alpha = 0.05, df = 1, Cohen’s w = 0.50 (Cohen, 1992), and power = 0.80, N = 32 participants are required. Given our total sample of N = 92 children, a sufficient statistical power level was reached.