Medical Records
ENT characteristics before surgery.
In our comprehensive analysis of medical records, historical data, and preoperative visits, we meticulously compiled the following essential information:
1. Historical cases of recurrent tonsillitis, with tonsil classification based on estimation using the Brodsky scale. Tonsillar hypertrophy was explicitly defined as grades III and IV [
18].
2. Adenoidal hypertrophy.
3. History of recurrent ear infections and/or hearing loss.
4. Diagnosis of hearing loss, carefully assessed through clinical presentations and audiometric tests. Pathological conditions were identified using type B and C tympanograms.
5. Antibiotic therapy undertaken as a treatment for recurrent tonsillitis before surgery was considered.
6. Preoperative PSG was conducted to evaluate sleep-related issues, mainly snoring, sleep apnea, and noisy breathing. In cases where PSG was unavailable, these symptoms were recorded in medical records and classified as SDB or OSA.
The motivation for the surgery was classified into four main groups:
(I) Tonsillar and adenoidal hypertrophy, carefully assessed through comprehensive preoperative physical examinations.
(II) Documented history of recurrent adenoiditis or tonsillitis.
(III) History of recurrent middle ear infections and hearing loss.
(IV) Evidence of SDB or OSA, either instrumental or clinical.
Respiratory polysomnography
During the overnight assessment, limited-channel polysomnography (PSG) was conducted using a portable ambulatory device (SOMNOscreen PSG, developed by SOMNOmedics GmbH in Randersacker, Germany). This advanced device continuously monitored various key physiological parameters relevant to sleep-related studies. These parameters included nasal airflow, thoracic and abdominal respiratory movements (measured using thoracic and abdominal belts), arterial oxygen saturation (SpO2), heart rate (measured via a finger probe), electrocardiogram (ECG), body position (measured via a mercury sensor), and tracheal sounds (recorded through a microphone) [
26].
The device was affixed to the participant between 9:00 PM and 8:00 AM, capturing data throughout the night while the individual remained comfortably at home. The recorded data underwent manual and automatic evaluations to ensure a comprehensive analysis. The DOMINO software, specifically SOMNOmedics v.2.6.0, was used for automated analysis. Additionally, experienced technicians manually reviewed the data to obtain accurate results. Respiratory events were assessed following the guidelines provided by the American Academy of Sleep Medicine (AASM), as described in the study of Marcus, Brooks et al. (2012) [
10].
To assess the severity of OSA, the total number of obstructive apneas, mixed apneas, and hypopneas was calculated and then divided by the real estimated sleep time per hour. This calculation provided the Obstructive Apnea-Hypopnea Index (oAHI), expressed as the number of events per hour. Desaturation was considered if there was a minimum 3% decrease in oxygen levels. All oxygen desaturations (SpO2) were quantified relative to the baseline SpO2 (%) mean and minimum SpO2 (%). The Oxygen Desaturation Index (ODI) was determined by dividing the total desaturations by the full estimated sleep time per hour. This metric provided additional information about the severity and frequency of oxygen level fluctuations during sleep.
Questionnaire at follow-up
The telephone questionnaire was divided into three sections: 1) questions about general and perinatal history, 2) administration of the PSQ; and 3) PedsQL—physical and school domains.
Anamnestic information
The first part of the questionnaire consisted of general questions regarding the perinatal period. In particular, the following information was investigated: birth weight and length, gestational age [
26], and pacifier use. Participants were divided into those who did not use a pacifier or used it for less than two years compared to those who maintained the habit longer [
27]. Other information included active maternal smoking during pregnancy, atopy, malocclusion, and neuropsychiatric disorders. Information regarding COVID-19 infection (symptomatic or asymptomatic) was also collected. Parameters at follow-up included the child’s weight, height, and BMI.
Pediatric Sleep Questionnaire (PSQ)
The Pediatric Sleep Questionnaire (PSQ) is a comprehensive tool consisting of 22 questions regarding various aspects of sleep problems in children, including snoring frequency, loud snoring, observed apneas, breathing difficulties during sleep, daytime sleepiness, distractible or hyperactive behaviour, and other common symptoms associated with pediatric OSAS. These questions can be methodically divided into three distinct groups, namely nighttime symptoms, daytime symptoms, and cognitive symptoms. The PSQ facilitates the identification of potential cases of pediatric OSAS through careful assessment of these symptoms.
Each question within the PSQ requires responses in the form of “yes,” “no,” or “don’t know,” reflecting the participant’s perception of the child’s sleep-related experiences. Consequently, the total score is calculated by dividing the sum of positive responses (yes) by the total number of reactions, excluding those “don’t know.” This standardized approach ensures a fair and accurate representation of the child’s sleep-related problems.
As a result, the derived total score ranges from 0 to 1, with a critical cutoff point of 0.33 used to distinguish patients exhibiting OSAS symptoms from those who do not. This established threshold has been validated and endorsed by Chervin et al. (2000) [
23].
It is important to note that the PSQ has garnered considerable recognition for its high diagnostic value in effectively screening patients suspected of having OSAS. Its ability to comprehensively assess various aspects of sleep disorders, combined with the reliability of the calculated total score, confirms its utility as a robust screening tool [
27,
28,
29].
Questionnaire on quality of life in children (PedsQL)
The PedsQL 4.0 Generic Core Scales is a comprehensive questionnaire to assess Health-Related Quality of Life (HRQOL) in children and adolescents aged 2 to 18. This instrument consists of 23 questions divided into four distinct domains. The assessment involves two parallel versions of the questionnaire, one to be completed by the child and the other by the parent or primary caregiver.
Each question asks participants to rate the frequency of occurrence of a specific problem over the past month on a scale from 0 to 4. These ratings reflect the perceived impact of various health-related aspects on the child’s life. The sum of all scores obtained for each question is divided by the total number of responses given to get the final score. This systematic approach ensures a comprehensive and reliable assessment of the child’s HRQOL across the specified domains. [
30].
It is essential to emphasize that in cases where more than 50% of responses are missing or not provided by the participant, the overall score cannot be accurately calculated.
Statistical Analysis
The statistical analysis described focuses on exploring differences in the percentage distributions of categorical variables among different groups of patients. The chi-square test was used for categorical variables. The percentages of patients with defined categorical variables within the PSQ- and PSQ+ groups are compared. Additionally, the percentages of patients with defined categorical variables between the Ad and A&T patient groups are compared.
Statistical analysis of continuous variables in the PSQ- and PSQ+ patient groups is performed. The mean and standard deviation (S.D.) are provided for each group, along with 98% confidence intervals (C.I. 98%) for the mean. Furthermore, the result of the Mann-Whitney test, including the associated p-value, is reported.
Multiple linear regression analysis is conducted. In this analysis, the dependent variable is the percentage of PSQ% at follow-up, including the independent variables (or predictors). For each independent variable, several parameters are provided T, indicating the value of the estimated coefficient for each independent variable in the regression; Standard Error, meaning the standard error of the estimated coefficient; Beta, representing the standardized beta coefficient, indicating how much the dependent variable changes in terms of standard deviations when the independent variable increases by one standard deviation; t, representing the t-value, which is the ratio of the estimated coefficient to the standard error of the coefficient; and P, indicating the associated p-value for each independent variable, i.e., the statistical significance of the variable’s effect on the dependent variable; the 95% C.I. for B, representing the 95% confidence interval for the estimated coefficient of the independent variable.
A value of P < 0.05 was considered statistically significant for all tests employed. However, we acknowledge that the P-value in our study may be influenced by factors such as the small sample size, risk of bias, and random error. Therefore, we will also consider statistical significance with a P-value between 0.05 and 0.1, providing a reasonable explanation for this value and other evidence supporting the relationship [
31].
The data were recorded in a Microsoft® Excel® database for Windows 11 and statistically analyzed using SPSS version 22.0 for Windows (SPSS Inc., Chicago, IL).