📄 medicine

Prevalence of obstructive sleep apnoea and validation of the OSA-5 questionnaire among children aged 2-7 years with adenotonsillar hypertrophy at Mulago National Referral Hospital, Uganda: a prospective cross-sectional study

This prospective cross-sectional study at Mulago National Referral Hospital found that obstructive sleep apnoea affects 88.5% of children aged 2–7 with adenotonsillar hypertrophy and demonstrated that the OSA-5 questionnaire achieves perfect sensitivity and specificity in diagnosing the condition within this high-risk population, suggesting its utility for prioritizing care in low-resource settings.

Original authors: Geoffrey Tabu, Justine Namwagala, Rym Hidour, Hellen T. Aanyu, Ivan Kimuli

Published 2026-06-28

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Geoffrey Tabu, Justine Namwagala, Rym Hidour, Hellen T. Aanyu, Ivan Kimuli

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a busy hospital in Uganda, Mulago National Referral Hospital, acting like a giant sorting station for children who have trouble breathing while they sleep. The main culprit for these breathing troubles is often adenotonsillar hypertrophy—a fancy way of saying the children's tonsils and adenoids (the little glands in the back of the throat) have grown too big, like overgrown bushes blocking a garden path.

Here is the story of what the researchers found, broken down simply:

The Problem: A Missing Tool

Usually, to know for sure if a child has Obstructive Sleep Apnoea (OSA)—where breathing stops and starts repeatedly during sleep—you need a "Gold Standard" test called Polysomnography. Think of this test as a high-tech, all-night sleep camp where doctors hook the child up to many wires to watch their brain, heart, and breathing.

However, in many places like Uganda, this "sleep camp" is rare. It's expensive, takes a long time, and there aren't enough machines or experts. The doctors needed a simpler, faster way to sort the children who really need help from those who might just be snoring a bit.

The Solution: The OSA-5 Questionnaire

The researchers tested a simple tool called the OSA-5 questionnaire. Imagine this as a 5-question "breathing checklist" that a parent fills out. It asks about five specific things:

Does the child snore?
Do they hold their breath while sleeping?
Do they choke or gasp?
Do they breathe through their mouth?
Are the parents worried about the breathing?

The idea was: If the score is high (5 or more), the child likely has the problem. If it's low, they probably don't.

The Experiment

The team gathered 104 children (ages 2 to 7) who already had big tonsils and adenoids. They did two things for every child:

The Checklist: They had the parents fill out the OSA-5 questionnaire.
The Real Test: They sent the children home with a portable machine (a "sleep tracker" that isn't as heavy as the full lab version) to record their sleep for one night. This portable test was the "truth" they compared the checklist against.

The Results: A Perfect Match

The findings were surprisingly strong:

High Prevalence: Almost 9 out of 10 children (88.5%) in this group actually had sleep apnoea. This makes sense because they were all selected specifically because they had the "overgrown bushes" (big tonsils) that block breathing.
The Checklist Worked Flawlessly: In this specific group of high-risk children, the OSA-5 questionnaire got a perfect score.
- It correctly identified every single child who had the breathing problem.
- It correctly identified every single child who did not have the problem.
- In math terms, it had 100% accuracy for this group.

The Catch (What the Paper Actually Says)

While the results sound like a miracle cure, the paper is very careful about what this means:

The "High-Risk" Filter: The study only looked at children who already had confirmed big tonsils and adenoids. It's like testing a metal detector only on people who are already standing in a pile of gold coins. The detector worked perfectly, but that doesn't mean it will work just as well if you use it on a random crowd of people where most don't have gold.
Not a Replacement Yet: The authors say this tool is great for prioritizing children in busy clinics where sleep labs are scarce. It helps doctors decide who needs to be seen first. However, they explicitly state that in places where full sleep labs are available, the checklist should not replace the real test, especially for children with other health issues.
Need for More Testing: The paper concludes that while this tool is excellent for this specific group of sick children, it needs to be tested on "lower-risk" children (like kids in regular schools) before it can be used as a standalone tool for everyone.

The Bottom Line

In this specific hospital setting, the simple 5-question checklist was a perfect match for the complex sleep machines when looking at children with very large tonsils. It offers a practical way for doctors to quickly spot the children who need urgent care, but the researchers warn that we need to test it on a wider, healthier group of children before declaring it a one-size-fits-all solution.

Technical Summary: Prevalence of OSA and Validation of the OSA-5 Questionnaire in Ugandan Children

Problem Statement
Obstructive sleep apnoea (OSA) in children is a significant health concern associated with pulmonary hypertension, neurocognitive impairment, and growth failure. While adenotonsillar hypertrophy is the primary anatomical cause, the gold standard for diagnosis—overnight polysomnography (PSG)—is often inaccessible in low-resource settings like Uganda due to high costs, equipment requirements, and personnel constraints. Consequently, there is a critical need for validated, low-cost screening tools to triage children with adenotonsillar hypertrophy for specialist evaluation and surgery. The OSA-5 questionnaire, a brief five-item caregiver-reported tool, has been proposed for this purpose but lacked validation in the specific context of Ugandan pediatric populations.

Methodology
This study employed a prospective cross-sectional diagnostic accuracy design conducted at the Ear, Nose, and Throat (ENT) Clinic of Mulago National Referral Hospital, Kampala, Uganda, between February 1, 2020, and January 31, 2022.

Study Population: The cohort consisted of 104 children aged 2–7 years with clinically and radiologically confirmed adenotonsillar hypertrophy. Exclusion criteria included a body mass index (BMI) above the 95th percentile, known cardiac or chronic lung disease, craniofacial deformities, neuromuscular disorders, and other specific causes of upper airway obstruction.
Data Collection:
- Screening: Caregivers completed the OSA-5 questionnaire, which assesses snoring, breath-holding, choking, mouth breathing, and parental concern regarding breathing. Scores range from 0 to 15, with a cut-off of $\ge$ 5 indicating a likelihood of OSA.
- Reference Standard: All participants underwent overnight portable sleep monitoring using the Alice PDX system. The device recorded oxygen saturation, pulse rate, airflow, respiratory effort, and body position.
- Clinical Assessment: Tonsillar size was graded via the Brodsky classification, and adenoid size was assessed via lateral neck radiography using the adenoidal-nasopharyngeal ratio.
Definitions: OSA was defined as an Apnoea-Hypopnoea Index (AHI) greater than 1 event per hour. Diagnostic accuracy metrics (sensitivity, specificity, positive predictive value, and negative predictive value) were calculated using the portable sleep monitoring results as the reference standard.

Key Results

Demographics: The mean age of participants was 4.6 years, with 54.8% being female. The mean BMI was 15.9 kg/m².
Symptomatology: Snoring (96.1%), mouth breathing (95.2%), parental concern (94.2%), and breath-holding episodes (89.4%) were the most prevalent symptoms.
OSA Prevalence: Of the 104 enrolled children, 92 were diagnosed with OSA via portable sleep monitoring, yielding a prevalence of 88.5% (95% CI: 82.2–94.7).
Diagnostic Accuracy of OSA-5: Using the predefined cut-off score of 5:
- The questionnaire identified all 92 children with OSA as positive and all 12 children without OSA as negative.
- Sensitivity: 100% (95% CI: 96.1–100.0)
- Specificity: 100% (95% CI: 73.5–100.0)
- Positive Predictive Value (PPV): 100% (95% CI: 96.1–100.0)
- Negative Predictive Value (NPV): 100% (95% CI: 73.5–100.0)
Item Analysis: Within the questionnaire, breath-holding, mouth breathing, and parental concern were significantly associated with OSA on sleep monitoring, whereas snoring and choking did not show significant predictive value in this specific sample.

Significance and Claims
The authors conclude that OSA is highly prevalent among children aged 2–7 years with adenotonsillar hypertrophy in this high-risk referral setting. The study claims that the OSA-5 questionnaire demonstrated excellent diagnostic accuracy within this specific population, correctly classifying every participant against the reference standard.

The paper posits that in resource-limited settings where formal sleep laboratory services are scarce, the OSA-5 questionnaire serves as a practical tool for:

Identifying children with adenotonsillar hypertrophy who are likely to have OSA.
Prioritizing children for further specialist evaluation and surgical care.

Limitations and Scope
The authors explicitly note that these findings are specific to a high-risk clinical population (children with confirmed grade III/IV tonsillar hypertrophy) and may not be generalizable to the broader, lower-risk pediatric population or community settings. They acknowledge that the small number of children without OSA (n=12) resulted in wide confidence intervals for specificity and NPV. Furthermore, the study cautions that the questionnaire should not replace objective sleep testing where such testing is available, particularly for children with comorbidities or atypical presentations. The authors recommend further validation in general pediatric and lower-risk populations before the tool is adopted as a standalone screening mechanism.