Systematic review of pediatric tonsil size and polysomnogram-measured obstructive sleep apnea severity

Systematically review the biomedical literature for data comparing clinical, subjective tonsil size (0-4+ scale) to objectively measured obstructive sleep apnea syndrome (OSAS) using polysomnography (PSG).

PubMed database.

A comprehensive PubMed MeSH search was conducted to identify articles comparing subjective tonsil size to objectively measured OSAS. Inclusion criteria included pediatric patients only, sample size greater than 5, and sufficient data to extract for analysis. Exclusion criteria included patients with obesity or craniofacial syndromes.

Twenty articles were included in the final data set. The mean sample size was 161 (range, 32-700) and grand mean age was 6.7 (range, 2.7-11.7). Case series (evidence based medicine [EBM] level 4) was the predominant study design (16 studies). Eleven of 20 studies concluded there was an association between subjective tonsil size and objective OSAS, whereas 9 did not. Varying statistical techniques were used including simple diagnostic tables (k = 8), linear or logistic regression (k = 19), correlation (k = 5), and analysis of variance (k = 2). A customized quality assessment of each study was performed. Studies showing no association between tonsil size and OSAS had a higher quality score than studies showing an association (3.22 vs 2.36, P = .0317).

The association between subjective pediatric tonsil size using 0-4+ scale and objective OSAS severity is weak at best. High-quality studies suggest no association. Providers must recognize the limitations of using tonsil size in clinical decision making.