Pediatric tonsil size: objective vs subjective measurements correlated to overnight polysomnogram

To compare subjective and objective tonsil size measurements (weight, volume, intertonsillar distance) along with simple pharyngeal measurements in the prediction of pediatric obstructive sleep apnea hypopnea syndrome (OSAHS) severity.

Prospective case series.

Tonsil size (0-4+) and Mallampati/Friedman palate position were subjectively measured. During adenotonsillectomy, tonsil size was objectively measured along with simple pharyngeal dimensions. Spearman rank correlation coefficients were calculated and forward and reverse stepwise multivariate regression modeling was used to assess the prediction of preoperative OSAHS severity as determined by polysomnogram (PSG). Evaluation of possible outlier influence was also performed.

Thirty-four pediatric patients (median age, 4; range, 2-9) were included. Objective tonsil weight (Spearman's rho = 0.6143, P = 0.0002), tonsil volume (rho = 0.4960, P = 0.0039), and intertonsillar distance (rho = -0.7559, P < 0.0001) were strongly correlated with subjective tonsil size but not with age, body mass index (BMI), or preoperative apnea-hypopnea index (AHI). Regression modeling demonstrated that only measured tonsil weight (beta = 1.43, P = 0.003), age (beta = -3.21, P = 0.001), and hard palate length (beta = 0.979, P = 0.003) were significant predictors of the preoperative AHI (R(2) = 0.5358). Evaluation of possible outliers indicated tonsil weight was the most robust predictor of preoperative AHI.

Subjective tonsil size measurements correlate well with objective tonsil volume measurements. However, only objective tonsil measurements were significantly predictive of objective PSG-measured OSAHS severity.