Dados do Trabalho

Título

Morphometric variables and desaturations: proposal for a combined, more accurate index for OSA screening

Introdução

Mathematical models have been long used to identify and stratify risk for obstructive sleep apnea (OSA). The team headed by Clete Kushida developed a predictive morphometric model (KMM) with potential advantages for the accuracy and speed of diagnosis. However, their actual predictive value is still uncertain.

Objetivo

To assess whether the Kushida morphometric and oxy-hemoglobin desaturation indices present sensitivity and specificity to identify individuals with OSA.

Métodos

Cross-sectional study with adults >18 years, both genders, polysomnography (PSG), Epworth Sleepiness Scale, body mass index, neck circumference and intraoral measurements belonging to the KMM. The following were not included: individuals in a situation that could prevent the performance of PSG; previous treatment for sleep-disordered breathing; previous history of malignant tumor in the upper airways; pregnant women. Individuals with the absence of central incisor teeth and/or permanent second molars were excluded.

Resultados

144 patients were selected, 75 of which were excluded. Among the 69 patients studied, women predominated (38; 55.1%), the mean age was 47.3 ±13.4 years, 55 individuals presented AHI≥ 5 ev/h and 14, an AHI< 5 ev/h. When adopting the cut-off point of AHI≥ 5 ev/h, the KMM showed sensitivity (SE)= 60.0%, specificity (SP)= 71.4% and 95% confidence interval of the area under the curve (95% CI of AUC)= 0.655; the KMM+ODI revealed SE= 73.0%, SP= 71.4% (95% CI of AUC= 0.779) and the ODI showed SE= 76.4% and SP= 92.9% (95% CI of AUC= 0.815). At the cut-off point of AHI≥ 15 ev/h, the KMM presented SE= 64.1%, SP= 76.7% (95% CI of AUC= 0.735); the KMM+ODI showed SE= 82.1%, SP= 83.3% (95% CI of AUC= 0.895); and the ODI presented SE= 76.9%, SP= 100.0% (95% CI of AUC= 0.903). For the cut-off point of AHI≥ 30 ev/h, the KMM showed SE= 56.0%, SP= 77.2% (95% CI of AUC= 0.722); the KMM+ODI revealed SE= 92.0%, SP= 79.5% (95% CI of AUC= 0.926); and the ODI showed SE= 92.0%, SP= 90.9% (95% CI of AUC= 0.941).

Conclusões

The isolated Kushida morphometric model failed to show good screening discrimination for OSA in this study sample. However, the combination of ODI and KMM significantly improved the sensitivity and specificity of this model for all OSA severity divisions suggesting that this association may be more effective in risk prediction.

Palavras-chave

Sleep. Obstructive Sleep Apnea. Risk Factors.