Acurácia de um questionário sobre xerostomia como instrumento de rastreio de hipossalivação

Abstract

This study aimed to determine the accuracy of a questionnaire on xerostomia as a tool for screening hyposalivation. Overall, 402 adult subjects answered a questionnaire with 08 questions about xerostomia and were submitted to stimulated sialometry, considering hyposalivation values ≤ 0.7 mL / min. Data on sociodemographic conditions and general health were collected. The reproducibility and internal consistency of this questionnaire were analyzed. The correlation between their score and the values of sialometry was verified, besides comparing the total score between the groups with and without hyposalivation. Bivariate and multivariate analyses were also performed using the Decision Tree. Hyposalivation was identified in 162 participants (40.3%). The responses to each question individually revealed variable reproducibility (Kappa = 0.450-0.785) and satisfactory internal consistency (Cronbach's alpha = 0.70). Subjects with higher number of positive responses had lower salivary flow values (Spearman's rho = -0.193, p <0.001) and when the questionnaire score was compared among groups, the hyposalivation group responded positively to a greater number of items. The instrument had sensitivity of 64.8% and area under the curve of 0.60 (95% CI = 0.547- 0.645, p <0.001). Associations with age group of 50-59 years (p = 0.011), female gender (p <0.001), menopause (p = 0.001), alcohol consumption (p = 0.033), systemic disease (p = 0.002) and use of medication (p <0.001) were found. In the multivariate analysis, in addition to gender (p <0.001) and hypertension (p = 0.005), there was an association between hyposalivation and diabetes (p = 0.014). It was found that diabetic and hypertensive women constitute a high-risk group for hyposalivation. The questionnaire may be useful in the epidemiological screening of possible cases of hyposalivation.