Acurácia dos questionários de rastreamento de disfagia orofaríngea em adultos e idosos: uma revisão sistemática

Abstract

Introduction: Screening of oropharyngeal dysphagia in adults is usually performed using questionnaires and allows determining the prevalence of this condition in the general population. However, it is necessary that these questionnaires have adequate accuracy (above 0,8). Objective: to verify the accuracy of oropharyngeal dysphagia screening questionnaires in adults. Method: This is a systematic review study, whose survey was carried out in seven electronic databases (CINAHL, EMBASE, LILACS, LIVIVO, PubMed, SCOPUS, Web of Science) and in the gray literature (Google Scholar, Open Gray and quest). The search, selection and extraction of studies were performed blindly by two independent evaluators, without restriction of language or publication date, according to the following eligibility criteria: population 18 years and over; use of oropharyngeal dysphagia screening questionnaires; performance of swallowing videofluoroscopy or swallowing videoendoscopy as a reference standard; and presentation of accuracy data. The risk of bias of the selected studies was assessed using the QUADAS-2 tool and the certainty of evidence using GRADE. Results: Of the 1,759 articles found after removing duplicates, nine underwent title and abstract reading and two were included. The questionnaires used were The Easy Dysphagia Symptom Questionnaire (EDSQ), which showed sensitivity (90.9%), specificity (67.5%) and area under the curve (AUC) values of 0.84, but presented a high risk of bias; and the Screening of Oropharyngeal Dysphagia in the Elderly (RaDI), with values of sensitivity (80%), specificity (89%), positive predictive value (95%), negative predictive value (63%), positive likelihood ratio (7, 64), negative likelihood ratio (0.22) and AUC of 0.88, with low risk of bias. The certainty of evidence was very low. Conclusion: There are only two questionnaires for screening for oropharyngeal dysphagia, both specific for elderly people. The accuracy of the questionnaires varies between 0.84 and 0.88, with a low risk of bias in one of them and high in the other and very low certainty of evidence.