Adaptação de teste para rastreio de delirium em idosos admitidos em serviço de urgência

Abstract

OBJECTIVE: Delirium is one of the most frequent syndromes among elderly patients admitted to emergency units and, despite presenting well-established symptoms and signs, there are still diagnostic failures. Thus, the aim of the study was to adapt the Abbreviated Mental Test (AMT) as a screening tool for delirium in elderly adults admitted to an emergency department. METHODS: This cross-sectional study was conducted at the emergency department of a university hospital in Brasília, Brazil between April and June, 2014. We evaluated 90 patients of both sexes, aged 60 years or older. The Confusion Assessment Method O (CAM) Instrument was considered the gold standard for diagnosing delirium. The complete translated AMT, administered in four different culturally adapted models, as well as condensed models, were compared to the CAM receiver operating characteristics curve (significance < 0.05; H0: AUC = 0.5). Inter-rater agreement was evaluated with the kappa test, using SPSS version 22.0.0.0. RESULTS: The prevalence of delirium was 25.6%. The best of the four AMT models presented sensitivity and specificity of 78.3% and 85.1%, respectively, with good inter-rater reproducibility (Kappa = 0.793). The best condensed model included four questions, with sensitivity and specificity of 82.6% and 82.1%, respectively, and a kappa of 0.746. CONCLUSIONS: Compared to the gold standard, the adapted AMT (complete or condensed) was adequate as an alternative for quick delirium screening in elderly patients admitted to an emergency department, especially for unaccompanied patients with no prior cognitive deficit.