BACKGROUND: Although dyspepsia is a highly prevalent complaint, only a small percentage of patients are found to have serious disease and to need full investigation. The selection of dyspeptic patients who really need upper gastrointestinal endoscopy (UGE) is crucial in order to avoid unnecessary costs and overload services. MATERIAL AND METHODS: 200 primary care patients (age range 16-76 years) with dyspepsia and without any evidence of organic disease were interviewed with a structured questionnaire, and afterwards they underwent UGE. Multiple logistic regression analysis identified variables that could discriminate individuals with abnormal UGE from individuals with normal UGE, as well as individuals with peptic ulcer from individuals with normal UGE. RESULTS: The variables to discriminate individuals with abnormal UGE from individuals with normal UGE were age 45 years or above, and male sex, whereas to discriminate individuals with peptic ulcer in UGE from individuals with normal UGE were age 45 years or above, male sex, smoking, fullness, and absenteeism. A practical model based on these variables was developed, showing sensitivity of 78,9%, specificity of 51,7%, positive predictive value (PPV) of 39,4 % and negative predictive value (NPV) of 86% for the detection of individuals with UGE abnormal; and sensitivity of 84%, specificity of 70%, PPV of 38% and NPV of 95% for the identification of individuals with UGE identifying peptic ulcer. CONCLUSIONS: A strategy based on a model containing clinical and social characteristics of dyspeptic patients could potentially decrease the number of unnecessary UGEs.
Dyspepsia; Endoscopy; Upper gastrointestinal endoscopy; Primary care
