Clinical application of Carlsson's questionnaire to predict erosive GERD among healthy Chinese

Gastroesophageal reflux disease (GERD) is a common gastrointestinal disease, yet there is no definitive gold standard to describe and diagnose it.

We used endoscopic examination and Carlsson's questionnaire to evaluate the prevalence of erosive esophagitis during health examinations of individuals in Taiwan.

From October 2001 to March 2002, 778 people underwent self-paid health examinations including esophagogastroduodenoscopic examinations. All subjects completed Carlsson's questionnaire before endoscopy. We determined the positive predict rate, negative predict rate, sensitivity, and specificity of the Carlsson's score for predicting esophagitis and relationships of the score (score > or =4 vs score <4) and esophagitis based on sex, age, body mass index (BMI), smoking, peptic ulcer and drinking habits.

One hundred and thirty-one people with scores > or =4 were highly suspected to have GERD. Of them, 21 were diagnosed as having reflux erosive esophagitis (16.0%) on endoscopic examination. Of 647 people whose scores were <4, 49 were diagnosed with having reflux erosive esophagitis (7.6%). Thus, 70 people were diagnosed as having erosive esophagitis for a prevalence of 9% (70 of 778). The difference between scores > or =4 and <4 to detect esophagitis differed significantly (P < 0.001). Total esophagitis differed significantly according to BMI, drinking habit and sex.

The prevalence of reflux esophagitis is 9.00% at a single medical center in Taiwan. Esophagitis is positively related to higher BMI, alcohol consumption and being of male sex. Predicting the prevalence of esophagitis in a general population by using Carlsson's questionnaire was unsatisfactory.