Hendra Koncoro*, I Ketut Mariadi**, Gde Somayana**, IGA Suryadarma**, Nyoman Purwadi**, IDN Wibawa**

* Department of Internal Medicine, Faculty of Medicine University of Udayana/Sanglah Hospital, Denpasar

** Division of Gastroenterology, Department of Internal Medicine University of Udayana /Sanglah Hospital, Denpasar

ABSTRACT

Background: Upper gastrointestinal (UGI) malignancy is one of the major causes of cancer related death. Endoscopy in dyspeptic patients above 45 years, or those with alarm symptoms may detect this condition. There were only limited data in Indonesia about age and alarm symptoms to predict UGI malignancy. This study was aimed to determine the prevalence of UGI malignancy among dyspepsia patients and to develop a simple clinical prediction model.

Method: A cross-sectional study to 390 patients with dyspepsia underwent endoscopy in Endoscopy Unit of Sanglah Hospital Denpasar between July 2012 and June 2013 was conducted. Demography and alarm symptoms were documented. Chi-square and logistic regression test analysis were conducted to analyze variables associated with UGI malignancy.

Results: Twenty (5.13%) of 390 patients with dyspepsia had UGI malignancy. Of the 20 patients, 65% were gastric cancer and 30% were esophageal cancer. The mean age was 59 ± 12 years. Variables associated with UGI malignancy were weight loss (OR = 8.2), dysphagia (OR = 6.2), age > 45 years old (OR = 5.6), gastrointestinal bleeding (OR = 5.5), persistent vomiting (OR = 5.4), and anemia (OR = 4.9). Using a simplified rule of age > 45 years and the presence of any alarm symptom, sensitivity was 85% and specificity was 67.57%.

Conclusions: UGI malignancy was found in 5.13% of patients with dyspepsia who underwent endoscopy. Simple clinical prediction model states that age above 45 years and alarm symptoms may be used as a screening tool to predict UGI malignancy.

Keywords: dyspepsia, alarm symptoms, upper GI malignancy, clinical prediction model

Indones J Gastroenterol Hepatol Dig Endosc Vol.14, No.2, August 2013 (FULL TEXT)