METHODOLOGY: Twenty-four children referred consecutively to the University of Malaya Medical Centre who fulfilled Apley's criteria (at least three episodes of abdominal pain severe enough to affect normal activity over a period longer than 3 months) were tested for lactase deficiency using a pocket breath test analyser (BreatH2 meter; Europa Scientific, Cheshire, England). Lactulose was used to check for hydrogen-producing capacity.

RESULTS: There were 14 males and 10 females in the study, consisting of five Malays, 14 Chinese and five Indians. Mean age was 9.9 years. Seventeen of the 24 children (70.8%) with recurrent abdominal pain who underwent the breath hydrogen test had a positive result. In those with a negative result, subsequent lactulose administration resulted in a positive rise in breath hydrogen. None of the 24 children developed abdominal pain during the test. All the Indian subjects, 71.4% of the Chinese subjects and 40% of the Malay subjects with recurrent abdominal pain had lactase deficiency. The proportion of boys and girls with lactase deficiency was similar (71.4 vs 70.0%, respectively). There was no significant difference between lactase sufficient and deficient children with recurrent abdominal pain with regard to sex, age, ethnic group and clinical features. Following a lactose-free diet, none of the children in the breath hydrogen positive and negative groups reported any appreciable difference in pain symptoms.

METHODOLOGY: Breath samples of patients referred to the Gastroenterology Unit, Women's and Children's Hospital, North Adelaide, South Australia, for confirmation of the diagnosis of carbohydrate malabsorption were analysed with the Quintron microlyzer (Quintron Instrument Co., Milwaukee, USA) and the BreatH2 analyser, using the Quintron microlyzer as the gold standard.

RESULTS: Twenty-nine breath H2 tests (BHT) were performed in 29 patients aged 2 months to 61 years. The sensitivity and specificity of the BreatH2 analyser in detecting a positive BHT using the Quintron microlyser as the gold standard were 0.90 and 0.95 with positive and negative predictive values of 0.90 and 0.95, respectively. There was one false positive and one false negative reading. Bland-Altman plots showed a high degree of agreement between the values obtained with two different methods.