• Author / Uploaded
• breathanalysis

• Categories
• Medicina CLINICA
• Medicina
• Especialidades Medicas
• Matemáticas

Citation preview

EVALUATION OF THE ACCURACY OF A PORTABLE DEVICE, GASTROLYZER EC60 BEDFONT, FOR HYDROGEN (H2) DETECTION IN EXPIRED AIR Di Stefano M, Missanelli A, Miceli E, Mazzocchi S, Corazza GR. Gastroenterology Unit, IRCCS Policlinico S. Matteo, University of Pavia, Pavia; Italy.

Background and Aims: H2 Breath Test (H2BT) is a simple, noninvasive, repeatable and cheap test for diagnosis of carbohydrate malabsorption. Recently, a portable device, Gastrolyzer EC60 Bedfont, was proposed for detection of H2 in expired air, but until now there were no studies which evaluated its accuracy. Patients and Methods: 32 patients (20 female, mean age 40±16 yrs) with functional bowel complaints, who underwent H2breath test as a part of their diagnostic algorithm, took part in this study. Fifteen of them underwent H2BT after lactose due to lactose malabsorption, 9 after glucose due to small bowel bacterial overgrowth, 8 after lactulose for an evaluation of cumulative H2 production capacity. In each breath sample, H2 levels were measured with the new device and with the conventional instrument, Quintron Microlyzer DP 12, considered the gold standard. Lactose breath test was positive if an increase >20ppm over baseline was present. Glucose breath test was positive if an increase >12 ppm over baseline was present. Cumulative H2 production capacity was considered elevated if the area under the time-concentration curve (AUC) was >5000 ppm x min. Statistical analysis was performed by the Bland and Altman graphical method and by the Lin’s concordance correlation coefficient. The Kappa statistics were calculated to evaluate the agreement between the two methods in terms of test positivity. Results: The concordance calculated with the Lin’s correlation coefficient was 97,2% with a confidence gap between 95 and 99%. The agreement between the positivity/negativity of the lactose and glucose tests was 100% (Kappa). The repeatability for the AUC (Bland and Altman) was estimated around 98,5% with a uniform distribution of the data.

N ote: Da ta mu st o ve rla y da she d l ine fo r pe rfect co nco rda n ce

95% Limits Of Agreement

Difference of lnauc_quintronandlnauc_bedfont

lnauc_quintron

10

8

6

Concordance calculated with the Lin’s correlation coefficent

4 4

6

ln au c_ be dfo nt

8

10

2

1

0

Repeatability for the AUC performed by the Bland and Altman graphical method

-1 4

6 8 Mean o f lnauc_qui ntron and lnauc_bed font

10

Conclusions: The gastrolyzer is a very accurate device. Due to a simple and quick procedure for breath collection, it can be used in any location, allowing a prolonged, ambulatory measurement of breath H2 excretion. The adoption of recently described, more accurate diagnostic criteria for carbohydrate malabsorption, based on a 7-hour measurement of breath H2 excretion could be easier using such a portable device.