Introduction Hydrogen breath testing is a valuable tool to aid in the diagnosis of carbohydrate malabsorption, however this procedure can be time consuming and costly for both the patient and clinician. Here, an existing bench-top clinical breath test device is compared to a hand-held, app-connected breath test device, which could allow for at-home use, enabling a more time and cost-effective physician-patient interaction and treatment pathway.

Methods Fourteen healthy adult volunteers mean age 31 (2–3) yrs, were recruited. Breath tests were performed by each volunteer using two of the proposed devices and the results were compared against those of the existing clinical device, a reference standard benchtop device. A baseline reading was recorded using each device prior to the ingestion of 10g of lactulose, a non-absorbable carbohydrate substrate. Each volunteer took a breath reading every five minutes, sequentially switching between each device, such that over a 15-minute interval, three breath samples were recorded, each with a different device. Over the course of three hours, this yielded 39 data points per volunteer. The exclusion criteria for analysis was a baseline breath hydrogen concentration of >15ppm on any of the devices. One volunteer was excluded from the final data analysis due to a high H₂ baseline leaving 13 subjects for analysis. The H₂ concentrations recorded by the reference device were used to assess whether malabsorption had occurred. In instances where there was a >20ppm increase in H₂ over the initial baseline (Pimentel et al, Am J Gastroenterol. 2000 Dec;95(12):350–), the test was considered positive, otherwise it was considered negative. The H₂ concentrations from each device were compared at 15-minute intervals, using concentration plots created for each volunteer, by linearly interpolating the H₂ concentrations over time.

Results There was diagnostic agreement in 13 out of 13 (12 positive and one negative) cases between the proposed devices and the reference device. The mean difference was 3.6 ppm and the mean absolute difference was 7.4 ppm.

Conclusions Initial testing on healthy volunteers suggests that the proposed device may offer results comparable to those delivered by the reference benchtop device. Furthermore, the use of the proposed device via the associated smartphone app could allow for the immediate transmission of results to the clinician, for use in further consultation and guidance. Further validation studies are necessary to assess the performance of proposed device over a longer time period, using a larger cohort of patients, with a focus on those with functional digestive disorders.