Sweat test for cystic fibrosis: Wearable sweat sensor vs. standard laboratory test

• Published in: Journal of cystic fibrosis Vol. 17; no. 4; pp. e35 - e38
• Main Authors: Choi, Dong-Hoon, Thaxton, Abigail, Jeong, In cheol, Kim, Kain, Sosnay, Patrick R., Cutting, Garry R., Searson, Peter C.
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01.07.2018
• Subjects: Cystic fibrosis; Personalized medicine; Sweat rate measurement; Sweat test; Wearable sweat chloride sensor; Cystic fibrosis; Wearable sweat chloride sensor; Sweat test; Sweat rate measurement; Personalized medicine
• DOI: 10.1016/j.jcf.2018.03.005
• ISSN: 1569-1993; 1873-5010

Sweat chloride testing for diagnosis of cystic fibrosis (CF) involves sweat induction, collection and handling, and measurement in an analytical lab. We have developed a wearable sensor with an integrated salt bridge for real-time measurement of sweat chloride concentration. Here, in a proof-of-concept study, we compare the performance of the sensor to current clinical practice in CF patients and healthy subjects. Sweat was induced on both forearms of 10 individuals with CF and 10 healthy subjects using pilocarpine iontophoresis. A Macroduct sweat collection device was attached to one arm and sweat was collected for 30 min and then sent for laboratory analysis. A sensor was attached to the other arm and the chloride ion concentration monitored in real time for 30 min using a Bluetooth transceiver and smart phone app. Stable sweat chloride measurements were obtained within 15 min following sweat induction using the wearable sensor. We define the detection time as the time at which the standard deviation of the real-time chloride ion concentration remained below 2 mEq/L for 5 min. The sweat volume for sensor measurements at the detection time was 13.1 ± 11.4 μL (SD), in many cases lower than the minimum sweat volume of 15 μL for conventional testing. The mean difference between sweat chloride concentrations measured by the sensor and the conventional laboratory practice was 6.2 ± 9.5 mEq/L (SD), close to the arm-to-arm variation of about 3 mEq/L. The Pearson correlation coefficient between the two measurements was 0.97 highlighting the excellent agreement between the two methods. A wearable sensor can be used to make real-time measurements of sweat chloride within 15 min following sweat induction, requiring a small sweat volume, and with excellent agreement to standard methods. [Display omitted]