AUTHOR=Riveline Jean-Pierre , Mallone Roberto , Tiercelin Clarisse , Yaker Fetta , Alexandre-Heymann Laure , Khelifaoui Lysa , Travert Florence , Fertichon Claire , Julla Jean-Baptiste , Vidal-Trecan Tiphaine , Potier Louis , Gautier Jean-Francois , Larger Etienne , Lefaucheur Jean-Pascal 

TITLE=Validation of the Body Scan®, a new device to detect small fiber neuropathy by assessment of the sudomotor function: agreement with the Sudoscan®

JOURNAL=Frontiers in Neurology

VOLUME=Volume 14 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2023.1256984

DOI=10.3389/fneur.2023.1256984

ISSN=1664-2295

ABSTRACT=Background: Sudomotor dysfunction is one of the earliest manifestations of small fiber neuropathy (SFN), reflecting the alteration of sympathetic C fiber innervation of the sweat glands. Among other techniques, such innervation can be assessed by measuring electrochemical skin conductance (ESC) in micro-Siemens (µS). In this study, ESC was measured at the feet to detect distal SFN. For this objective, the performance of a new device, the Body Scan® (Withings, France), intended for home use, was compared with that of a reference device, the Sudoscan® (Impeto Medical, France), which requires a hospital setting.
Methods: In patients having diabetes with or without neuropathy or non-diabetic patients having lower-limb neuropathy, the diagnostic performance of the Body Scan® measurement was assessed by calculating its sensitivity (Se) and specificity (Sp) to detect at least moderate SFN (Se70 and Sp70), defined by a value of feet ESC ≤70µS and >50µS on the Sudoscan® measure, or severe SFN (Se50 and Sp50), defined by a value of feet ESC ≤50µS on the Sudoscan® measure. The agreement between the two devices was assessed with the analysis of Bland-Altman plots, mean absolute error (MAE), and root mean squared error (RMSE) calculations. The repeatability of the measurements was also compared between the two devices.
Results: A total of 147 patients (52% men, mean age 59 years old, 76% diabetic) were included in the analysis. The sensitivity and specificity to detect at least moderate or severe SFN were: Se70=0.91 ([0.83, 0.96]), Sp70=0.97 ([0.88, 0.99]), Se50=0.91 ([0.80, 0.98]), and Sp50=0.99 ([0.94, 1]), respectively. The bias and 95% limits of agreement were 1.5 [-5.4, 8.4]. The MAE was 2.9 and the RMSE 3.8. The intra-sample variability was 2.0 for the Body Scan® and 2.3 for the Sudoscan®.
Conclusion: The ESC measurements provided by the Body Scan® were in almost perfect agreement with those provided by the reference device, the Sudoscan®, which validates the accuracy of the Body Scan® for the detection of SFN. By enabling simple, rapid and autonomous use by the patient at home, this new technique will facilitate screening and monitoring of SFN in daily practice.