Improvement in the precision of capillary refill time measurements for diagnosing hypovolemic status

• Published in: Journal of Biomechanical Science and Engineering Vol. 18; no. 3; p. 23-00002
• Main Authors: SUGITA, Shukei, NAKAMURA, Masanori, MIZUNO, Takato, UJIHARA, Yoshihiro
• Format: Journal Article
• Language: English
• Published: Tokyo The Japan Society of Mechanical Engineers 01.01.2023; Japan Science and Technology Agency
• Subjects: Algorithms; Capillary refill time; Coefficient of variance; Color; Data collection; Dehydration; Exponential functions; Fingernail; Heart; Image processing; Invasive measurement; Temperature effects; Time measurement; Unloading; Vascular function
• ISSN: 1880-9863; 1880-9863
• DOI: 10.1299/jbse.23-00002

The capillary refill time (CRT) is the time that elapses before the fingernail color returns to red from white after releasing a load applied to the nail. We aimed to estimate hypovolemic status using the CRT. However, the precision of CRT is low. The first step toward improving precision was estimation of the coefficient of variance (CV) via repeated measurements. The best and next-best indices of the CRT change rates in the hypovolemic state were 0.19 and 0.48 respectively, and the corresponding target CVs were ≤ 0.10 and 0.24. The CRTs of subjects were measured by imaging nails during loading and unloading and fitting of an exponential function to the curve of intensity changes. The CRT was 0.77 ± 0.35 s (n = 108), which is comparable to previous studies. However, the CV was 0.37 ± 0.13, which is higher than even the second-best index. Thus, we used only the green color channels of nail images to generate algorithms to define the optimal start time of the CRT calculations; this provided a lower CV (0.31 ± 0.12). Finally, we investigated the effects of fingertip temperature and load release timing on CRT during a cardiac cycle. Correction of the temperature effect with a care on a phase during the cardiac cycle to trigger data collection reduced CV to 0.15 ± 0.11, which cleared the second-best index. Thus, our protocol improves CRT precision; it may be possible to estimate hypovolemic status using the CRT.