Studies reporting the reliability of image analysis when assessing skin temperature of the foot are scarce. The aim of this study was to assess the interrater and intrarater reliability of the analysis of foot skin temperature based on the angiosome concept and the association between skin temperature differences and the differences in size of the ROIs. Methodology: Thermograms from 26 feet were analysed by two independent assessors and each assessor analysed the same images on different occasions. Mean temperature values of each of the six ROIs were extracted for analysis. Relative reliability was assessed by Intraclass Correlation Coefficient (ICC) measures and absolute reliability was assessed using Bland and Altman agreement measures and standard error of measurement (SEM). The Spearman correlation coefficient was used to assess the association between the skin temperature differences and the differences in size of the ROIs in the interrater and intrarater analysis. Results: The ICC values evidenced excellent interrater and intrarater reliability with the 95% confidence intervals (CI) ranging between 0.962 and 1.000 and the SEM ranged between 0.00°C and 0.36°C. The mean absolute difference (bias) between the measurements ranged between 0.002°C and 0.117°C and small to moderate associations between the differences in skin temperature and the difference in the number of pixels were identified. Conclusion: The excellent interrater and intrarater reliability measures suggest that the methodology of analysis was reliable and may be used in research and clinical settings. Although statistical significant associations between the skin temperature differences and the differences in size of the ROIs were found, the magnitude of the skin temperature differences between assessments and between assessors (0.02-0.17°C) is not clinically relevant.

No abstract