Bland-Altman图评价高频超声和病理测量大鼠皮肤厚度的一致性研究

目的探讨高频超声测量大鼠供皮区皮肤各层厚度的准确性。 方法以购自海军军医大学动物实验中心的10只SD大鼠为对象，每只大鼠构建4个供皮区模型，共40个供皮区模型，分别应用高频超声和组织病理学方法，测量大鼠供皮区表皮厚度、真皮厚度、全层皮肤厚度以及断层皮片厚度，Bland-Altman散点图计算测量数据的95%一致性界限(95%LOA)。 结果Bland-Altman散点图显示两种方法测量表皮厚度、真皮厚度、全层皮肤厚度以及断层皮片厚度差值的95%LOA分别为(－15.80～14.30 μm，P＝0.54)、(－9.99～8.59 μm, P＝0.36)、(－32.33～32.03 μm, P＝0.95)、(－33.62～32.87 μm, P＝ 0.89)。两种方法测量表皮厚度差值和断层皮片厚度差值的点都分别在其95%LOA内，真皮厚度、全层皮肤厚度差值各仅有1个点落其95%LOA外。真皮厚度、全层皮肤厚度以及断层皮片厚度测量差值的95%LOA全部在临床可接受的范围内，表皮厚度测量差值的95%LOA有95%的点在临床可接受的范围内。 结论高频超声测量大鼠供皮区皮肤各层厚度的数据与组织病理学测量数据具有较好的一致性。高频超声作为一种无创、快速、准确、客观的方法，可望在供皮区取皮深度评估中得到广泛应用。

Application of Bland-Altman plot in evaluating the agreement between high-frequency and histology for measuring skin thickness in rats

ObjectiveTo investigate the accuracy of high-frequency ultrasound for measuring the thickness of each layer of rat skin from a donor site. MethodsEstablishing 40 models of the donor site which came from ten SD rats. The thickness of each layer of skin from rat donor sites and the thickness of split skin were measured using high-frequency ultrasound and histology, respectively, and then the 95% limits of agreement (95%LOA) for measurement data was calculated by the Bland-Altman approach. ResultThe Bland-Altman plot showed that the 95% LOA of the differences in the epidermis thickness, dermal thickness, full-thickness skin thickness and split skin thickness was (-15.80 μm to 14.30 μm, P=0.54), (-9.99 μm to 8.59 μm, P=0.36), (-32.33 μm to 32.03 μm, P=0.95 ), and (-33.62 μm to 32.87 μm, P= 0.89), respectively. In this study, the Bland-Altman plot shows that the difference in dermal thickness and the thickness of the split skin calculated by the two methods were all within the 95% LOA. Only one spot representing the difference in dermal thickness was outside the 95% LOA, and one spot representing the difference in full-thickness skin thickness was outside the 95% LOA. Compared with the maximum allowable clinical error range (±10% of the mean thickness of histological measurements), the 95% LOA of the dermal thickness, full-thickness skin thickness and split skin thickness were all within the clinically acceptable range, and the results showed that 95% of the points of the epidermal thickness were within the 95%LOA and within the clinically acceptable range. ConclusionA high level of agreement was discovered between high-frequency ultrasound and histology for measuring skin thickness of donor sites. High-frequency ultrasound can provide non-invasive , qucik, accurate and objective measurement for skin thickness, which have promising application in assessing the depth of donor site.