Interobserver study of liver histopathology using the Ishak score in patients with chronic hepatitis C virus infection
Interobserver study of liver histopathology using the Ishak score in patients with chronic hepatitis C virus infection
复制标题
DOI：
10.1111/j.1478-3231.1999.tb00033.x
复制
发表时间：
1999-06-01
期刊：
影响因子：
--
通讯作者：
Dhillon, AP
中科院分区：
文献类型：
--
作者：
Westin, J;Lagging, LM;Dhillon, AP
背景资料：评估慢性丙型肝炎病毒（HCV）感染患者的肝损害的组织病理学程度是必不可少的常规护理。已经提出了几种评分系统，试图使组织学评估标准化。Ishak等人提出了一种评分系统，尽管得到广泛认可，但其观察者间的可靠性尚未得到评估。方法：95例慢性HCV感染患者的肝活检由三名独立观察员进行评分。评估界面肝炎、融合性坏死、局灶性坏死、门静脉炎症和纤维化。结果融合性坏死在急性肝炎中更常见，但在任何活检中均未观察到。对于其余的炎症变量（门静脉周围肝炎、局灶性坏死和门静脉炎症），我们发现所有三名观察者的一致率为95-96%。Kappa评分范围为0.11 - 0.41，加权Kappa评分范围为0.18 - 0.53。对于分期，我们注意到84%的一致性，kappa评分为0.26-0.47，加权kappa评分为0.57-0.69。结论Ishak系统与良好的观察者间的可靠性，如果在系统的每个变量的一个分类水平的偏差被接受为协议。与Knodell系统相比，它提供了更详细的信息，但关于纤维化的可靠性较低。
Background: Assessing the histopathological degree of liver damage is essential to the routine care of patients with chronic hepatitis C virus (HCV) infection. Several scoring systems have been proposed in attempts to standardize the histological assessment. One scoring system has been proposed by Ishak et al, Although widely endorsed, its interobserver reliability has not been evaluated. Methods: 95 liver biopsies from patients with chronic HCV infection were scored by three independent observers. Interface hepatitis, confluent necrosis, focal necrosis, portal inflammation, and fibrosis were assessed. Results. Confluent necrosis, which is more common in acute hepatitis, was not seen in any biopsy. For each of the remaining variables of inflammation (periportal hepatitis, focal necrosis, and portal inflammation) we found agreement in 95-96% for all three observers. Kappa scores ranged from 0.11 to 0.41 and weighted kappa scores from 0.18 to 0.53. For staging we noted 84% agreement, kappa scores of 0.26-0.47 and weighted kappa scores of 0.57-0.69. Conclusion. The Ishak system is associated with good interobserver reliability if a deviance of one categorical level in each variable of the system is accepted as agreement. Compared to the Knodell system it provides more detailed information but is less reliable regarding fibrosis.