医学
自身免疫性肝炎
金标准(测试)
胃肠病学
内科学
疾病
病理
作者
Li Wang,Zhenning Du,Hong‐Li Liu,Yu Zhang,S J Wang,Y. Hu,Liqiu Li,Ping Zhu,Yan‐Dan Zhong,Qing‐Fang Xiong,Yongfeng Yang
摘要
Abstract Background and Aims The International AIH Pathology Group (IAIH‐PG) put forward the new histological criteria of autoimmune hepatitis (AIH) in 2022, which have not undergone adequate verification. In this study, we verified the applicability of the new histological criteria in the population of Chinese patients with chronic liver disease, comparing it with the simplified criteria. Methods The gold standard for diagnosis in all patients was based on histological findings, combined with clinical manifestations and laboratory tests and determined after a follow‐up period of at least 3 years. A total of 640 patients with various chronic liver diseases from multiple centres underwent scoring using the new histological criteria and the simplified criteria, comparing their diagnostic performance. Results In this study, the new histological criteria showed a sensitivity of 73.6% and 100% for likely and possible AIH, with specificities of 100% and 69.0% respectively. The coincidence rates of possible AIH for the new histological criteria, simplified histological criteria and simplified score were 81.7%, 72.8% and 69.7% respectively. For likely AIH, the rates were 89.2%, 75.9% and 65.6% respectively. Based on the new histological criteria, all patients with AIH were correctly diagnosed. Specifically, 73.6% were diagnosed with likely AIH and 26.4% were possible AIH. Additionally, the simplified histological criteria achieved a diagnosis rate of 98.6% for AIH, while the simplified score could only diagnose 53.8% of AIH. Conclusions Compared with the simplified score and simplified histological criteria, the sensitivity and specificity of the new histological criteria for AIH were significantly improved. The results indicate that the new histological criteria exhibit high sensitivity and specificity for diagnosing AIH in China.
科研通智能强力驱动
Strongly Powered by AbleSci AI