Interobserver reproducibility of cervical cytologic and histologic interpretations - Realistic estimates from the ASCUS-LSIL triage study
Interobserver reproducibility of cervical cytologic and histologic interpretations - Realistic estimates from the ASCUS-LSIL triage study
复制标题
DOI：
10.1001/jama.285.11.1500
复制
发表时间：
2001-03-21
影响因子：
120.7
通讯作者：
Schiffman, M
中科院分区：
文献类型：
--
作者：
Stoler, MH;Schiffman, M
尽管对可靠性有严格的假设，但对于宫颈病理标本的解释，病理学家之间的一致标准尚未明确定义。目的 确定多个训练有素的观察者之间细胞学、阴道镜组织学和环形电外科切除术 (LEEP) 组织学宫颈标本解释的再现性。设计和设置未确定的非典型鳞状细胞 显着性低度鳞状上皮内病变 (ASCUS-LSIL) 分类研究 (ALTS)，一项正在进行的美国多中心临床试验。受试者来自 1996-1998 年期间参加 ALTS 的女性，由 7 个临床中心和 4 个病理质量控制组对 4948 个单层细胞学载玻片、2237 个阴道镜活检和 535 个 LEEP 标本进行解读 (QC) 病理学家。主要结果测量 kappa 值，用于比较原始临床中心解释和第一个 QC 审查员对标本的屏蔽解释。结果 对于所有 3 种标本类型，临床中心病理学家给出的解释明显比审查 QC 病理学家更严格。单层细胞学解释的重现性为中等（kappa = 0.46；95% 置信区间 [CI]，0.44-0.48），相当于穿刺活检组织病理学解释（kappa = 0.46；95% CI，0.43-0.49）和 LEEP 组织病理学解释（kappa = 0.49；95%）的重现性。 CI，0.44-0.55），对于较不严格的解释，组织病理学缺乏可重复性最为明显。结论 对于所有类型的宫颈标本，解释变异性很大。宫颈活检的组织病理学并不比单层细胞学更具重复性，甚至 LEEP 结果的解释也是可变的。考虑到训练有素的病理学家之间存在的不可重复性程度，现实的表现期望应该指导他们的解释的使用。
Context Despite a critical presumption of reliability, standards of interpathologist agreement have not been well defined for interpretation of cervical pathology specimens.Objective To determine the reproducibility of cytologic, colposcopic histologic, and loop electrosurgical excision procedure (LEEP) histologic cervical specimen interpretations among multiple well-trained observers.Design and Setting The Atypical Squamous Cells of Undetermined Significance-Low-grade Squamous Intraepithelial Lesion (ASCUS-LSIL) Triage Study (ALTS), an ongoing US multicenter clinical trial.Subjects From women enrolled in ALTS during 1996-1998, 4948 monolayer cytologic slides, 2237 colposcopic biopsies, and 535 LEEP specimens were interpreted by 7 clinical center and 4 Pathology Quality Control Group (QC) pathologists.Main Outcome Measures kappa Values calculated for comparison of the original clinical center interpretation and the first QC reviewer's masked interpretation of specimens.Results For all 3 specimen types, the clinical center pathologists rendered significantly more severe interpretations than did reviewing QC pathologists. The reproducibility of monolayer cytologic interpretations was moderate (kappa =0.46; 95% confidence interval [CI], 0.44-0.48) and equivalent to the reproducibility of punch biopsy histopathologic interpretations (kappa =0.46; 95% CI, 0.43-0.49) and LEEP histopathologic interpretations (kappa =0.49; 95% CI, 0.44-0.55), The lack of reproducibility of histopathology was most evident for less severe interpretations.Conclusions Interpretive variability is substantial for all types of cervical specimens. Histopathology of cervical biopsies is not more reproducible than monolayer cytology, and even the interpretation of LEEP results is variable. Given the degree of irreproducibility that exists among well-trained pathologists, realistic performance expectations should guide use of their interpretations.