Golam Mostafa, Queen Zarat
National Institute of Cancer Research and Hospital, Department of Histopathology, Mohakhali, Dhaka-1212, Bangladesh.
Malays J Med Sci. 2015 Jul-Aug;22(4):32-9.
While intraoperative consultation has been used in Bangladesh for a long period of time, to date, there has been no published reporting on the performance of frozen sections. The current audit evaluates the performance of frozen sections in a well reputed medical center in Bangladesh, Anowara Medical Services.
This retrospective study has been designed to measure the accuracy of frozen section diagnosis in a medical center in a third-world country, where many surgical procedures rely on intraoperative consultation.
A series of 1379 intra- and peri-operative frozen section cases, from 2007 to 2014, was reviewed. Intraoperative tissue specimens received at Anowara Medical Services were processed for frozen sections. After examination of the frozen section that yielded the initial frozen section diagnoses, the frozen tissues were reprocessed for regular paraffin sectioning. These paraffin sections were examined by a second pathologist, and a final diagnosis was issued. The frozen section diagnosis and final diagnoses of all cases were retrospectively analysed to determine the accuracy of frozen section examination.
Overall, accurate diagnosis was made on frozen sections in 98.2% of the cases. The discrepant diagnoses were all clinically significant, i.e., there were discrepancies between benign and malignant diagnoses on frozen and paraffin sections. In 1% of the cases, diagnosis was deferred. Fifty percent of the deferred cases were benign. Two cases, received in formalin, were excluded. In both cases, the diagnosis was positive for malignancy. The number of false negative results (4 false negatives) was slightly lower than that of false positives (5 false positives). Specificity and sensitivity of 99.3% and 99.4% were achieved, respectively. In this study, the positive predictive value was 99.2% and the negative predictive value was 99.5%. Over the years, the number of discrepant diagnoses remained fairly constant.
The present method has a satisfactory rate of accuracy of frozen section diagnosis, which is comparable to other remote and recent published series. The results of this study offer a testament to the reliability of frozen section diagnosis rendered by qualified pathologists in Bangladesh and may serve as evidence in building confidence among the surgeons who use this service for improved patient care.
虽然术中会诊在孟加拉国已使用很长时间，但迄今为止，尚未有关于冰冻切片性能的公开报告。当前的审计评估了孟加拉国一家著名医疗中心——阿诺瓦拉医疗服务中心冰冻切片的性能。
这项回顾性研究旨在衡量一个第三世界国家的医疗中心冰冻切片诊断的准确性，在该中心许多外科手术依赖术中会诊。
回顾了2007年至2014年期间的一系列1379例术中和围手术期冰冻切片病例。阿诺瓦拉医疗服务中心接收的术中组织标本被处理用于冰冻切片。在检查产生初始冰冻切片诊断的冰冻切片后，将冰冻组织重新处理用于常规石蜡切片。这些石蜡切片由另一位病理学家检查，并给出最终诊断。对所有病例的冰冻切片诊断和最终诊断进行回顾性分析，以确定冰冻切片检查的准确性。
总体而言，98.2%的病例在冰冻切片上做出了准确诊断。差异诊断均具有临床意义，即冰冻切片和石蜡切片上的良性和恶性诊断存在差异。1%的病例诊断被推迟。50%的推迟病例为良性。排除了2例用福尔马林固定的病例。在这两例中，诊断为恶性阳性。假阴性结果的数量（4例假阴性）略低于假阳性结果的数量（5例假阳性）。特异性和敏感性分别达到99.3%和99.4%。在本研究中，阳性预测值为99.2%，阴性预测值为99.5%。多年来，差异诊断的数量保持相当稳定。
本方法冰冻切片诊断的准确率令人满意，与其他近期发表的系列研究相当。本研究结果证明了孟加拉国合格病理学家进行的冰冻切片诊断的可靠性，并可为使用该服务以改善患者护理的外科医生建立信心提供证据。
