A Retrospective Analysis of Diagnostic Accuracy of Frozen Sections Compared to Permanent Sections: A 5 Years Study in A Single Tertiary Care Hospital In South Delhi

Introduction: Intraoperative frozen section pathology consultation has long been applied to verify/categorize a lesion, evaluate surgical margins, determine the organ of origin, and assess tissue adequacy for further diagnostic studies. Correlation of intraoperative frozen section diagnosis with final diagnosis can be an important component of an institution's quality assurance process. The aim of this study was to evaluate the concordance and discordance between frozen section and histopathology diagnosis and to assess the reliability and challenges associated with this rapid diagnostic method. Materials and Methods: A retrospective review of all intraoperative consultation in the Frozen Section and Histopathology Department of Pathology, Hamdard Institute of Medical Sciences and Research and Hospital, New Delhi from 1st January 2018 to 31st December 2022 was performed. During this five-year study period, a total of 185 patients from surgical departments underwent intraoperative consultation. The final diagnosis from the surgical pathology report was compared to the intraoperative consultation diagnosis. Results: Frozen section diagnoses were evaluated on all the 185 cases, with 88 cases identified as positive for malignancy and 97 cases as negative for malignancy. On subsequent histopathology paraffin sections, 92 cases were positive and 93 negative for malignancy. The comparison revealed an approximate result concordance of 97.0% and a discordance of 3.0%. Conclusion: Frozen section intraoperative consultation is a reliable rapid method for patient diagnosis and management. Monitoring diagnostic accuracy of intraoperative consultation provides substantial information towards the causes of the errors. Our study suggests that more accurate sampling, and knowledge about clinical history and presentation can reduce the limitation and increase the diagnostic accuracy thus avoiding technical errors and reducing result discordance.