Dynamic Telepatholgy-Assisted Review of Previously Misdiagnosed Frozen Sections Shows Increased Accuracy over Single Pathologist-Rendered Diagnoses
Anand Shah, Silvat Sheikh-Fayyaz, Tawfiqul Bhuiya, Nora J Morgenstern. Hofstra North Shore-LIJ School of Medicine, Lake Success, NY
Background: Dynamic telepathology is a branch of medical telecommunications in which live video feed is transmitted in real-time from a primary location to a remote site. Our institution recently implemented this technology to transmit images of frozen sections (FZ) from two of our tertiary hospitals to our department, which has recently relocated to a centralized off-site facility, in order to assist with difficult cases in real-time. The purpose of this study is to determine whether the use of telepathology can improve FZ diagnostic accuracy within our institution.
Design: Institutional cases in which major errors were committed in original FZ diagnoses (false positives, false negatives and significant misclassification with potential for impacting patient care) over the last five years were collected. These slides were shown via telepathology to either a senior pathologist or a subspecialty service chief, as though they were real-time FZ. Original diagnoses were kept blind to both the host and the remote consultant pathologist. Telepathology-assisted diagnoses were then compared to the original FZ and final sign-out diagnoses in order to assess changes in accuracy.
Results: Forty cases were reviewed(14 pulmonary, 6 gynecologic, 6 genitourinary, 5 head and neck, 2 dermatology, 2 gastrointestinal, 2 bone & soft tissue, 1 neurology, 1 liver, 1 gallbladder). Of these, 26(65%) had improved accuracy with telepathology. Use of telepathology corrected 15 false positives and 10 false negatives. However, telepathology was unable to avoid 3 false positives and 5 false negatives. Only 1 of 5 misclassified diagnoses was improved by telepathology. Three cases (2 false negatives, 1 misclassification) were deferred by telepathology, due to suboptimal slide preparation.
|Original FZ diagnosis||Cases with improved accuracy using telepathology-assistance (%)||Cases deferred by remote telepathologist (%)|
|False Positives||18||15 (83)||0 (0)|
|False Negative||17||10 (58)||2 (12)|
|Misclassification||5||1 (20)||1 (20)|
|Total Cases||40||26 (65)||3 (7.5)|