操作者經驗對提升上泌尿道上皮癌視覺診斷準確性之研究

鄭苑伶¹、卓育慶¹、陳進利¹、楊明昕¹、高建璋¹、

曹智惟¹、蒙恩¹、孫光煥¹、于大雄¹、丁慧恭¹

The Trained Eye Speaks: Elevating Visual Diagnostic Accuracy in Upper Tract Urothelial Carcinoma

Yuan-Ling Cheng¹, Yu-Cing Jhuo¹, Jin-Li Chen¹, Ming-Hsin Yang¹, Chien-Chang Kao¹, 

Chi-Wei Tsao¹, En Meng¹, Guang-Huan Sun¹, Dah-Shyong Yu¹, Hui-Kung Ting¹

¹Division of Urology, Department of Surgery, Tri-Service General Hospital, National

Defense Medical University, Taipei 114024, Taiwan

Purpose:

  Ureteroscopic visual impression plays a key role in diagnosing upper tract urothelial carcinoma (UTUC), but its accuracy may vary with the operator’s experience. This study aimed to evaluate the diagnostic performance of visual impression and assess whether intraoperative urine cytology offers additional diagnostic value in selected clinical scenarios.

Materials and Methods:

  We retrospectively analyzed 81 patients who underwent diagnostic ureteroscopy with biopsy for suspected UTUC at a single tertiary center between February 2021 and December 2023. Patients without biopsy results or adequate endoscopic images were excluded. Visual impressions (“suspected UC” vs “not UC”) were recorded intraoperatively and stratified by operator experience (attending vs. senior resident). Intraoperative cytology was performed selectively at the discretion of the operating urologist. Final surgical pathology was available only for patients who subsequently underwent definitive surgery. Diagnostic performance was assessed using sensitivity, specificity, predictive values, accuracy, and Fisher’s exact test.

Results:

  Visual impression achieved a sensitivity of 98.2%, specificity of 52.0%, positive predictive value of 82.1%, negative predictive value of 92.9%, and overall accuracy of 84.0%. Among the 20 procedures performed by the attending physician, visual and biopsy diagnoses were 100% concordant. In comparison, the resident group demonstrated a concordance rate of 78.7% (48/61), showing a statistically significant difference (p = 0.031).
 Among patients visually assessed as “not UC,” cytology was evaluated for its potential to reduce missed diagnoses. However, it failed to detect the only missed UC case in this group, and no cytology-positive results were recorded. This suggests limited diagnostic utility of cytology in low-suspicion settings (p = 1.0).
 Similarly, in cases visually assessed as “suspected UC,” cytology did not significantly differentiate between final UC and non-UC cases (p = 0.648), further suggesting limited utility in this context.

Conclusion:

  Operator experience significantly improves the diagnostic accuracy of visual impression during ureteroscopy. Intraoperative cytology appears to offer minimal additional value, particularly in cases with low clinical suspicion or when performed by experienced urologists. Emphasizing training and experience may improve diagnostic confidence and reduce the need for adjunctive testing.