Answers are grounded in this paper and cite their sources.
Background/Aim: Urine cytology is non-invasive, but its performance for detecting high-grade urothelial carcinoma in patients with hematuria varies. This study aimed to assess the accuracy of urine cytology classified using the Paris System (second edition) against histopathologically confirmed high-grade invasive urothelial carcinoma (HGIUC) at Dr. Soetomo General Academic Hospital, Surabaya.Methods: This retrospective study reviewed 223 specimens from unique patients between May 2023 and May 2024. Histopathology was available for 160 patients; excluding 68 ineligible cases yielded 92 paired cases. Among the 80 paired cases with recorded timing, bladder histopathology was obtained 3–6 months after cytology and served as the reference standard (HGIUC positive; non-neoplastic tissue negative). Positivity thresholds were atypical urothelial cells (AUC)/HGUC (primary) and HGUC alone (secondary). Diagnostic indices and 95% confidence intervals (CIs) were calculated.Results: Among 92 paired cases, 75 were HGIUC-positive and 17 were reference-negative. At the AUC/HGUC threshold, there were 10 true-positive, 0 false-positive, 65 false-negative, and 17 true-negative results; sensitivity was 13.3% (95% CI, 6.6–23.2%), specificity 100% (80.5–100%), positive predictive value 100% (69.2–100%), negative predictive value 20.7% (12.6–31.1%), accuracy 29.3% (20.3–39.8%), and the negative likelihood ratio was 0.87 (0.79–1.00). The positive likelihood ratio was not estimable. At the HGUC-only threshold, sensitivity was 8.0% and specificity 100%. Including six unsatisfactory cytology specimens among 81 HGIUC cases as test failures reduced primary sensitivity to 12.3%.Conclusion: Cytology missed most HGIUC cases; negative findings cannot exclude malignancy. Specificity is imprecise given only 17 reference-negative cases, and the 3–6-month reference interval may affect the estimates. Cystoscopy and histopathology remain essential when bladder cancer is suspected.
No takes yet. Share an insight, caveat, or question.
Rahmanti et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: