Article Sidebar
Main Article Content
Abstract
Background: Distinguishing benign from malignant extrahepatic obstructive jaundice before surgery determines whether patients proceed to stone clearance or to oncological staging and resection.
Objective: To evaluate the diagnostic accuracy of admission serum bilirubin fractions and preoperative magnetic resonance cholangiopancreatography (MRCP) for differentiating choledocholithiasis from pancreatobiliary malignancy.
Methods: This single-center retrospective study, reported according to STARD 2015, included 50 consecutive patients (71 screened) treated at Dr. Mohammad Hoesin Central General Hospital, Palembang, Indonesia, in 2024–2025. Admission bilirubin fractions and 1.5-T MRCP were compared with operative or biopsy histopathology using receiver operating characteristic analysis, Wilson 95% confidence intervals (CIs), likelihood ratios, and exploratory Firth logistic regression.
Results: Histopathology showed choledocholithiasis in 39 patients (78.0%) and pancreatobiliary carcinoma in 11 (22.0%). Direct bilirubin (median 13.7 vs 8.7 mg/dL; p = 0.003), total bilirubin (20.17 ± 6.78 vs 13.02 ± 7.64 mg/dL; p = 0.007), and common bile duct diameter (26.00 ± 8.60 vs 18.49 ± 9.06 mm; p = 0.018) were greater in malignancy. Direct bilirubin ≥9.65 mg/dL (AUC 0.797) and total bilirubin ≥14.3 mg/dL (AUC 0.773) each had a sensitivity of 90.9% (95% CI 62.3–98.4%), with specificities of 56.4% and 61.5% and negative predictive values of 95.7% and 96.0%. For MRCP, sensitivity was 63.6% and specificity 94.9%; accuracy reached 88.0%, LR+ was 12.4, and the adjusted odds ratio was 21.65 (95% CI 3.12–150.24).
Conclusion: Admission bilirubin flagged most cancers but also many stone patients, whereas MRCP rarely mislabeled stones but missed some tumors. A sequential pathway using both needs external validation.
Keywords
Article Details
Authors retain copyright and grant Sriwijaya Journal of Surgery the non-exclusive right of first publication.
The journal’s publishing license is Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0). Reuse requires appropriate attribution, must be non-commercial, and adaptations must be shared under the same license.
Authors may deposit their submitted manuscript, accepted manuscript, and published version without an embargo, with acknowledgement of the journal and a link to the version of record.
See the Copyright and Licensing Policy for details.