Bile Exposure and Postoperative Cholangitis After Laparoscopic Resection of Choledochal Cysts: Clarifying Exposure, Controlling Confounding

We commend Yano and colleagues for evaluating whether intraoperative bile exposure (BE) affects recovery and complications after laparoscopic choledochal cyst resection with hepaticojejunostomy. In a multi-institutional series supervised by a single surgeon (n = 43), BE was associated with a higher incidence of postoperative cholangitis (33.3% vs. 4.0%) [1]. We offer methodological clarifications to strengthen inference and interpretation. A central issue is how BE is defined. The current composite mixes (i) early operative sequencing that prolongs exposure, (ii) iatrogenic bile duct injury, and (iii) “bile leakage immediately after surgery,” while postoperative leakage is also analyzed as an outcome [1]. This may introduce incorporation bias and inflate associations. A clearer primary definition—restricting BE to intraoperative contamination before closure and excluding postoperative leak—paired with a prespecified sensitivity analysis using the original composite would clarify whether BE independently relates to cholangitis. Importantly, surgical technique and laparoscopic experience evolved over the study period. Operative sequencing changed in ways that shortened bile exposure in later years [1]. Grouping cases by BE may therefore partially compare early- versus late-era surgery and different experience levels rather than BE alone. Interrupted time-series (segmented regression) would best disentangle protocol/era effects from BE itself [2]; if underpowered, modeling calendar year/era and center as covariates or fixed effects is a pragmatic alternative. Statistical handling can be strengthened. The analyses are largely univariate [1]. Multivariable logistic regression or propensity-score methods (e.g., inverse probability weighting) can balance age, cyst type—including IVa—cyst diameter, preoperative cholangitis, and center [3]. With small cells (e.g., 6/18 vs. 1/25) and zeros, Fisher's exact tests are preferable; if multiple endpoints are retained, a transparent family-wise or FDR approach is advisable. To sharpen clinical signals, separate mechanisms, and standardize outcomes. The BE construct blends prolonged exposure from early transection with true bile duct injury (and immediate postoperative leak). Reporting these pathways separately would be informative. Given variation in follow-up protocols and diagnostic criteria, standardized—ideally blinded—adjudication for cholangitis and related endpoints would reduce misclassification [1]. The follow-up horizon is short for late biliary endpoints. Mean surveillance of ~4–5 years is briefer than the decade-scale window over which anastomotic strictures and intrahepatic stones often emerge after pediatric choledochal cyst surgery [1]. We encourage extended follow-up with time-to-event methods (Kaplan–Meier or cumulative incidence) and more tempered causal language about “late” complications until ≥ 10-year data accrue [4]. Finally, two reporting discrepancies in the paper by Yano et al. [1] merit correction: their results text indicates later drain removal in the BE group (7.4 ± 2.4 vs. 12.7 ± 9.7 days; p = 0.030), whereas Table 2 in their publication shows the opposite direction and non-significance (9.7 ± 7.6 vs. 7.2 ± 3.4 days; p = 0.30). In the same paper, the Abstract lists postoperative cholangitis p = 0.003, whereas the main text reports p = 0.020. Harmonizing these would improve clarity. In sum, this series would be strengthened by (1) an incorporation-bias-free primary exposure definition with sensitivity analysis, (2) era/learning-curve adjustment (ideally via interrupted time-series, or minimally era and center covariates), (3) multivariable or propensity-weighted modeling with exact inference for sparse data, and (4) standardized adjudication with longer, survival-based follow-up. These steps would clarify whether bile exposure itself—and which mechanism—drives postoperative cholangitis. The authors declare no conflicts of interest.