We conducted a single-center retrospective cohort study. We used records from patients who underwent LC + LCBDE at the Affiliated Hospital of Youjiang Medical University for Nationalities. The operations took place between January 2018 and December 2025. The analytic database contained 202 eligible operations. It did not retain the number of patients screened or each reason for exclusion. The institutional ethics committee approved the study. All procedures followed the applicable ethical standards. All patients received study information and gave written informed consent.

We included patients aged 65 years or older. Imaging and liver tests had to support a diagnosis of common bile duct stones. Patients had to undergo LC + LCBDE with attempted stone extraction. The surgeon then used primary closure or T-tube drainage. The record had to document an intraoperative assessment of duct clearance by choledochoscopy, cholangiography, or the operative note. Core baseline and perioperative data also had to be available. We excluded biliary malignancy, a confirmed benign bile duct stricture, and obstruction from an ampullary lesion. We also excluded complex stones that required hepatectomy or biliary-enteric anastomosis. Conversion to open surgery or another major concurrent procedure also led to exclusion. For repeat admissions, we kept only the first eligible operation. We excluded records without key exposure or primary-outcome data. All 202 records contained the 13 PSM covariates and the reported perioperative outcomes. We did not impute missing values. The final cohort included 81 primary-closure patients and 121 T-tube patients.

We grouped patients by the final treatment of the choledochotomy. Direct closure was classified as primary closure. Closure around an external T-tube was classified as T-tube drainage. The surgeon removed stones and assessed clearance before choosing the closure method. Treatment was not randomized. The surgeon considered the bile duct wall, confidence in clearance, distal duct patency, biliary pressure, inflammatory edema, and the patient’s overall condition. We treated choledochoscopy as an intraoperative process measure. We did not use it as a baseline matching variable. Operative complexity and procedural progress may affect its use.

2.2.1. Primary Closure Group

After induction of general anesthesia, pneumoperitoneum was established and laparoscopic cholecystectomy and exposure of the common bile duct were performed routinely. A longitudinal incision was made on the anterior wall of the common bile duct. Stones were extracted using forceps, a retrieval basket, or choledochoscopy; lithotripsy and saline irrigation were used when necessary. After confirming the absence of obvious residual stones, distal duct patency, and acceptable blood supply and suturability of the bile duct wall, the choledochotomy was closed with appropriately sized absorbable sutures using a continuous or interrupted full-thickness technique. After confirming the absence of active bile leakage, a subhepatic drain was placed at the surgeon’s discretion.

2.2.2. T-Tube Drainage Group

Surgeons exposed the common bile duct and removed stones as described for the primary-closure group. They chose a silicone T-tube according to the duct diameter. The short limbs entered the proximal and distal ducts. The surgeon closed the choledochotomy around the tube with absorbable sutures. The team checked patency and looked for leakage. The long limb exited through the abdominal wall and was secured. A subhepatic drain was placed. After surgery, T-tube output, liver tests, symptoms, and imaging guided care. T-tube removal was not a discharge requirement. Patients who met the usual discharge criteria could leave with the tube in place. During follow-up, the team generally completed clamping and T-tube cholangiography before removal. When needed, surgeons treated residual stones through the mature tract.

Both groups received individualized antimicrobial treatment, analgesia, nutritional support, and fluid management. Early ambulation and oral intake were encouraged when clinically appropriate. Discharge criteria included stable vital signs, no uncontrolled infection or bile leakage, adequate oral intake, controllable pain, and no need for inpatient-level intervention.

Postoperative hospital stay was the primary outcome. We counted calendar days from the day after surgery until discharge. Secondary outcomes included operative time, blood loss, intraoperative stone clearance, choledochoscopy use, ICU admission, and time to first flatus. We also recorded surgical reintervention, ERCP, and unplanned readmission within 30 days. Intraoperative stone clearance meant that no definite stone required immediate further treatment at the end of exploration. The surgeon based this assessment on choledochoscopy, cholangiography, or the operative record. This measure differs from stones found later during follow-up.

Safety and follow-up outcomes included complications within 30 days, major complications, bile duct stricture, stones found within 6 months, and recurrent stones. We defined any postoperative complication as a departure from expected recovery at Clavien-Dindo grade I or higher. Major complications were grade III or higher [13]. The database recorded overall and major-complication indicators. It did not consistently record biliary, infectious, or T-tube-related subtypes. We could not report those details without reviewing the source records again. Bile duct stricture required compatible symptoms or biochemical abnormalities plus imaging or endoscopic confirmation. Stones found within 6 months may have been missed at surgery or identified later. Their number can therefore exceed the number of intraoperative clearance failures. Recurrent stones were identified 6 months or more after documented initial clearance.

Follow-up data were obtained from outpatient records, subsequent hospitalization records, and telephone interviews. In principle, symptoms, liver function, and abdominal ultrasonography were assessed 1 month after surgery, followed by evaluations at 3 and 6 months and annually thereafter. Computed tomography, magnetic resonance cholangiopancreatography, endoscopic ultrasonography, or ERCP was performed when fever, jaundice, or abdominal pain occurred. Follow-up was censored in June 2026.

We used SPSS version 27.0 and R version 4.4.3. We present continuous variables as median (25th percentile, 75th percentile). We present categorical variables as n (%). Before matching, we used the Mann-Whitney U test, chi-square test, or Fisher exact test, as appropriate. A binary logistic model estimated each patient’s propensity for primary closure. It included age, sex, common bile duct diameter, maximum stone diameter, stone number, BMI, WBC, CRP, TBIL, ALT, AST, gamma-glutamyl transferase, and serum amylase. These covariates had no missing values, so we used no imputation. We performed 1:1 nearest-neighbor matching without replacement. The absolute caliper was 0.1. We assessed balance mainly with absolute SMDs. Values below 0.10 indicated good balance [14]. We also inspected propensity-score overlap and within-pair distances. For the matched sample, we used paired Wilcoxon signed-rank tests for continuous outcomes. We used exact McNemar tests for binary outcomes [15]. We report median within-pair differences for continuous outcomes. We report paired risk differences for binary outcomes, calculated as primary closure minus T-tube drainage. We estimated 95% CIs using 20,000 paired bootstrap resamples. All tests were two-sided. We used P < 0.05 as the significance threshold. Analyses of secondary outcomes were exploratory and had no multiplicity adjustment.

The analytic database included 202 eligible older patients. It contained 81 primary-closure patients and 121 T-tube patients. None of the 13 matching variables was missing. Matching produced 73 pairs. Eight primary-closure patients and 48 T-tube patients remained unmatched (Figure 1). The largest absolute SMD fell from 0.364 to 0.134. Twelve of the 13 post-matching SMDs were below 0.10. TBIL still showed mild imbalance (SMD = 0.134). In the matched sample, propensity scores ranged from 0.095 to 0.690 for primary closure. They ranged from 0.100 to 0.709 for T-tube drainage. The mean within-pair absolute difference was 0.015, and the largest was 0.089 (Table 1 and Figure 2).

Figure 1. Patient flow and propensity score matching.

Table 1. Baseline characteristics before and after propensity score matching.

Note: Continuous variables are reported as median (25th percentile, 75th percentile). Sex is reported as n (%). SMD, standardized mean difference. An absolute SMD below 0.10 indicates good balance. TBIL showed mild residual imbalance after matching (SMD = 0.134).

Figure 2. Covariate balance before and after propensity score matching. The dashed line marks an absolute SMD of 0.10.

In the matched analysis, primary closure was linked to shorter operating time (P = 0.026) and lower blood loss (P = 0.037). It was also linked to earlier flatus (P = 0.011) and a shorter hospital stay (P < 0.001). The median within-pair difference in hospital stay was −3.0 days (95% CI, −5.0 to −2.0). Intraoperative stone clearance was 97.3% with primary closure and 95.9% with T-tube drainage. The paired risk difference was 1.4 percentage points (95% CI, −4.1 to 6.8; P = 1.000). Choledochoscopy use, ICU admission, 30-day reintervention, ERCP, and readmission did not differ between groups (Table 2).

Table 2. Intraoperative and 30-day postoperative outcomes after propensity score matching.

Note: For continuous outcomes, effects are median within-pair differences. For binary outcomes, effects are paired risk differences in percentage points. Each risk difference is primary closure minus T-tube drainage. We used 20,000 paired bootstrap resamples to calculate CIs. P values come from paired Wilcoxon signed-rank or exact McNemar tests.

Overall 30-day complications occurred in 11 primary-closure patients (15.1%). They occurred in 21 T-tube patients (28.8%). The paired comparison was not significant (P = 0.087). The paired risk difference was −13.7 percentage points (95% CI, −27.4 to 0.0). Major complications occurred in 1 patient (1.4%) after primary closure and 3 patients (4.1%) after T-tube drainage (P = 0.625). Stones found within 6 months occurred in 5 patients (6.8%) and 12 patients (16.4%), respectively (P = 0.092). Recurrent stones at 6 months or later occurred in 4 patients (5.5%) in each group (P = 1.000) (Table 3). The source dataset did not consistently record biliary, infectious, or tube-related event types. We did not reconstruct these events after the analysis.

Table 3. Postoperative complications and follow-up outcomes after propensity score matching.

Note: Any complication was Clavien-Dindo grade I or higher. Major complications were grade III or higher. We analyzed stones found within 6 months separately from intraoperative clearance. They may include stones missed during surgery or found later. Effects are paired risk differences in percentage points. CIs come from paired bootstrap resampling. P values come from exact McNemar tests.