Clinical outcomes of immediate, 24-h, and 24–72-h laparoscopic cholecystectomy after endoscopic stone clearance: a matched cohort analysis

Abstract

Objective

This study aimed to systematically compare the perioperative outcomes of patients undergoing laparoscopic cholecystectomy (LC) at three specific intervals following endoscopic retrograde cholangiopancreatography (ERCP) for choledocholithiasis: immediate (3–6 h), 24-h, and 24–72-h. A propensity score-matched (PSM) cohort analysis was utilized to minimize potential selection bias.

Methods

A retrospective analysis was conducted on 200 patients treated between January 2021 and December 2025. Patients were categorized into Immediate (3–6 h, n = 45), 24-h (n = 49), and 24–72-h (n = 106) LC groups. To control for confounders, sequential 1:1 nearest-neighbor propensity score matching was performed based on demographics, comorbidities, clinical presentation, stone characteristics, and preoperative laboratory values, yielding 84 well-matched patients (n = 25, 21, and 38 per group, respectively). Perioperative outcomes, postoperative complications, laboratory changes (Δ), and length of stay, were compared among the matched groups.

Results

After PSM, baseline characteristics were similar across groups (P > 0.05). The Immediate group had significantly shorter operative time (P = 0.043), Calot’s triangle dissection time (P < 0.001), and less intraoperative blood loss (P = 0.036) versus the 24–72‑h group. On postoperative day 1, the Immediate group showed lower absolute serum amylase (AMY, P = 0.006), alanine aminotransferase (ALT, P = 0.007), and γ‑glutamyl transferase (GGT, P = 0.016). However, change‑from‑baseline (ΔAMY, ΔALT, ΔGGT) did not differ significantly (all P > 0.05), indicating equivalent physiological recovery. The Immediate group had a lower incidence of hyperamylasemia (P = 0.019), while acute pancreatitis rates were similar (P = 0.481). Total hospital stay was shorter in the Immediate group (P = 0.002), but postoperative stay did not differ significantly (P = 0.081).

Conclusion

Immediate LC (3–6 h) following ERCP is a safe and efficient strategy. This “extreme-early window” facilitates anatomical dissection before inflammatory changes become restrictive. The significantly reduced total hospital stay stems from clinical process optimization rather than accelerated physiological recovery. Therefore, immediate LC should be recommended as the preferred clinical pathway for eligible patients.

Choledocholithiasis, the presence of gallstones within the common bile duct, is a prevalent and potentially life-threatening manifestation of biliary disease, frequently leading to severe complications such as obstructive jaundice, acute cholangitis, or biliary pancreatitis [1–3]. To mitigate the risk of recurrent biliary events and address the primary source of stones, current clinical guidelines advocate for a sequential, two-stage treatment paradigm: endoscopic retrograde cholangiopancreatography (ERCP) for ductal clearance followed by laparoscopic cholecystectomy (LC) [4, 5]. While this strategy is established as the standard of care, its successful implementation depends on optimizing the sequence and timing of these procedures to balance clinical safety with healthcare efficiency.

Despite the established efficacy of this sequential approach, the optimal timing for performing LC following ERCP remains a matter of significant clinical debate. Historically, conventional surgical wisdom advocated for a delay of several days to weeks, based on the rationale that ERCP-induced inflammation and potential subclinical pancreatitis might obscure anatomical planes, thereby increasing surgical difficulty and the risk of ductal injury [6–8]. Conversely, emerging evidence and Enhanced Recovery After Surgery (ERAS) protocols suggest that early intervention may mitigate the risk of recurrent biliary events and reduce the total length of hospital stay [9–12]. Nevertheless, the definition of "early" intervention remains inconsistent, ranging from within 24 h to one week post-ERCP, and high-level evidence specifically comparing immediate (within hours) versus delayed intervention within the critical 72-h window is remarkably sparse.

Therefore, the present study was designed to systematically evaluate and compare the perioperative outcomes of LC at three distinct intervals following successful ERCP: immediate (3–6 h), 24-h, and 24–72-h. By utilizing a propensity score-matched (PSM) analysis to minimize potential selection bias, the investigation focused on whether the extreme-early window following ERCP provides a superior surgical environment for anatomical dissection. These findings are intended to provide a robust clinical basis for standardizing the timing of sequential treatment in patients with choledocholithiasis.

Materials and methods

Study setting

This retrospective cohort study has been reported in line with the STROCSS (Strengthening the Reporting of Cohort, Cross-sectional and Case–control Studies in Surgery) guidelines [13]. The study was conducted at Shijiazhuang People's Hospital (Shijiazhuang, China). Consecutive patients who underwent ERCP followed by LC for choledocholithiasis between January 2021 and December 2025 were identified from the institutional database. A total of 206 patients were initially screened for eligibility. After excluding 4 patients with incomplete medical records and 2 patients who were lost to follow-up after discharge, the final study cohort comprised 200 patients for analysis. The study protocol was reviewed and approved by the Medical Research Ethics Committee of Shijiazhuang People's Hospital (Approval No. 2026-011). Due to the retrospective nature of the study, the requirement for informed consent was waived. All methods were conducted in accordance with the relevant guidelines and regulations.

Further details are provided in Fig. 1.

Fig. 1.

Flowchart of the study

Inclusion and exclusion criteria

Inclusion criteria:

• Patients aged 18–85 years;

• Diagnosed with choledocholithiasis confirmed by MRCP endoscopic ultrasound, or intraoperative findings during ERCP;

• Underwent successful ERCP with biliary sphincterotomy and stone clearance (confirmed by occlusion cholangiogram);

• Subsequently underwent laparoscopic cholecystectomy during the same hospitalization;

• Complete clinical, laboratory, and operative data available.

Exclusion criteria:

• Prior history of any abdominal surgery;

• Acute pancreatitis of any severity at the time of hospital admission (including mild biliary pancreatitis);

• Severe acute cholangitis (Tokyo Guidelines 2018 Grade III);

• Child–Pugh class B or C liver function;

• Pregnancy or lactation;

• Patients with a history of malignant tumors or those with immunocompromised status, defined as chronic corticosteroid use equivalent to prednisone ≥ 10 mg/day for > 3 months, active chemotherapy, radiation therapy, or known immunodeficiency disorders.

Study groups and timing definition

Patients were categorized into three groups based on the time interval between the completion of ERCP and the start of laparoscopic cholecystectomy. The upper limit of 72 h for the delayed group was selected based on institutional practice patterns and aligns with findings from previous studies, which have demonstrated that performing LC within 72 h after ERCP is feasible, safe, and associated with improved outcomes, including reduced hospital stay and lower complication rates [14, 15].

• Immediate LC Group: Patients who underwent LC within 3 to 6 h after ERCP completion. This interval was chosen to represent surgery performed on the same day after a brief observation period, allowing for initial post-ERCP monitoring while avoiding an overnight waiting period.

• 24-h Delayed LC Group: Patients who underwent LC between > 6 and 24 h after ERCP completion.

• 24–72-h Delayed LC Group: Patients who underwent LC between > 24 and 72 h after ERCP completion.

The median time intervals from ERCP to LC for each group were as follows: Immediate group, 4.6 h (range 3.1–5.5 h); 24-h Delayed group, 14.3 h (range 7.2–21.5 h); 24–72-h Delayed group, 47.5 h (range 28.5–65.0 h).

Surgical equipment

Laparoscopic cholecystectomy was performed using the STORZ laparoscopic system (Karl Storz, Tuttlingen, Germany). The system included a high-definition camera, xenon light source, and CO₂ insufflator. Standard laparoscopic instruments (graspers, dissectors, clip appliers) were used for all procedures. All surgeries were performed by the same team of experienced hepatobiliary surgeons to ensure technical consistency.

Surgical procedure

All ERCP procedures were performed by experienced endoscopists using a standard side-viewing duodenoscope. Following successful stone extraction with a retrieval balloon or basket, a temporary plastic biliary stent or nasobiliary drainage tube was placed based on the difficulty of the procedure and the presence of residual sludge.

Subsequent LC was performed under general anesthesia using a standard four-port technique. Special attention was directed toward the dissection of Calot’s triangle. For the Immediate group (3–6 h post-ERCP), the dissection focused on identifying the cystic duct and artery within the early edematous window, where anatomical planes remained relatively distinct. For the delayed groups, any inflammatory adhesions encountered were managed using blunt or sharp dissection to achieve a "critical view of safety" before clipping and dividing the cystic structures. Intraoperative cholangiography was performed only if biliary anatomy was unclear or if synchronous stones were suspected.

Perioperative care protocol

Given the minimally invasive nature of LC, postoperative recovery is generally rapid. To optimize patient outcomes while maintaining safety, a structured perioperative care pathway was followed for all patients. An evidence-based Enhanced Recovery After Surgery (ERAS) protocol was routinely implemented, incorporating multimodal strategies for postoperative management [16, 17]. Vital signs were continuously monitored until hemodynamic and clinical stability were achieved. Intravenous fluids, electrolytes, and caloric support were administered as needed to maintain metabolic balance, alongside appropriate analgesic therapy to ensure adequate pain control. Patients were encouraged to ambulate as early as tolerated to reduce the risk of thromboembolic events and promote functional recovery.

For patients in the Immediate LC group, a standardized post-ERCP observation protocol was implemented. Serum amylase was measured approximately 2 h after ERCP completion, prior to proceeding with laparoscopic cholecystectomy. In all patients in this group, amylase levels remained within normal limits or showed only mild elevation well below three times the upper limit of normal, and no patient developed new-onset abdominal pain suggestive of pancreatitis. The safety of performing LC shortly after ERCP is supported by studies demonstrating the feasibility of a single-anesthetic “rendezvous” approach, in which ERCP and LC are performed sequentially during the same operative session [18, 19].

On postoperative day 1 (POD 1), routine laboratory tests—including complete blood count, renal and liver function panels—were routinely obtained between 6:00 and 8:30 AM to assess early recovery and detect any occult complications.

Following hospital discharge, a structured follow-up was conducted by designated personnel at one week post-surgery, with subsequent outpatient visits scheduled as clinically indicated.

Observational indicators

Preoperative indicators

Demographic and clinical characteristics included age, gender, and body mass index (BMI). Comorbidities such as hypertension, diabetes mellitus, and coronary artery disease (CAD) were documented. Clinical presentation at admission was categorized into the following types: asymptomatic/biliary colic only, obstructive jaundice, and acute cholangitis of mild (Grade I) or moderate (Grade II) severity according to the Tokyo Guidelines 2018. Gallstone-related parameters included the size and number of gallstones, as well as the diameter of the common bile duct (CBD) and the size and number of CBD stones. Preoperative laboratory assessments were obtained from the last blood test performed within 24 h prior to ERCP, including total bilirubin (TBiL), direct bilirubin (DBiL), gamma-glutamyl transferase (GGT), alkaline phosphatase (ALP), alanine aminotransferase (ALT), aspartate aminotransferase (AST), albumin (ALB), white blood cell count (WBC), and amylase (AMY). These baseline indicators were used to assess group comparability prior to propensity score matching.

Intraoperative indicators

Intraoperative indicators included operative time, Calot’s triangle dissection time, estimated blood loss, and the rate of intraoperative conversion to open surgery. Operative time was defined as the duration from the initial skin incision to the completion of skin closure. Calot’s triangle dissection time was recorded from the beginning of Calot’s triangle exposure until the cystic duct and cystic artery were clearly identified and divided. Estimated blood loss was determined by calculating the volume of fluid in the suction canister and the increase in weight of the surgical swabs.

Postoperative indicators

Postoperative indicators included the time to first flatus, stone clearance rate, postoperative complications, laboratory tests, and hospital stay, the latter of which was categorized into total hospital stay and postoperative length of stay. Postoperative complications were specifically monitored and recorded, including fever, acute pancreatitis, hyperamylasemia, and bile leak.

Laboratory recovery was evaluated using both absolute serum levels on POD 1 and the change-from-baseline (Δ) values. The Δ values for key parameters, including ΔAMY, ΔALT, ΔAST, and ΔGGT, were calculated by subtracting the preoperative baseline value from the POD 1 value. This approach was employed to assess the intrinsic rate of physiological recovery while accounting for preoperative variations in liver and pancreatic functions.

Statistical analysis

Statistical analyses were performed using SPSS 25.0 software. Continuous variables were tested for normality using the Shapiro–Wilk test. Normally distributed continuous data were presented as mean ± standard deviation ($\overline{\text{x}}\pm \text{s}$) and compared among the three groups using one-way analysis of variance (ANOVA), with post-hoc pairwise comparisons performed using the least significant difference (LSD) test. Non-normally distributed continuous data were expressed as median with interquartile range [M (P25, P75)] and analyzed using the Kruskal–Wallis H test, followed by pairwise comparisons using the Mann–Whitney U test with Bonferroni correction. Categorical variables were presented as frequencies and percentages [n (%)] and compared using the chi-square test or Fisher's exact test, as appropriate.

To minimize selection bias, propensity score matching (PSM) was performed using a logistic regression model. Covariates included demographics (age, gender, BMI), comorbidities (hypertension, diabetes, CAD), clinical presentation, stone characteristics, and a panel of preoperative laboratory values (TBiL, DBiL, GGT, ALP, ALT, AST, ALB, WBC, and AMY).

Given the three-group comparison, a sequential pairwise matching strategy was employed with the 24–72-h Delayed LC Group serving as the reference group. Patients in the Immediate LC Group were first matched 1:1 to patients in the 24–72-h Delayed LC Group using nearest-neighbor matching with a caliper width of 0.2 times the standard deviation of the propensity score, yielding 25 matched pairs. Subsequently, patients in the 24-h Delayed LC Group were matched 1:1 to patients in the 24–72-h Delayed LC Group using the same algorithm, yielding 21 matched pairs. After excluding eight duplicate patients from the 24–72-h Delayed LC Group who were matched to both groups, the final matched cohort comprised 25 patients in the Immediate LC Group, 21 patients in the 24-h Delayed LC Group, and 38 patients in the 24–72-h Delayed LC Group.

Standardized mean differences were used to assess the balance of covariates before and after matching, with a difference of < 0.1 indicating adequate balance.

All statistical tests were two-tailed, and a P-value < 0.05 was considered statistically significant.

Results

Baseline characteristics of patients

A total of 200 patients who underwent ERCP followed by LC for choledocholithiasis were included in this study. Based on the timing of LC after ERCP, patients were categorized into three groups: the Immediate LC Group (n = 45), the 24-h Delayed LC Group (n = 49), and the 24–72-h Delayed LC Group (n = 106).

The baseline demographic and clinical characteristics of the three groups before propensity score matching are presented in Table 1. No statistically significant differences were observed among the groups in terms of gender distribution, age, BMI, clinical presentation, or the prevalence of hypertension, diabetes, and coronary artery disease (all P > 0.05). Gallstone characteristics, including size and number of gallstones, diameter of the common bile duct, and size and number of CBD stones, also showed no statistically significant differences across the three groups (all P > 0.05). Preoperative laboratory parameters, including TBiL, DBiL, GGT, ALP, ALT, AST, ALB, WBC, and AMY, showed no significant differences among the groups (all P > 0.05). These findings indicate that the three groups were generally well-balanced prior to matching, although minor numerical variations existed.

Table 1.

General information of patients

	Immediate LC Group (n = 45)	24-h Delayed LC Group (n = 49)	24–72-h Delayed LC Group (n = 106)	Statistical test values	P value
Gender [n (%)]				1.514	0.469
Men	20 (44.44%)	28 (57.14%)	54 (50.94%)
Women	25 (55.56%)	21 (42.86%)	52 (49.06%)
Age ($\overline{\text{x}}\pm \text{s}$,years)	58.44 ± 13.66	60.06 ± 14.58	57.33 ± 16.37	0.533	0.588
BMI ($\overline{\text{x}}\pm \text{s}$,kg/m2)	25.34 ± 3.18	25.21 ± 4.11	25.43 ± 4.82	0.045	0.956
Hypertension [n (%)]				1.451	0.484
Yes	13 (28.89%)	19 (38.78%)	41 (38.68%)
No	32 (71.11%)	30 (61.22%)	65 (61.32%)
Diabetes [n (%)]				0.469	0.791
Yes	6 (13.33%)	9 (18.37%)	18 (16.98%)
No	39 (86.67%)	40 (81.63%)	88 (83.02%)
CAD [n (%)]				0.607	0.738
Yes	8 (17.78%)	8 (16.33%)	14 (13.21%)
No	37 (82.22%)	41 (83.67%)	92 (86.79%)
Clinical presentation [n (%)]
Asymptomatic/Biliary colic only	24 (53.33%)	31 (63.27%)	66 (62.26%)	1.262	0.532
Obstructive jaundice	7 (15.56%)	4 (8.16%)	11 (10.38%)	1.388	0.498
Acute cholangitis
Grade I (mild)	3 (6.67%)	4 (8.16%)	8 (7.55%)	0.159	1.000
Grade II (moderate)	11 (24.44%)	10 (20.41%)	21 (19.81%)	0.422	0.810
Size of gallstone [M(P25,P75),cm]	0.60 (0.40,0.90)	0.70 (0.40,0.80)	0.70 (0.50,1.0250)	4.771	0.092
Number of gallstones [n (%)]				0.330	0.848
Single	4 (8.89%)	3 (6.12%)	7 (6.60%)
Multiple	41 (91.11%)	46 (93.88%)	99 (93.40%)
Diameter of CBD [M(P25,P75),cm]	1.00 (0.80,1.20)	1.00 (0.80,1.20)	1.00 (0.80,1.20)	0.044	0.978
Size of CBD stone [M(P25,P75),cm]	0.60 (0.30,0.80)	0.60 (0.40,0.80)	0.60 (0.40,0.80)	1.682	0.431
Number of CBD stones [n (%)]				5.578	0.061
Single	9 (20.00%)	9 (18.37%)	36 (33.96%)
Multiple	36 (80.00%)	40 (81.63%)	70 (66.04%)
TBiL ($\overline{\text{x}}\pm \text{s}$,μmo/Ll)	49.68 ± 44.64	47.49 ± 61.81	39.86 ± 43.61	0.807	0.448
DBiL ($\overline{\text{x}}\pm \text{s}$,μmo/Ll)	26.07 ± 30.78	23.88 ± 35.74	19.32 ± 27.73	0.900	0.408
GGT ($\overline{\text{x}}\pm \text{s}$,U/L)	344.00 ± 328.16	291.88 ± 305.40	321.84 ± 293.84	0.350	0.705
ALP ($\overline{\text{x}}\pm \text{s}$,U/L)	133.13 ± 75.71	146.33 ± 108.75	154.63 ± 112.06	0.511	0.601
ALT ($\overline{\text{x}}\pm \text{s}$,U/L)	205.26 ± 226.37	179.27 ± 123.20	204.96 ± 221.95	0.298	0.743
AST ($\overline{\text{x}}\pm \text{s}$,U/L)	154.27 ± 166.90	88.88 ± 135.06	146.91 ± 206.99	2.028	0.134
ALB ($\overline{\text{x}}\pm \text{s}$,g/L)	39.84 ± 4.25	40.71 ± 9.61	39.56 ± 4.28	0.610	0.544
WBC ($\overline{\text{x}}\pm \text{s}$,109/L)	8.22 ± 4.74	7.52 ± 3.80	6.80 ± 3.82	2.610	0.076
AMY ($\overline{\text{x}}\pm \text{s}$,U/L)	97.13 ± 194.59	79.12 ± 61.02	66.82 ± 40.61	1.436	0.240

Following 1:1 nearest-neighbor matching, a matched cohort of 84 patients was generated, comprising 25 patients in the Immediate LC Group, 21 in the 24-h Delayed LC Group, and 38 in the 24–72-h Delayed LC Group. As presented in Table 2, all baseline characteristics remained well-balanced across the matched groups (all P > 0.05). Standardized mean differences for all covariates were less than 0.1, confirming that the matched cohort provided a robust basis for comparing perioperative outcomes without the interference of baseline imbalances.

Table 2.

Baseline characteristics of patients after propensity score matching

	Immediate LC group (n = 25)	24-h delayed LC group (n = 21)	24–72-h delayed LC group (n = 38)	Statistical test values	P value
Gender [n (%)]				0.912	0.634
Men	13 (52.00%)	8 (38.10%)	18 (47.37%)
Women	12 (48.00%)	13 (61.90%)	20 (52.63%)
Age ($\overline{\text{x}}\pm \text{s}$,years)	56.96 ± 13.00	54.24 ± 15.91	60.00 ± 14.23	1.137	0.326
BMI ($\overline{\text{x}}\pm \text{s}$,kg/m2)	25.83 ± 3.50	24.91 ± 4.19	25.29 ± 3.25	0.388	0.680
Hypertension [n (%)]				0.198	0.906
Yes	8 (32.00%)	6 (28.57%)	13 (34.21%)
No	17 (68.00%)	15 (71.43%)	25 (65.79%)
Diabetes [n (%)]				2.334	0.344
Yes	8 (32.00%)	3 (14.29%)	7 (18.42%)
No	17 (68.00%)	18 (85.71%)	31 (81.58%)
CAD [n (%)]				1.202	0.530
Yes	5 (20.00%)	3 (14.29%)	4 (10.53%)
No	20 (80.00%)	19 (85.71%)	34 (89.47%)
Clinical presentation [n (%)]
Asymptomatic/Biliary colic only	13 (52.00%)	12 (57.14%)	20 (52.63%)	0.146	0.930
Obstructive jaundice	4 (16.00%)	3 (14.29%)	6 (15.79%)	0.124	1.000
Acute cholangitis
Grade I (mild)	3 (12.00%)	1 (4.76%)	3 (7.89%)	0.815	0.780
Grade II (moderate)	5 (20.00%)	5 (23.81%)	9 (23.69%)	0.201	1.000
Size of gallstone [M(P25,P75),cm]	0.70 (0.40,1.10)	0.70 (0.55,0.85)	0.70 (0.40,1.00)	0.180	0.914
Number of gallstones [n (%)]				1.581	0.529
Single	1 (4.00%)	3 (14.29%)	3 (7.89%)
Multiple	24 (96.00%)	18 (85.71%)	36 (92.11%)
Diameter of CBD [M(P25,P75),cm]	1.00 (0.80,1.15)	1.00 (0.80,1.20)	1.00 (0.80,1.0250)	1.701	0.427
Size of CBD stone [M(P25,P75),cm]	0.50 (0.30,0.80)	0.60 (0.55,1.00)	0.60 (0.30,0.80)	1.972	0.373
Number of CBD stones [n (%)]				0.273	0.857
Single	6 (24.00%)	5 (23.81%)	11 (28.95%)
Multiple	19 (76.00%)	16 (76.19%)	27 (71.05%)
TBiL ($\overline{\text{x}}\pm \text{s}$,μmol)	42.54 ± 37.90	43.24 ± 39.29	52.83 ± 62.10	0.404	0.669
DBiL ($\overline{\text{x}}\pm \text{s}$,μmol)	20.63 ± 24.92	21.63 ± 25.08	27.79 ± 39.46	0.451	0.639
GGT ($\overline{\text{x}}\pm \text{s}$,U/L)	270.56 ± 307.92	357.00 ± 372.24	353.84 ± 328.13	0.562	0.572
ALP ($\overline{\text{x}}\pm \text{s}$,U/L)	136.96 ± 88.90	173.90 ± 143.76	160.53 ± 97.63	0.700	0.500
ALT ($\overline{\text{x}}\pm \text{s}$,U/L)	192.49 ± 239.41	181.71 ± 112.50	205.13 ± 213.33	0.095	0.910
AST ($\overline{\text{x}}\pm \text{s}$,U/L)	137.67 ± 140.06	97.71 ± 147.79	134.73 ± 154.92	0.486	0.617
ALB ($\overline{\text{x}}\pm \text{s}$,g/L)	39.64 ± 4.38	38.43 ± 5.58	39.56 ± 4.62	0.459	0.633
WBC ($\overline{\text{x}}\pm \text{s}$,109/L)	7.21 ± 2.98	6.89 ± 2.72	7.08 ± 3.37	0.059	0.943
AMY ($\overline{\text{x}}\pm \text{s}$,U/L)	58.80 ± 34.57	64.71 ± 35.44	71.79 ± 49.51	0.729	0.485

Intraoperative outcomes

Before propensity score matching

In the unmatched cohort, as shown in Table 3, the Immediate LC Group had a mean operative time of 106.31 ± 46.03 min, which was significantly shorter than that of the 24-h Delayed LC Group (135.02 ± 55.52 min) and the 24–72-h Delayed LC Group (140.22 ± 72.45 min). The overall difference among the three groups was statistically significant (P = 0.011), and post-hoc comparisons confirmed that the Immediate LC Group differed significantly from both delayed groups. Calot's triangle dissection time was also significantly shorter in the Immediate LC Group (26.72 ± 4.23 min) compared to the 24-h Delayed Group (33.25 ± 3.94 min) and the 24–72-h Delayed Group (32.47 ± 5.97 min), with an overall difference that was highly significant (P < 0.001). With regard to estimated blood loss, the Immediate LC Group had the lowest value (11.71 ± 8.86 mL), followed by the 24-h Delayed LC Group (18.98 ± 41.58 mL) and the 24–72-h Delayed LC Group (19.79 ± 21.87 mL); however, these differences did not reach statistical significance (P = 0.214). The rate of conversion to open surgery did not differ significantly among the three groups (P = 1.000), with no conversions in the Immediate LC Group, one conversion (2.04%) in the 24-h Delayed LC Group, and two conversions (1.89%) in the 24–72-h Delayed LC Group.

Table 3.

Comparison of intraoperative outcomes among the three groups

	Immediate LC group (n = 45)	24-h delayed LC group (n = 49)	24–72-h delayed LC group (n = 106)	Statistical test values	P value
Blood loss ($\overline{\text{x}}\pm \text{s}$, mL)	11.71 ± 8.86	18.98 ± 41.58	19.79 ± 21.87	1.555	0.214
Calot's triangle dissection time ($\overline{\text{x}}\pm \text{s}$, min)	26.72 ± 4.23a	33.25 ± 3.94b	32.47 ± 5.97b	23.779	< 0.001
Operative time ($\overline{\text{x}}\pm \text{s}$, min)	106.31 ± 46.03a	135.02 ± 55.52b	140.22 ± 72.45b	4.625	0.011
Intraoperative conversion to open [n (%)]				0.834	1.000
Yes	0 (0.00%)	1 (2.04%)	2 (1.89%)
No	45 (100.00%)	48 (97.96%)	104 (98.11%)

Please note the superscript letters (a, b, c) for each number; if the same letter appears, it indicates no statistical significance between them

After propensity score matching

After propensity score matching, the differences in intraoperative outcomes were further accentuated. As shown in Table 4, the Immediate LC Group still had the shortest operative time (103.48 ± 40.31 min), whereas the 24‑72‑h Delayed LC Group had the longest (140.03 ± 69.52 min); the 24‑h Delayed LC Group showed an intermediate value (132.29 ± 44.50 min). The overall difference among the three groups was statistically significant (P = 0.043). Pairwise comparisons revealed that the Immediate LC Group had a significantly shorter operative time than the 24‑72‑h Delayed LC Group (P < 0.05), but no significant difference was observed between the Immediate and 24‑h Delayed groups.

Table 4.

Intraoperative outcomes among the three groups after propensity score matching

	Immediate LC group (n = 25)	24-h delayed LC group (n = 21)	24–72-h delayed LC group (n = 38)	Statistical test values	P value
Blood loss ($\overline{\text{x}}\pm \text{s}$, mL)	10.48 ± 10.22a	12.62 ± 7.85a,b	24.61 ± 32.47b	3.457	0.036
Calot's triangle dissection time ($\overline{\text{x}}\pm \text{s}$, min)	26.53 ± 4.08a	33.46 ± 4.35b	32.45 ± 6.09b	13.309	< 0.001
Operative time ($\overline{\text{x}}\pm \text{s}$, min)	103.48 ± 40.31a	132.29 ± 44.50a,b	140.03 ± 69.52b	3.282	0.043
Intraoperative conversion to open [n (%)]				1.802	0.567
Yes	0 (0.00%)	0 (0.00%)	1 (2.63%)
No	25 (100.00%)	21 (100.00%)	37 (97.37%)

Please note the superscript letters (a, b, c) for each number; if the same letter appears, it indicates no statistical significance between them

Calot’s triangle dissection time was also significantly shorter in the Immediate LC Group (26.53 ± 4.08 min) than in the 24‑h Delayed Group (33.46 ± 4.35 min) and the 24–72‑h Delayed Group (32.45 ± 6.09 min), with a highly significant overall difference (P < 0.001).

Estimated blood loss differed significantly after matching (P = 0.036). The Immediate LC Group had the lowest blood loss (10.48 ± 10.22 mL), followed by the 24‑h Delayed LC Group (12.62 ± 7.85 mL), while the 24–72‑h Delayed LC Group had the highest (24.61 ± 32.47 mL). Pairwise comparisons indicated that the Immediate LC Group lost significantly less blood than the 24–72‑h Delayed LC Group (P < 0.05); no significant differences were found between the Immediate and 24‑h Delayed groups, nor between the 24‑h Delayed and 24–72‑h Delayed groups.

The conversion rate to open surgery remained low across the three matched groups, with no significant difference (P = 0.567). No conversion occurred in either the Immediate or the 24‑h Delayed LC Group, whereas one patient (2.63%) in the 24–72‑h Delayed LC Group required conversion.

Collectively, these findings indicate that performing LC immediately after ERCP yields superior intraoperative outcomes—particularly shorter operative time and reduced blood loss—compared with delaying surgery to 24–72 h. The 24‑hour delay strategy produced intermediate results that did not differ significantly from either extreme in most comparisons.

Postoperative outcomes

Before propensity score matching

In the unmatched cohort, as shown in Table 5, the time to first flatus differed significantly among the three groups (P = 0.014). The Immediate LC Group had the shortest time (1.00 [0.75, 1.50] days), followed by the 24-h Delayed LC Group (1.50 [1.00, 2.00] days) and the 24–72-h Delayed LC Group (1.00 [1.00, 2.00] days).

Table 5.

Comparison of postoperative outcomes among the three groups

	Immediate LC group (n = 45)	24-h delayed LC group (n = 49)	24–72-h delayed LC group (n = 106)	Statistical test values	P value
Postoperative time to first flatus [M(P25,P75),d]	1.00 (0.75,1.50)a	1.50 (1.00,2.00)a,b	1.00 (1.00,2.00)b	8.521	0.014
Postoperative complications [n (%)]
Fever	7 (15.56%)a	13 (26.53%)b	41 (38.68%)b	8.451	0.015
Acute pancreatitis	5 (11.11%)	10 (20.41%)	24 (22.64%)	2.710	0.258
Hyperamylasemia	5 (11.11%)a	13 (26.53%)a,b	33 (31.13%)b	6.286	0.043
Bile leak	0 (0.00%)	1 (2.04%)	1 (0.94%)	1.178	0.720
Stone clearance [n (%)]				0.513	0.774
Complete	44 (97.78%)	40 (95.92%)	94 (95.28%)
Incomplete	1 (2.22%)	2 (4.08%)	5 (4.72%)
TBiL ($\overline{\text{x}}\pm \text{s}$,μmol/L)	37.46 ± 31.08	29.54 ± 31.95	27.89 ± 18.28	2.289	0.104
DBiL ($\overline{\text{x}}\pm \text{s}$,μmol/L)	12.14 ± 10.28	11.73 ± 17.40	10.28 ± 9.95	0.462	0.631
GGT ($\overline{\text{x}}\pm \text{s}$,U/L)	131.13 ± 110.55a	190.62 ± 185.31a,b	199.40 ± 149.71b	3.302	0.039
ALP ($\overline{\text{x}}\pm \text{s}$,U/L)	124.91 ± 50.55	119.43 ± 71.74	112.34 ± 57.52	0.753	0.472
ALT ($\overline{\text{x}}\pm \text{s}$,U/L)	55.53 ± 30.23a	76.65.66.17a,b	87.96 ± 67.61b	4.494	0.012
AST ($\overline{\text{x}}\pm \text{s}$,U/L)	48.96 ± 23.48	39.88 ± 26.14	49.15 ± 41.12	1.319	0.270
ALB ($\overline{\text{x}}\pm \text{s}$,g/L)	34.93 ± 5.43	33.22 ± 3.84	34.77 ± 3.81	2.660	0.072
WBC ($\overline{\text{x}}\pm \text{s}$,109/L)	8.59 ± 2.48	8.75 ± 3.01	9.09 ± 3.19	0.524	0.593
AMY ($\overline{\text{x}}\pm \text{s}$,U/L)	47.20 ± 22.84a	69.81 ± 33.50b	67.39 ± 41.93b	5.844	0.003
Hospital Stay Indicators [M(P25,P75),d]
Total postoperative length of stay	8.00 (7.00,10.00)a	9.00 (7.00,11.00)a	10.00 (9.00,12.00)b	24.231	< 0.001
Length of stay after LC	5.00 (4.00,7.00)	5.00 (4.00,6.00)	5.00 (4.00,6.00)	4.182	0.124

Please note the superscript letters (a, b, c) for each number; if the same letter appears, it indicates no statistical significance between them

Regarding postoperative complications, the incidence of fever was lowest in the Immediate LC Group (15.56%), followed by the 24-h Delayed LC Group (26.53%) and the 24–72-h Delayed LC Group (38.68%), with a statistically significant difference (P = 0.015). Similarly, the incidence of hyperamylasemia was significantly lower in the Immediate LC Group (11.11%) than in the 24-h Delayed LC Group (26.53%) and the 24–72-h Delayed LC Group (31.13%) (P = 0.043). No significant differences were observed among the groups for acute pancreatitis (P = 0.258), bile leak (P = 0.720), or stone clearance status (P = 0.774).

Postoperative laboratory parameters showed significant differences for GGT (P = 0.039), ALT (P = 0.012), and AMY (P = 0.003). The Immediate LC Group had the lowest GGT levels (131.13 ± 110.55 U/L), while the 24–72-h Delayed LC Group had the highest (199.40 ± 149.71 U/L). For ALT, the Immediate LC Group also had the lowest levels (55.53 ± 30.23 U/L), followed by the 24-h Delayed LC Group (76.65 ± 66.17 U/L) and the 24–72-h Delayed LC Group (87.96 ± 67.61 U/L). Postoperative AMY levels were lowest in the Immediate LC Group (47.20 ± 22.84 U/L) compared with the 24-h Delayed LC Group (69.81 ± 33.50 U/L) and the 24–72-h Delayed LC Group (67.39 ± 41.93 U/L). No significant differences were found for TBiL, DBiL, ALP, AST, ALB, or WBC (all P > 0.05).

Regarding hospital stay, the total postoperative length of stay differed significantly among the three groups (P < 0.001). The Immediate LC Group had the shortest total stay (8.00 [7.00, 10.00] days), followed by the 24-h Delayed LC Group (9.00 [7.00, 11.00] days) and the 24–72-h Delayed LC Group (10.00 [9.00, 12.00] days). However, the length of stay after LC did not differ significantly across the groups (P = 0.124).

After propensity score matching

Following propensity score matching, as shown in Table 6, the differences in postoperative outcomes were further refined. The time to first flatus showed no significant difference among the three matched groups (P = 0.402), with median values of 1.00 [0.50, 1.50] days in the Immediate LC Group, 1.00 [1.00, 1.50] days in the 24‑h Delayed LC Group, and 1.00 [1.00, 2.00] days in the 24‑72‑h Delayed LC Group.

Table 6.

Postoperative outcomes among the three groups after propensity score matching

	Immediate LC group (n = 25)	24-h delayed LC group (n = 21)	24–72-h delayed LC group (n = 38)	Statistical test values	P value
Postoperative time to first flatus [M(P25,P75),d]	1.00 (0.50,1.50)	1.00 (1.00,1.50)	1.00 (1.00,2.00)	1.821	0.402
Postoperative complications [n (%)]
Fever	5 (20.00%)	3 (14.29%)	10 (26.32%)	1.115	0.554
Acute pancreatitis	2 (8.00%)	3 (14.29%)	6 (15.79%)	1.307	0.481
Hyperamylasemia	1 (4.00%)a	4 (19.05%)b	5 (13.16%)b	7.621	0.019
Bile leak	0 (0.00%)	0 (0.00%)	0 (0.00%)	0.000	1.000
Stone clearance [n (%)]				2.164	0.297
Complete	24 (96.00%)	20 (95.24%)	38 (100.00%)
Incomplete	1 (4.00%)	1 (4.76%)	0 (0.00%)
TBiL ($\overline{\text{x}}\pm \text{s}$,μmol/L)	35.32 ± 32.40	26.44 ± 12.28	33.35 ± 25.64	2.289	0.104
DBiL ($\overline{\text{x}}\pm \text{s}$,μmol/L)	10.70 ± 10.04	10.94 ± 6.59	13.06 ± 14.20	0.400	0.671
GGT ($\overline{\text{x}}\pm \text{s}$,U/L)	119.84 ± 86.17a	225.78 ± 217.13b	223.68 ± 180.53b	3.344	0.040
ALP ($\overline{\text{x}}\pm \text{s}$,U/L)	116.88 ± 50.45	136.19 ± 88.71	124.21 ± 62.52	0.480	0.621
ALT ($\overline{\text{x}}\pm \text{s}$,U/L)	58.12 ± 25.77a	72.71 ± 60.897a,b	97.00 ± 75.22b	3.250	0.044
AST ($\overline{\text{x}}\pm \text{s}$,U/L)	45.24 ± 22.56	36.90 ± 15.88	45.74 ± 28.48	1.009	0.369
ALB ($\overline{\text{x}}\pm \text{s}$,g/L)	35.22 ± 6.26	33.99 ± 4.39	34.82 ± 4.22	0.366	0.695
WBC ($\overline{\text{x}}\pm \text{s}$,109/L)	9.00 ± 2.38	8.27 ± 1.71	8.65 ± 3.10	0.448	0.641
AMY ($\overline{\text{x}}\pm \text{s}$,U/L)	44.04 ± 20.20a	63.24 ± 29.98b	62.68 ± 36.57b	3.271	0.043
Hospital Stay Indicators [M(P25,P75),d]
Total postoperative length of stay	8.00 (7.00,9.00)a	8.00 (7.00,9.00)a	9.50 (8.00,10.00)b	12.840	0.002
Length of stay after LC	5.00 (4.00,6.50)	4.00 (4.00,5.50)	4.00 (4.00,6.00)	5.018	0.081

Please note the superscript letters (a, b, c) for each number; if the same letter appears, it indicates no statistical significance between them

With regard to postoperative complications, the incidence of hyperamylasemia remained significantly different (P = 0.019). The Immediate LC Group had the lowest rate (4.00%), compared with the 24‑h Delayed LC Group (19.05%) and the 24–72‑h Delayed LC Group (13.16%). No significant differences were observed among the groups for fever (P = 0.554), acute pancreatitis (P = 0.481), bile leak (P = 1.000), or stone clearance status (P = 0.297).

Postoperative laboratory parameters showed patterns consistent with the unmatched analysis. Significant differences persisted for GGT (P = 0.040), ALT (P = 0.044), and AMY (P = 0.043). The Immediate LC Group demonstrated the lowest GGT levels (119.84 ± 86.17 U/L), whereas the 24‑h Delayed LC Group (225.78 ± 217.13 U/L) and the 24‑72‑h Delayed LC Group (223.68 ± 180.53 U/L) had higher values. For ALT, the Immediate LC Group also had the lowest levels (58.12 ± 25.77 U/L), followed by the 24‑h Delayed LC Group (72.71 ± 60.89 U/L) and the 24–72‑h Delayed LC Group (97.00 ± 75.22 U/L). Postoperative AMY levels were lowest in the Immediate LC Group (44.04 ± 20.20 U/L) compared with the 24‑h Delayed LC Group (63.24 ± 29.98 U/L) and the 24‑72‑h Delayed LC Group (62.68 ± 36.57 U/L). No significant differences were observed for TBiL, DBiL, ALP, AST, ALB, or WBC (all P > 0.05).

To further assess the resolution of hepatobiliary and pancreatic injury beyond absolute postoperative values, the changes from baseline (Δ) were calculated as the POD 1 value minus the pre‑ERCP value. The Immediate LC Group had a mean ΔALT of a decrease of 134.37 ± 230.38 U/L, compared with a decrease of 109.00 ± 73.63 U/L in the 24-h Delayed Group and a decrease of 108.13 ± 175.13 U/L in the 24–72-h Delayed Group (P = 0.827). For ΔAST, the values were a decrease of 89.43 ± 136.27, a decrease of 60.81 ± 148.71, and a decrease of 88.99 ± 150.72 U/L, respectively (P = 0.744). ΔGGT values were a decrease of 150.72 ± 267.15, a decrease of 131.22 ± 170.66, and a decrease of 130.16 ± 181.28 U/L (P = 0.921). ΔAMY values were a decrease of 14.76 ± 39.71, a decrease of 1.48 ± 23.76, and a decrease of 9.11 ± 35.37 U/L (P = 0.429). None of these differences reached statistical significance, suggesting equivalent physiological recovery rates across the three groups.

Regarding hospital stay, the total postoperative length of stay remained significantly different among the three matched groups (P = 0.002). The Immediate LC Group and the 24‑h Delayed LC Group had similar median stays (8.00 [7.00, 9.00] days and 8.00 [7.00, 9.00] days, respectively), both of which were significantly shorter than that of the 24–72‑h Delayed LC Group (9.50 [8.00, 10.00] days). No significant difference was observed in the length of stay after LC among the three matched groups (P = 0.081).

The median follow-up duration for the matched cohort was 7 (6, 9) days. During this period, no late complications such as biliary strictures or bile leak were identified. Following our institutional protocol for minor abdominal surgery, an active follow-up (via telephone or outpatient visit) was conducted at 1 week post‑discharge, after which a passive surveillance approach was adopted, with patients instructed to return only if new symptoms occurred.

Discussion

The optimal timing of laparoscopic cholecystectomy (LC) following endoscopic retrograde cholangiopancreatography (ERCP) for choledocholithiasis has remained a subject of ongoing debate [20]. The present propensity score‑matched cohort study sought to address this clinical uncertainty by systematically comparing perioperative outcomes among patients undergoing LC at three precisely defined time points: immediately (3–6 h), at 24 h, and at 24–72 h after successful ERCP. Immediate LC was associated with superior intraoperative and postoperative outcomes compared to delayed approaches, particularly when contrasted with the 24–72‑hour delayed strategy.

The key findings of this study can be summarized as follows. First, immediate LC significantly reduced operative time, Calot's triangle dissection time, and intraoperative blood loss compared to delayed surgery, with the most pronounced differences observed between the Immediate and 24–72‑h Delayed groups. Second, postoperative recovery was enhanced in the immediate LC cohort, as evidenced by a lower incidence of hyperamylasemia, more favorable profiles of liver function parameters (GGT and ALT), and reduced postoperative AMY levels. Third, total postoperative hospital stay was significantly shortened in patients undergoing immediate LC. Importantly, these benefits were achieved without any increase in postoperative complications or conversion rates, underscoring the safety of this approach.

Operative efficiency and technical considerations

The finding that immediate LC was associated with significantly shorter operative time and reduced blood loss aligns with the concept of an optimal "surgical window" shortly after ERCP. This was further substantiated by the objective measurement of Calot's triangle dissection time, which was significantly shorter in the Immediate LC Group, indicating reduced technical difficulty during the critical phase of dissection. During this early period, the inflammatory response following endoscopic sphincterotomy and stone extraction has not yet fully developed, and local tissue edema remains minimal. Consequently, the anatomical planes within Calot's triangle are relatively clear, which facilitates safe and efficient dissection. In contrast, waiting 24–72 h allows the inflammatory reaction to reach its peak, characterized by pronounced tissue edema, hyperemia, and early adhesion formation. These changes may obscure the critical anatomical structures, thereby increasing technical difficulty and prolonging operative time [21–23]. The finding that the 24‑h Delayed Group exhibited intermediate outcomes—not significantly different from either extreme in most comparisons—further supports the notion that the optimal window may be narrow and that prolonged delay confers no advantage and may offer no additional benefit.

Postoperative recovery and complications

The significantly lower incidence of hyperamylasemia in the immediate LC group represents a clinically important finding. Post-ERCP hyperamylasemia, even in the absence of overt pancreatitis, reflects subclinical pancreatic injury that may be exacerbated by subsequent surgical manipulation [24, 25]. Delaying LC for 24–72 h may allow this pancreatic inflammation to persist or worsen, potentially increasing susceptibility to surgical trauma during gallbladder dissection. By proceeding immediately with LC, the pancreas is spared this additional insult during a vulnerable period. However, analysis of the changes from baseline (Δ) in AMY, ALT, AST, and GGT revealed no significant differences among the three groups, suggesting equivalent physiological recovery rates across the three groups. This finding aligns with the observation that the length of stay after LC did not differ among groups (P = 0.081), suggesting that the shorter total hospital stay in the immediate group was primarily attributable to the elimination of the waiting period between ERCP and surgery, rather than accelerated postoperative recovery.

The safety of immediate LC was further supported by the comparable incidence of postoperative pancreatitis across the three groups. In the matched cohort, acute pancreatitis occurred in 8.00% of the Immediate LC Group, 14.29% of the 24‑h Delayed Group, and 15.79% of the 24‑72‑h Delayed Group, with no statistically significant difference (P = 0.481). This finding suggests that proceeding with immediate LC does not increase the risk of post‑ERCP pancreatitis. Moreover, the well‑established safety of the “rendezvous” technique—in which LC is performed immediately after ERCP under the same anesthesia—provides strong precedent that delaying surgery by 3–6 h does not introduce additional pancreatic risk [26].

The shorter total postoperative hospital stay observed in the immediate LC group is particularly noteworthy from a healthcare resource utilization perspective. This finding likely reflects the cumulative benefits of improved perioperative outcomes, including reduced operative time, lower blood loss, and fewer complications such as hyperamylasemia. Importantly, the length of stay after LC itself was comparable among groups, indicating that the overall reduction in hospitalization was primarily attributable to eliminating the waiting period between ERCP and surgery—a finding with direct implications for patient throughput and cost-effectiveness.

Comparison with previous studies

The existing literature on the timing of LC after ERCP has yielded heterogeneous findings. Some earlier retrospective studies reported increased technical difficulty or higher complication rates with early intervention, leading to caution among surgeons [27, 28]. However, many of these studies were limited by small sample sizes, inconsistent definitions of "early" and "delayed" surgery, and inadequate adjustment for confounding variables. More recent investigations, including several meta-analyses, have suggested that same-admission or early LC may be safe and potentially beneficial [29, 30]. The present study extends these findings by employing rigorous propensity score matching to control for a comprehensive set of potential confounders, thereby providing more robust evidence that immediate LC is not merely non-inferior but actually superior to delayed approaches in selected patients.

The findings of the present study align with and strengthen the conclusions of a recent systematic review that advocated for same-admission cholecystectomy after ERCP [31]. However, the present study offers greater granularity by distinguishing between 24‑hour and 24‑72‑hour delays, demonstrating that the benefits of early intervention are most pronounced when surgery is performed without any planned delay. This finding has practical implications for hospital workflow and operating room scheduling.

Clinical implications and alignment with ERAS principles

The findings of this study support the integration of immediate LC into ERAS protocols for patients with choledocholithiasis. ERAS pathways emphasize minimizing preoperative waiting, reducing surgical stress, and accelerating postoperative recovery—all of which are facilitated by the immediate approach [32]. By eliminating the waiting period between ERCP and LC within the same hospitalization, immediate LC reduces total hospital stay, potentially lowers costs, and minimizes the risk of recurrent biliary events during the interval between procedures. For healthcare systems seeking to optimize resource utilization without compromising patient outcomes, the immediate strategy represents an attractive option.

It is important to note, however, that patient selection remains critical. The benefits observed in this study were achieved in a cohort that excluded patients with severe cholangitis, severe acute pancreatitis, or significant comorbidities. Immediate LC may not be appropriate for patients with ongoing sepsis, organ failure, or those requiring intensive care management [33, 34]. Clinical judgment and individual patient assessment should always guide decision-making.

Limitations

Several limitations of this study should be acknowledged. First, despite propensity score matching, the retrospective design carries a risk of residual confounding from unmeasured variables, such as surgeon experience or subtle variations in ERCP technique. Second, this was a single-center study, which may limit generalizability to other institutions with different practice patterns. Third, the matched sample size, while adequate for major outcomes, may have been insufficient to detect small but clinically meaningful differences in rare complications (e.g., bile leak). Fourth, the follow-up period was relatively short (median 7 days), as our institutional protocol for minor abdominal surgery uses active follow-up at 1 week followed by passive, symptom-triggered surveillance. Although LC has a low incidence of late events, this limited window may not capture rare or late-onset complications such as biliary strictures. Future studies with extended, structured follow-up are needed to better characterize long-term outcomes.

Conclusions

In conclusion, this propensity score-matched cohort study demonstrates that immediate laparoscopic cholecystectomy following successful ERCP for choledocholithiasis is associated with superior perioperative outcomes compared to delayed surgical strategies. Immediate LC significantly reduces operative time and blood loss, lowers the incidence of postoperative hyperamylasemia, improves early recovery of liver function, and shortens total hospital stay, all without increasing complication rates.

Author contributions

Conception and design of study: Ming Li, Zhen Liu, Zhong Li; Acquisition of data: Ming Li, Zhen Liu, Dake Liu, Zhenduo Zhang, Liwei Wang, Zhong Li, Yingxin Wang, Long Zhao; Data analysis and interpretation: Ming Li, Zhen Liu, Dake Liu, Zhenduo Zhang, Liwei Wang, Zhong Li, Yingxin Wang, Long Zhao; Writing—original draft preparation: Zhen Liu, Dake Liu, Zhong Li; Writing—review and editing: Ming Li, Zhen Liu, Dake Liu, Zhenduo Zhang, Liwei Wang, Zhong Li, Yingxin Wang, Long Zhao; All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by the Medical Science Research Project Plan of Hebei Province (2024), issued by the Health Commission of Hebei Province (Grant No. 20241612).

Data availability

The datasets utilized and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Disclosures

Zhen Liu, Zhong Li, Dake Liu, Zhenduo Zhang, Liwei Wang, Yingxin Wang, Long Zhao, and Ming Li have no conflicts of interest or financial ties to disclose.