Predictors of the technical success of ultrasound-guided percutaneous transhepatic biliary drainage for biliary obstruction

Abstract

BACKGROUND

Percutaneous transhepatic biliary drainage (PTBD) is a critical minimally invasive intervention for biliary obstruction, with technical success being a key determinant of clinical outcomes.

AIM

To identify and quantify the clinically modifiable factors predicting technical success of ultrasound-guided PTBD, and thereby to develop an evidence-based framework for procedural triage and standardization.

METHODS

A total of 263 ultrasound-guided PTBD procedures performed at three tertiary hospitals (2019-2024) were retrospectively analyzed. Technical success (defined as catheter patency at 2 weeks) and complications were evaluated. Multivariate logistic regression was used to assess provider-related, patient-related, and technical factors, with statistical significance set at P < 0.05.

RESULTS

The cumulative success rate was 94.3% (248/263 cases), with a first-attempt success rate of 72.6% (191/263). Repeat attempts significantly increased cumulative success (P < 0.01) but did not significantly increase the overall complication rate (P = 0.415). The significant predictors in the final multivariable model included operator experience [> 20 cases; odds ratio (OR) = 1.83; 95%CI: 1.12-2.99; P = 0.015], use of dedicated puncture probes (OR = 2.46; 95%CI: 1.16-5.24; P = 0.019), adequate periprocedural cooperation (OR = 2.73; 95%CI: 1.14-6.57; P = 0.025), a bile duct diameter ≥ 6 mm (OR = 2.13; 95%CI: 1.30-3.48; P = 0.003), and a bilirubin concentration > 340 μmol/L (OR = 2.58; 95%CI: 1.52-4.36; P < 0.001). These predictors were incorporated into a practical bedside scoring system to stratify procedural success probability.

CONCLUSION

This study proposes a preliminary framework for standardizing ultrasound-guided PTBD by identifying three actionable determinants: Operator experience (> 20 procedures), a defined reattempt limit (≤ 2 attempts), and specific anatomic-biochemical criteria (bile duct diameter ≥ 6 mm and bilirubin concentration > 340 μmol/L). The use of dedicated puncture probes is a key technical modifier. This framework translates evidence into practical decision support, providing a foundation for protocol formalization and prospective validation.

Key Words: Percutaneous transhepatic biliary drainage; Ultrasound guidance; Biliary obstruction; Technical success; Predictive factors

Core Tip: This multicenter study established the first evidence-based clinical decision-making framework for predicting technical success in ultrasound-guided percutaneous transhepatic biliary drainage (PTBD). Key actionable determinants were identified: Operator experience (> 20 procedures), bile duct diameter (≥ 6 mm), and the severity of obstruction (bilirubin > 340 μmol/L). The framework also defines a safe procedural limit (≤ 2 attempts) and identifies the use of dedicated puncture probes as a critical technical enhancer. This work translates evidence into practical guidance and provides a foundation for standardizing PTBD protocols.

INTRODUCTION

Biliary tract patency is essential for maintaining normal hepatobiliary physiology. Obstruction may arise from choledocholithiasis, biliary tumors, or extrinsic compression caused by malignancies of adjacent organs (e.g., the pancreas, duodenum, gallbladder, or liver)[1,2]. This disruption of enterohepatic circulation may lead to obstructive jaundice, hepatocellular damage, and functional impairment. In such cases, reestablishing biliary drainage is the primary therapeutic intervention, with percutaneous transhepatic biliary drainage (PTBD) serving as an important therapeutic option[1,2].

PTBD is a minimally invasive procedure that yields significant clinical benefits through three primary roles: (1) Therapeutic intervention: Draining bile to relieve biliary obstruction and secondary complications (e.g., cholangitis and hepatic dysfunction) and managing biliary/pancreatic leaks[3-5]; (2) Diagnostic tool: Enabling cholangiography, bile culture, pathological examination, and cholangioscopy-guided biopsy[6]; and (3) Access route for further interventions: Serving as a critical step for percutaneous cholangioscopic lithotomy and facilitating tumor-directed therapies, including radiofrequency ablation, stent placement, and brachytherapy[2,7-9].

The technical success of PTBD serves as the fundamental prerequisite for achieving clinical benefits. Given its broad diagnostic and therapeutic applications across various diseases, PTBD has become a commonly performed procedure. In large medical centers, PTBD is typically performed by dedicated interventional teams. However, challenges such as the absence of specialized puncture centers, scheduling constraints, or emergency clinical scenarios where PTBD must be performed as an adjunct to other surgical procedures may arise. Therefore, in this study, we aimed to establish an evidence-based, practical framework for ultrasound-guided PTBD by systematically identifying and integrating the key modifiable determinants of technical success.

MATERIALS AND METHODS

Study design

This multicenter retrospective study analyzed PTBD procedures performed under ultrasound guidance at three regional tertiary hospitals from January 2019 to December 2024 (6-year period). All consecutive PTBD cases performed by clinicians from internal medicine, surgery, and emergency departments were included in this study, excluding those conducted by dedicated interventional radiology teams. We evaluated success/failure rates and systematically assessed key influencing factors, including operator-related variables (e.g., specialty and experience), patient characteristics (e.g., disease severity), and technical strategies (e.g., access route and device selection).

PTBD technical protocol

All procedures were performed under ultrasound guidance using disposable PTBD kits, with multiside-hole pigtail drainage catheters (7-10F diameter) selected on the basis of clinical requirements. The established PTBD guidelines were strictly followed during the procedure, with either the two-step or one-step technique used at the discretion of the operator. In cases of initial failure, repeat attempts (≥ 3 procedures permitted) were allowed until technical success was achieved; alternatively, conversion to percutaneous transhepatic gallbladder drainage (PTGD)/endoscopic retrograde cholangiopancreatography (ERCP) or procedure abandonment, which was classified as failure was permitted. Postprocedural management included: (1) Close monitoring for 2 weeks, with failure defined as the occurrence of catheter occlusion/dislocation/fracture; and (2) Clavien-Dindo grade ≥ III complications (e.g., major hemorrhage, bile leakage) requiring catheter replacement or surgical intervention.

Operational procedure

The PTBD procedure was performed as follows: (1) Preoperative evaluation: Ultrasonographic assessment of the liver and intrahepatic bile ducts to determine the puncture site and patient position, followed by standard disinfection and local anesthesia; (2) Two-step method: (a) Ultrasound-guided insertion of a PTC needle into the target bile duct with confirmation of bile aspiration; (b) Guidewire advancement to the predetermined position; and (c) Tract dilation and catheter placement over the guidewire, followed by fixation after position confirmation; and (3) One-step method: Steps (a)-(b) were identical to those of the two-step method; these steps were followed by direct advancement of the preloaded drainage catheter under continuous ultrasound guidance, which was then secured. The catheter position was reconfirmed before the drainage bag was connected for both methods.

Outcome measures

Procedure metrics: The attempt frequency, success/failure rates per attempt (defined as catheter placement with patency maintained for 2 weeks), and complications. Each attempt was stratified into steps: Step 1: Puncture (skin-to-bile duct access with bile return, corresponding to technical step 1); step 2: Cannulation (tract dilation/guidewire insertion/catheter deployment via a two-step method (steps 2-3) or analogous one-step maneuvers); and step 3: Drainage (2-week functional maintenance per previous criteria). Failure was defined as any unmet metric. The predictors of success were as follows: Multivariate analysis of provider-related (e.g., specialty, experience), patient-related (e.g., bilirubin level, duct diameter), and technical (e.g., approach selection) factors.

Statistical analysis

Categorical variables are presented as n (%), and continuous variables are presented as the means ± SD. All analyses were performed using IBM SPSS Statistics 26.0 (IBM Corp., Armonk, NY, United States). Normality was assessed using Shapiro-Wilk tests; nonnormally distributed data were either transformed or analyzed using nonparametric equivalents. Comparative analyses included independent t tests (normal continuous data), χ² tests (categorical data, expected counts ≥ 5 per cell), and Fisher’s exact tests (categorical data, expected counts < 5). The learning curve was analyzed using the cumulative summation (CUSUM) method. Univariate screening preceded multivariate logistic regression (forward logistic regression method) to identify independent predictors, with statistical significance set at P < 0.05 (two-tailed).

RESULTS

Baseline procedure outcomes

For the 263 initial PTBD attempts, the first-attempt success rate was 72.6% (191/263), with 72 failures (3 cases abandoned, 69 cases that involved second attempts). The success rate for second attempts was 62.3% (43/69), with 26 failures, whereas that for third/subsequent attempts was 60.9% (14/23). Cumulatively, 355 procedures were performed (248 successes, 107 failures), with an overall per-procedure success rate of 69.9% (248/355). Case-based analysis demonstrated: (1) A success rate of 89.0% (234/263) for cases involving ≤ 2 attempts; and (2) An overall success rate of 94.3% (248/263) for all attempts, with a failure rate of 5.7% (15/263) (Figure 1).

Figure 1 Flowchart of patient enrollment and procedural outcomes. Diagram illustrates the pathway of 263 patients undergoing initial ultrasound-guided percutaneous transhepatic biliary drainage through sequential attempts. Boxes indicate the number of procedures at each attempt stage (first, second, and third/subsequent) and their immediate outcomes (success or failure). The cumulative technical success rate after all attempts was 94.3% (248/263).

First, second, and third/subsequent attempts [A1: 72.6% (191/263) vs A2: 62.3% (43/69) vs A3: 60.9% (14/23); P = 0.172], indicating that the per-attempt success rate did not significantly differ across attempts. Case-based analysis revealed no difference between the success of initial attempts and that of cumulative single attempts (A1 vs B1: P = 0.454), whereas the occurrence of a second attempt after initial failure significantly improved outcomes (cumulative success rate after two attempts: 89.0% (234/263) vs that of an initial attempt alone; A1+A2 vs B2: P < 0.01). Additional attempts (≥ 3) provided incremental benefits [overall success: 94.3% (248/263) vs success after ≤ 2 attempts; A1+A2+A3 vs B3: P < 0.01], but no significant difference was observed when comparing cases requiring third attempts vs those requiring only second attempts (B2: 89.0% vs B3: 94.3%; P = 0.062), suggesting diminishing returns beyond two attempts (Figure 2).

Figure 2 Success rates stratified by attempt sequence. A: Per-attempt technical success rates for first (A1: 73%), second (A2: 62%), and third/subsequent (A3: 61%) attempts; B: Cumulative case-based success rates after one (B1: 70%), two (B2: 89%), and three or more (B3: 94%) attempts. Statistical comparisons (P values) between groups are detailed in the main text.

Complications and failure-step analysis

Among the 263 patients, 28 had complications (10.6%, 28/263), all of which were Clavien-Dindo grade I-III complications. Complication rates stratified by number of attempts were not significantly different [single attempt: 9.3% (18/194); two attempts: 13.0% (6/46); ≥ 3 attempts: 17.4% (4/23); P = 0.415], with comparable severity distributions (P = 0.487). The distribution of failure steps across attempt sequences did not differ significantly (P = 0.139). Although not statistically significant, a numerically increasing trend in overall complications was observed with repeated attempts. A similar numerical trend was also noted for severe complications (Clavien-Dindo grade III: 1.0%, 4.4%, and 9.1%, respectively). These findings suggest that after two failed attempts, clinicians should carefully weigh the risk-benefit ratio of continuing PTBD vs considering alternative drainage methods (e.g., PTGD/ERCP), as repeated attempts may be associated with a numerically higher complication burden (Figure 3). The corresponding numerical data are detailed in Supplementary Table 1.

Figure 3 Analysis of complications and procedural failure steps. A: Overall and attempt-stratified rates of complications (Clavien-Dindo grades I-III); B: Distribution of the failed technical step (Puncture, Cannulation, or Drainage) across first, second, and third/subsequent attempts. The P-value for the trend in drainage failure across attempts is indicated.

Medical provider-related factors

Operator experience significantly affected the technical success, with procedures performed by operators with experience of > 20 cases demonstrating higher success rates [odds ratio (OR) = 1.85; 95%CI: 1.19-3.10; P = 0.007]. The learning curve analysis (CUSUM) method identified > 20 cases as the critical threshold for achieving stabilized operational efficiency (Figure 4). Compared with junior physicians, senior physicians achieved better outcomes (OR = 3.65, 95%CI: 1.77-10.06; P = 0.001). Effective coordination among operator, procedural assistant (“guide”), and patient was another significant predictor (good vs poor cooperation: OR = 2.65, 95%CI: 1.05-6.09; P = 0.039). In contrast, guide experience, guide title, and operator specialty were not significantly associated (all P > 0.05) (Table 1 and Supplementary Figure 1).

Figure 4 Receiver operating characteristic curve for the simplified predictive score. The curve depicts the discriminative ability of the multivariable logistic regression model (presented in Table 4) for predicting procedural success. The area under the curve was 0.714 (95%CI: 0.658-0.770, P < 0.001). The optimal probability cut-off was 0.371. AUC: Area under the curve.

Table 1 Domain-specific multivariable analysis of medical provider-related factors.

	Success/failure	P value	Exp(B)	95%CI for Exp(B)
Operator experience		0.007
> 20 cases	112/65		1.92	1.19	3.10
≤ 20 cases	136/42		0.52	0.32	0.84
Operator title		0.001
Senior	83/27		4.22	1.77	10.06
Junior	165/80		0.24	0.10	0.57
Guide experience		0.818
> 20 cases	86/38		1.06	0.64	1.75
≤ 20 cases	162/69		0.94	0.57	1.55
Guide title		0.327
Senior	89/32		1.29	0.78	2.14
Junior	159/75		0.78	0.47	1.29
Specialty		0.250
Surgery	125/49		0.68	0.35	1.32
Internal medicine	85/38		2.36	0.93	5.97
Emergency medicine	38/20		1.48	0.76	2.89
Level of cooperation		0.039
Good	235/95		2.53	1.05	6.09
Poor	13/12		0.40	0.16	0.95

“Senior” denotes attending physician or equivalent title.

Patient-related factors

The target bile duct diameter and bilirubin level were significantly associated with technical success. Procedures targeting ducts ≥ 6 mm had higher success rates (OR = 2.845; 95%CI: 1.738-4.656; P < 0.01), whereas a bilirubin concentration > 340 μmol/L further improved outcomes (OR = 3.366; 95%CI: 1.157-9.796; P = 0.0038). In contrast, age (> 65 years vs ≤ 65 years: OR = 0.93, P = 0.848), sex (male vs female: OR = 0.907, P = 0.823), BMI (> 24 kg/m² vs ≤ 24 kg/m²: OR = 1.097, P = 0.781), a history of upper abdominal surgery (OR = 0.616, P = 0.343), liver cirrhosis/fibrosis (OR = 0.967, P = 0.827), ascites (OR = 0.832, P = 0.695), obstruction etiology (tumor vs other: OR = 0.798, P = 0.917), and obstruction site (supra-confluence vs infra-confluence: OR = 0.964, P = 0.696) were not significantly associated (Table 2).

Table 2 Domain-specific multivariable analysis of patient-related factors.

	Success/failure	P value	Exp(B)	95%CI for Exp(B)
Age		0.848
> 65 years	151/67		0.93	0.568	1.522
≤ 65 years	97/40		1.075	0.657	1.759
Gender		0.823
Male	180/77		0.907	0.532	1.545
Female	68/30		1.103	0.647	1.879
BMI		0.781
≤ 24 (kg/m2)	208/90		1.097	0.567	2.125
> 24 (kg/m2)	40/17		0.911	0.47	1.765
History of upper abdominal surgery		0.343
No	239/102		0.616	0.19	1.996
Yes	9/5		1.625	0.501	5.269
Liver cirrhosis/fibrosis		0.827
No	229/92		0.967	0.41	2.283
Yes	20/11		1.034	0.438	2.439
Ascites		0.695
No	241/103		0.832	0.215	3.218
Yes	7/4		1.202	0.311	4.648
Etiology		0.917
Tumor	234/97		0.798	0.295	2.157
Other	15/6		1.254	0.464	3.392
Obstruction site		0.696
Supra-confluence	114/47		0.964	0.595	1.563
Infra-confluence	134/60		1.037	0.64	1.681
Bilirubin		0.038
> 340 (μmol/L)	123/27		3.366	1.157	9.796
170-340 (μmol/L)	119/75		1.737	0.627	4.811
< 170 (μmol/L)	6/5		0.576	0.208	1.595
Bile duct diameter		0.002
≥ 6 (mm)	167/60		2.845	1.738	4.656
< 6 (mm)	81/57		0.351	0.215	0.575

Bilirubin > 340 μmol/L is defined as severe hyperbilirubinemia. BMI: Body mass index.

Strategy selection-related factors

Compared with the use of conventional probes, the use of puncture-specific ultrasound probes was associated with higher technical success (OR = 2.219; 95%CI: 1.098-4.486; P = 0.026). In contrast, procedure timing (emergency vs elective: OR = 0.757, P = 0.300), procedure location (operating room vs bedside: OR = 0.863, P = 0.672), puncture angle (45°-135° vs others: OR = 0.945, P = 0.835), needle trajectory length (≥ 12 cm vs < 12 cm: OR = 1.129, P = 0.649), puncture position (right vs left side: OR = 1.214, P = 0.598), catheter diameter (≤ 8F vs > 8F: OR = 0.812, P = 0.440), intraprocedural operator change (OR = 1.242, P = 0.762), and target bile duct modification (OR = 5.962, P = 0.101) were not significantly associated with the outcomes (Table 3).

Table 3 Domain-specific multivariable analysis of strategy selection-related factors.

Characteristics	Success/failure	P value	Exp(B)	95%CI for Exp(B)
Timing		0.300
Emergency	56/30		0.757	0.447	1.282
Elective	192/77		1.321	0.78	2.237
Location		0.672
Operating room	129/50		1.155	0.593	2.25
Puncture room	81/40		0.863	0.43	1.732
Bedside	38/17		0.866	0.444	1.687
Angle		0.835
45°-135°	185/80		0.945	0.556	1.607
Other	63/27		1.058	0.622	1.799
Length		0.649
≥ 12 (cm)	63/30		1.129	0.67	1.901
< 12 (cm)	185/77		0.886	0.526	1.492
Position		0.598
Right side	31/12		1.214	0.59	2.497
Left side	217/95		0.824	0.4	1.695
Ultrasound probe		0.026
Conventional	198/96		0.451	0.223	0.911
Puncture-specific	50/11		2.219	1.098	4.486
Catheter diameter		0.440
≤ 8 French	191/78		0.812	0.478	1.379
> 8 French	57/29		1.232	0.725	2.094
Operator change		0.762
Yes	8/1		1.242	0.306	5.04
No	47/34		0.805	0.198	3.267
Target change		0.101
Yes	6/4		5.962	0.704	50.501
No	49/31		0.168	0.02	1.421

“Puncture-specific” refers to dedicated biopsy/guidance ultrasound probes.

To integrate these predictors into a practical tool, a consolidated multivariable logistic regression model was constructed (Table 4). This final model identified six factors independently associated with procedural success. Greater operator experience (OR = 1.83, 95%CI: 1.12-2.99), use of a dedicated puncture probe (OR = 2.46, 95%CI: 1.16-5.24), optimal periprocedural cooperation (OR = 2.73, 95%CI: 1.14-6.57), a bile duct diameter ≥ 6.0 mm (OR = 2.13, 95%CI: 1.30-3.48), and a bilirubin concentration > 340 μmol/L (OR = 2.58, 95%CI: 1.52-4.36) were significantly associated with higher odds of an uncomplicated, successful procedure (all P < 0.05). Operator seniority showed a positive but statistically nonsignificant association (OR = 1.42, 95%CI: 0.82-2.46; P = 0.207).

Table 4 Final comprehensive multivariable logistic regression model of factors associated with procedural success.

Variable	β coefficient	Odds ratio (95%CI)	P value
Puncture performer’s title (high vs low)	0.353	1.423 (0.823-2.461)	0.207
Puncture experience (more vs less)	0.606	1.833 (1.124-2.989)	0.015
Dedicated probe (special vs ordinary)	0.901	2.462 (1.157-5.238)	0.019
Patient cooperation (good vs poor)	1.005	2.732 (1.136-6.568)	0.025
Bile duct diameter (≥ 6 mm vs < 6 mm)	0.754	2.126 (1.300-3.478)	0.003
Bilirubin level (> 340 μmol/L vs < 340 μmol/L)	0.946	2.576 (1.523-4.357)	< 0.001
Constant	-3.187	0.041	< 0.001

This final multivariable model identified five independent predictors of procedural success (P < 0.05). Operator seniority was not statistically significant (P = 0.207). The model’s predictive performance is summarized in Figure 4.

The model demonstrated moderate discriminative ability. The area under the receiver operating characteristic curve was 0.714 (95%CI: 0.658-0.770; P < 0.001) (Figure 4). At the optimal probability cutoff of 0.371, which was determined by maximizing Youden’s index, the sensitivity and specificity for predicting procedural success were 56.1% and 77.0%, respectively.

For clinical application, the regression coefficients were translated into a bedside scoring system (Table 5). Points were assigned for the presence of each predictor associated with higher success: Greater operator experience (2 points), use of a specialized probe (3 points), good procedural cooperation (3 points), a bile duct diameter ≥ 6.0 mm (2 points), serum bilirubin concentration > 340 μmol/L (3 points), and higher operator seniority (1 point). The total score (range 0-14) categorized patients into three levels of predicted procedural success: Low (0-4 points), intermediate (5-9 points), and high (10-14 points).

Table 5 Bedside scoring system for predicting the probability of successful percutaneous transhepatic biliary drainage.

Predictor	Points assigned if:	Points
Operator experience	More experience	2
Dedicated probe	Specialized probe used	3
Procedural cooperation	Good cooperation	3
Bile duct diameter	≥ 6 mm	2
Serum bilirubin	> 340 μmol/L	3
Operator seniority	High title	1
Total score		0-14
Predicted success probability	Corresponding score range
Low	0-4
Intermediate	5-9
High	10-14

Points are assigned based on the regression coefficients and clinical weight. Operator seniority, while not statistically significant in the multivariate model (P = 0.207), was included due to its established clinical relevance in procedural settings.

DISCUSSION

The management of biliary obstruction involves three principal approaches: Percutaneous (interventional), endoscopic (duodenoscope-assisted), and surgical[1,10,11]. Surgical intervention remains the cornerstone for definitive disease eradication despite its invasive nature. Percutaneous and endoscopic techniques provide minimally invasive alternatives for biliary decompression[12], serving as either bridging therapies prior to definitive surgery or standalone palliative measures[2]. These approaches represent distinct technical paradigms that can be used individually or in combination, depending on disease characteristics, institutional capabilities, and therapeutic goals[2-4,13,14]. Consequently, optimal drainage strategy selection requires careful consideration of patient-specific factors, local expertise, and alignment with long-term treatment objectives.

ERCP enables physiological internal drainage, whereas endoscopic nasobiliary drainage provides an alternative external drainage approach with lower patient tolerance. ERCP facilitates comprehensive diagnostic evaluation, including upper gastrointestinal and biliary tract assessment with targeted biopsies, along with therapeutic interventions such as biliary stent placement and stone extraction[12,14,15]. While demonstrating particular efficacy for distal biliary obstructions (below the hepatic hilum) and having versatile therapeutic capabilities, ERCP presents several challenges: (1) A requirement for specialized operator expertise; (2) A dependence on advanced endoscopic equipment; (3) The need for fluoroscopic guidance with associated radiation exposure; (4) The need for patient sedation; and (5) Limited accessibility for proximal (hilar) lesions.

The ERCP procedure comprises three critical technical phases: (1) Duodenoscope advancement through the upper gastrointestinal tract to identify the major duodenal papilla; (2) Selective bile duct cannulation; and (3) Navigation across the obstructive segment for stent or drainage tube deployment. Technical challenges potentially compromising success include anatomical variations (particularly postsurgical reconstruction), luminal stenosis, periampullary diverticula, and impassable strictures[16]. Contemporary advancements, including Spyglass cholangioscopy and endoscopic ultrasound-guided access, have substantially addressed these limitations while avoiding radiation exposure. For ERCP-ineligible patients, alternative endoscopic approaches such as hepaticogastrostomy, choledochoduodenostomy, and cholecystoduodenostomy now provide viable salvage options[14,17,18].

PTBD primarily involves percutaneous external drainage, with the additional capability for internal drainage when obstructive segments can be traversed to reach the distal bile duct[19]. This technique has demonstrated efficacy across all biliary obstruction types (both intrahepatic and extrahepatic), particularly intrahepatic or bilateral hilar strictures, as it circumvents the need to traverse the obstruction—a key limitation of ERCP. The expanding clinical applications of the procedure, including percutaneous cholangioscopic lithotomy, stent placement, and intraluminal brachytherapy via PTBD access, have significantly expanded biliary interventions. PTBD involves three critical phases: (1) Percutaneous penetration through the abdominal wall layers (parietal/visceral peritoneum and hepatic capsule); (2) Intrahepatic needle navigation; and (3) Targeted bile duct cannulation with guidewire placement and catheter deployment. Successful execution depends on the interaction of three key elements: Patient-specific factors (e.g., anatomical variations and respiratory motion), real-time imaging quality, and operator expertise.

PTBD has a technical success rate of 85%-97%, demonstrating comparable or superior efficacy to ERCP in specific patient populations[12,20]. The spectrum of procedure-related complications encompasses both infectious sequelae (cholangitis, liver abscess, or sepsis) and noninfectious events (hemorrhage, bile leakage, catheter dislodgment, or drainage dysfunction)[21,22]. Comparative analyses revealed the distinct clinical advantages of PTBD: (1) Significantly lower postprocedural pancreatitis rates; (2) More effective infection control; and (3) Superior jaundice relief with greater hepatic function recovery[20,23]. Conversely, ERCP has advantages in terms of (1) A reduced rehospitalization frequency; (2) A shorter hospitalization duration; and (3) A potentially lower overall mortality risk[24]. These evidence-based considerations underscore that the selection of the optimal drainage modality requires a comprehensive evaluation of patient-specific anatomical and clinical factors, institutional technical capabilities, and planned subsequent therapeutic pathways, with both techniques offering valuable primary or salvage options depending on individual clinical scenarios.

Operator expertise and decision-making constitute the fundamental determinants of PTBD success, governing procedural execution and technical optimization across all phases. Multiple procedural factors were associated with both the complication rates and the technical outcomes[25]. Three key modifiable factors were identified: (1) Optimized team coordination; (2) Advanced guidance systems (e.g., dedicated puncture probes); and (3) Enhanced visualization through biliary duct dilation. While the other variables were not statistically significant, they may still have clinical effects. Our data establish 20 procedures as the critical competency threshold, corresponding to improved performance, a finding that is consistently observed among experienced operators. Repeated attempts significantly increased cumulative success (P < 0.01) without substantially increasing complications (P = 0.415), although a numerical increase in severe adverse events was observed with repeated attempts. This evidence underscores the need for balanced clinical judgment between persistence and risk mitigation, particularly when alternative drainage modalities (PTGD/ERCP) are considered in complex cases[26].

Optimal procedural management is essential for achieving successful technical outcomes. During the puncture phase, three critical factors determine bile duct access: (1) Precise needle manipulation (operator skill); (2) Accurate guidance (assistant and probe coordination); and (3) The management of dynamic factors (liver mobility and adjacent vasculature). The cannulation phase presents specific challenges: Inadequate tract dilation through anatomical layers (peritoneum, hepatic capsule, duct wall), suboptimal catheter/guidewire selection (stiffness or mismatched size), or anatomical resistance points causing wire/catheter angulation (e.g., peritoneal/hepatic capsule/duct wall penetration or respiratory-induced displacement).

For complex cases that require repeated punctures, key precautions include preventing hemorrhage (intra/extrabiliary and intra/extrahepatic), bile leakage, and gas ingress, which may degrade ultrasound images. The recommended strategies are as follows: (1) Saline flushing to avoid air introduction; (2) Limited bile aspiration before catheter placement to prevent duct collapse; and (3) Thorough saline flushing to remove air artifacts. When needed, intraductal contrast-enhanced ultrasound or combined X-ray cholangiography can increase catheter placement precision and facilitate stenting. Prolonged procedures risk compromising outcomes because of operator fatigue and reduced team coordination.

We found that while repeat attempts increased cumulative success, they altered the risk profile: Per-attempt success did not improve, and a numerical increase in drainage failure and severe complications was observed with repeated attempts, although these trends did not reach statistical significance[27]. This necessitates dynamic risk assessment. We therefore propose two attempts as a critical decision point. Guided by our prediction model, operators should reassess the case if success is not achieved after two attempts, clinicians should use our prediction model to reassess the case: Evaluate whether optimizing modifiable predictors (e.g., operator experience, use of a dedicated probe, team cooperation) could increase the likelihood of success; if not, alternative drainage methods should be strongly considered[26,28,29]. This framework is not a rigid rule but a practical tool for balancing risks against benefits and guiding individualized decision-making.

CONCLUSION

This study revealed that the success of ultrasound-guided PTBD is influenced by technical and operational factors. Key predictors include operator experience, a bile duct diameter ≥ 6 mm, and a bilirubin concentration > 340 μmol/L. Allowing up to three attempts was associated with a high cumulative success rate (94.3%) with an acceptable safety profile (10.6% complications). The use of a dedicated guidance system and optimal teamwork further significantly improved outcomes. Based on these findings, we developed a practical bedside scoring system, offering a tool for preoperative risk stratification and clinical decision support.

ACKNOWLEDGEMENTS

We gratefully acknowledge: All clinical staff for their dedicated operational support. Prof. Hai-Yang Li (Emergency Department, Peking Union Medical College Hospital) for expert guidance on ultrasound-guided puncture training.