Abstract
Background
Severe acute pancreatitis (SAP) complicated by acute respiratory distress syndrome (ARDS) features high mortality. This study aimed to clarify the clinical role of the double-lumen irrigation-suction tube (DLIST) in this patient population.
Material/Methods
This single-center retrospective study enrolled a total of 101 patients with SAP-ARDS from January 2020 to December 2025. After exclusion, 91 patients were eligible for analysis and divided into 2 groups: the DLIST group and a standard care group. Propensity score matching was used to balance age. Clinically observed outcomes were 28-, 60-, and 90-day all-cause mortality and ventilator-free days. Cox regression was applied to estimate the association between the DLIST intervention and mortality.
Results
We performed 1: 1 optimal pair matching without replacement. After propensity score matching, an age-matched cohort of 70 patients (35 pairs) was obtained. Compared with the standard care group, the DLIST group had lower 28-day (20.0% vs 54.3%, P = 0.003), 60-day (28.6% vs 60.0%, P = 0.008), and 90-day (34.3% vs 60.0%, P = 0.031) all-cause mortality, and more ventilator-free days at day 28 (12 ± 8 vs 8 ± 9 days, P = 0.041). Multivariable Cox regression showed that DLIST was independently associated with lower 90-day mortality (adjusted HR = 0.29, 95% CI 0.13–0.62, P = 0.002).
Conclusions
This exploratory retrospective cohort study indicates that early DLIST within 14 days of SAP symptom onset may be associated with short-term clinical outcomes among patients with SAP complicated by ARDS. However, prospective multicenter randomized trials are necessary to verify causality, identify the optimal timing, screen appropriate candidates, and formulate standardized procedures.
Keywords: Pancreatitis, Respiratory Distress Syndrome, Propensity Score
Introduction
Acute pancreatitis is a prevalent gastrointestinal disorder characterized by substantial heterogeneity in clinical progression and patient outcomes [1,2]. Severe acute pancreatitis (SAP), defined by persistent organ dysfunction lasting more than 48 hours [3], typically develops in the phase of the disease as a consequence of sterile systemic inflammatory responses [4,5]. This severe clinical phenotype confers a poor prognosis, with reported mortality rates ranging from 7.9% to 43.8% in contemporary clinical cohorts [6]. A substantial proportion of patients with SAP develop acute necrotic collections, which represent an early local complication of necrotizing pancreatitis [6,7]. In patients with concurrent acute necrotic collections and incipient organ failure, the sterile inflammatory phenotype often overlaps with secondary infectious complications, particularly during the second to third week of disease evolution [8]. Accumulating observational data suggest timely intervention for this high-risk group can alleviate systemic inflammation and facilitate organ function recovery [9].
SAP initiates a cytokine storm and severe systemic inflammatory response [10], triggering systemic capillary leakage and pulmonary edema. Circulating activated pancreatic enzymes directly injure lung tissue, while elevated intra-abdominal pressure evolves into abdominal compartment syndrome, further impairing ventilation and collectively inducing ARDS. This pathological cascade leads to multiple organ failure, prolonged ventilation, and high mortality.
In recent years, there has been a paradigm shift in the management of SAP from surgical to minimally invasive management [11,12]. Traditional single-lumen drainage merely drains peritoneal fluid passively without continuous lavage, while catheters are frequently occluded by necrotic debris, failing to control sustained intra-abdominal and pulmonary inflammation. As a minimally invasive approach for focal abdominal irrigation and drainage, the double-lumen irrigation-suction tube (DLIST) has been adopted in pancreatitis treatment [13]. It enables synchronous peritoneal lavage and negative-pressure aspiration, effectively reducing intraperitoneal inflammatory burden, avoiding catheter obstruction, and blocking secondary lung injury.
Peritoneal lavage was once proposed to eliminate intraperitoneal inflammatory mediators and activated pancreatic enzymes, theoretically suppressing systemic inflammation and distant organ damage [14]. Nevertheless, its optimal intervention window, technical details, and eligible patient population remain controversial clinically. Although DLIST has clear theoretical advantages, high-quality clinical studies focusing specifically on its use in patients with SAP combined with acute respiratory distress syndrome (ARDS) are still lacking, and its actual clinical value in this subgroup has not been fully verified.
This exploratory retrospective cohort study aimed to assess the correlation between DLIST and short-term prognosis for individualized drainage strategy guidance.
Material and Methods
Ethics Statement
The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This retrospective cohort study was approved by the Ethics Committee of the People’s Hospital of Guangxi Zhuang Autonomous Region (approval No.: KY-ZC-2025-049). The center has 5 wards and 88 treatment units; all the wards with sufficient expertise were assessed for eligibility. Given that this study was retrospective and conducted strict data anonymization procedures to protect patient privacy, the requirement for individual patient informed consent was waived by the above ethics committee.
Inclusion and Exclusion Criteria
We screened all adult patients (≥ 18 years old) admitted with SAP complicated by ARDS between January 1, 2020, and December 31, 2025. SAP was diagnosed according to the Revised Atlanta Classification [3]. Patients were eligible for additional screening if they met the following criteria: (1) diagnosed with acute pancreatitis within 7 days after the onset of abdominal pain; (2) had computed tomography (CT)-confirmed acute necrotic collections that were accessible percutaneously; and (3) had persistent (at least 48 h) and unresolved organ failure (respiratory, renal, or cardiovascular as per the Revised Atlanta Classification) on day 7 after the onset of abdominal pain. Organ failure in this study was defined by the modified Marshall score (respiratory and renal) or Sequential Organ Failure Assessment (SOFA) score ≥ 2 (cardiovascular) for the individual organ system. ARDS was defined according to the Berlin criteria [15]. Standard care for SAP [16] included fluid therapy, analgesics, management of organ failure and intensive care, antibiotics to prevent and control infectious complications, nutritional support, biliary tract management, vasopressor management, and intensive care unit (ICU) supportive care. The DLIST intervention followed standardized institutional criteria to minimize treatment-selection bias. Patients in the DLIST group were defined as those who underwent DLIST within 14 days of SAP symptom onset.
Exclusion criteria included pregnancy, advanced malignancy, chronic respiratory failure, admission to our center after open surgery for pancreatic necrotic tissue evacuation, or incomplete medical records.
Operation Procedure: DLIST
The double-lumen tube used in this study consists of 3 components: an outer tube, a suction tube, and an irrigation tube. Under CT guidance [8,17], a 16–26 F outer luminal tube was inserted, with its tip positioned in the peripancreatic necrotic or fluid-collection region and firmly secured. The outer luminal tube was a flexible and transparent outer luminal tube. Ten circular lateral holes, each measuring 0.3 to 0.5 cm in diameter, were uniformly cut at intervals of 0.3 to 0.5 cm along the anterior 3 to 5 cm of the tube wall using scissors. The outer tube was used for continuous normal saline irrigation [18,19]. An intravenous scalp needle tube was used as the irrigation tube, with its distal end fixed at the top of the outer tube and extending 0.3 to 0.5 cm beyond the outer tube, while its proximal end was connected to the infusion device. The infusion volume ranged from 3000 to 6000 mL/day, with a drip rate of 3 to 4 mL/min. The inner luminal tube was connected to a negative pressure of -10 to −20 kPa for simultaneous aspiration. The inner luminal tube served as the suction tube, which was inserted into the outer tube as a disposable suction catheter and secured to the distal end of the outer tube with No. 4 silk thread. The fixation thread should not be excessively short, to ensure that the tip of the suction tube has 2 to 3 cm of mobility within the outer tube. The suction tube is connected to a vacuum pump for continuous negative pressure suction.
Inflow and outflow volumes were balanced to eliminate pancreatic enzymes, necrotic debris, and inflammatory exudates. Daily irrigation volume and characteristics of the irrigation solution were recorded, and both sets of catheters were replaced weekly. When the drainage became clear and the output decreased markedly, the irrigation rate was gradually tapered, and the tube was removed.
Data Collection
At baseline (ICU admission as day 1), demographics, comorbidities, vital signs, laboratory values (hematological, inflammatory, and biochemical markers on ICU days 1, 3, 7, and 14), severity scores (SOFA, Acute Physiological and Chronic Health Evaluation [APACHE] II), and intervention details were extracted from the patients’ electronic medical records. Clinically observed outcomes included ventilator-free days at 28 days, lengths of ICU and hospital stay, changes in oxygenation (PaO2, FiO2), 28-day all-cause mortality, 60-day all-cause mortality, and 90-day all-cause mortality.
Propensity Score Matching
To minimize confounding, we performed 1: 1 propensity score matching (PSM) using the R package MatchIt (version 4.2.2). A common support region was imposed to ensure comparability. Balance was assessed using standardized mean differences. PSM was performed using age as the only matching variable, since age was the most significant confounder between the 2 groups. Pair matching without replacement and without a caliper was applied to minimize the total absolute distance across pairs.
Statistical Analysis
Continuous variables are expressed as mean with standard deviation (mean ± SD) or median with interquartile range (median (IQR) and were compared using the Welch t test or Mann-Whitney U test. Categorical variables are presented as frequencies (percentages) and were analyzed with the Pearson chi-square test. Survival analysis was performed with Kaplan-Meier curves and the log-rank test. Cox regression (univariable and multivariable) was used to estimate the association between DLIST intervention and mortality risk, after adjusting for potential confounders. The time origin for survival analysis was ICU admission. Patients were right-censored at the end of the 90-day follow-up, with no loss to follow-up. A 2-sided P value < 0.05 was considered statistically significant. All analyses were conducted using R software (version 4.2.2) and MSTATA (www.mstata.com).
Results
Patient Selection
A total of 101 patients with SAP complicated by ARDS were initially screened from the electronic medical records of the Department of Critical Care Medicine, The People’s Hospital of Guangxi Zhuang Autonomous Region, between 2020 and 2025. Exclusion criteria were applied as follows: age less than 18 years old (n = 1); pregnancy (n = 2); advanced malignancy (n = 0); chronic respiratory failure (n = 0); admitted to our center after open surgery for pancreatic necrotic tissue evacuation (n = 4); and incomplete clinical or laboratory data (n = 3). After exclusions, 91 patients were eligible for analysis and divided into 2 groups: the standard care group, consisting of 56 patients treated without peritoneal lavage; and the DLIST group, consisting of 35 patients who underwent dual-catheter peritoneal lavage within 14 days of SAP symptom onset (Figure 1).
Figure 1.
Flowchart of the patient recruitment process and dropout during the study. Abbreviations: SAP, severe acute pancreatitis; ARDS, acute respiratory distress syndrome; ICU, intensive care unit; DLIST, double-lumen irrigation-suction tube; PSM, propensity score matching.
Baseline Characteristics and Matching Process
Before matching, the DLIST group was significantly younger (39 ± 11 vs 51 ± 17 years old, P < 0.001), had lower hemoglobin levels, higher neutrophil percentages on day 14, and a lower prevalence of hypertension compared with the standard care group. To minimize baseline confounding, PSM was performed in a 1: 1 ratio, resulting in an age-balanced matched cohort of 70 patients (35 pairs). All subsequent outcome comparisons and survival analyses were conducted on this matched cohort. Significant differences were also observed in 28-day all-cause mortality, 60-day all-cause mortality, and 90-day all-cause mortality (all P < 0.05). After PSM (Tables 1–3), the balance of age was significantly improved, with a post-matching standardized mean difference of −0.152 (Table 4), confirming balance (Figure 2A, 2B). Age (41 ± 12 vs 39 ± 11, P = 0.305), SOFA scores (10.0 ± 4.3 vs 10.8 ± 3.2, P = 0.396), APACHE II scores (21.7 ± 6.4 vs 22.9 ± 4.9, P = 0.362), and comorbidity profiles were comparable between groups.
Table 1.
Baseline characteristics of patients in the standard care group and DLIST group before and after propensity score matching (PSM).
| Characteristic | Before PSM | p-value | After PSM | p-value | ||
|---|---|---|---|---|---|---|
| Standard care group (n=56) | DLIST group (n=35) | Standard care group (n=35) | DLIST group (n=35) | |||
| Age (years) | 51 ± 17 | 39 ± 11 | < 0.001 | 41 ± 12 | 39 ± 11 | 0.305 |
| Male (n,%) | 40 (71.4%) | 29 (82.9%) | 0.2152 | 30 (85.7%) | 29 (82.9%) | 0.743 |
| Hypertension (n,%) | 30 (53.6%) | 11 (31.4%) | 0.039 | 15 (42.9%) | 11 (31.4%) | 0.322 |
| Diabetes (n,%) | 25 (44.6%) | 15 (42.9%) | 0.867 | 19 (54.3%) | 15 (42.9%) | 0.339 |
| CHD (n,%) | 11 (19.6%) | 7 (20.0%) | 0.967 | 7 (20.0%) | 7 (20.0%) | > 0.999 |
| CKD (n,%) | 19 (33.9%) | 12 (34.3%) | 0.972 | 10 (28.6%) | 12 (34.3%) | 0.607 |
| ECMO (n,%) | 3 (5.4%) | 3 (8.6%) | 0.672 | 3 (8.6%) | 3 (8.6%) | > 0.999 |
| CRRT (n,%) | 17 (30.4%) | 12 (34.3%) | 0.696 | 14 (40.0%) | 12 (34.3%) | 0.621 |
| SAP symptom onset (day) | 9 ± 10 | 7 ± 7 | 0.376 | 9 ± 8 | 7 ± 7 | 0.371 |
| SOFA (score) | 9.9 ± 4.1 | 10.8 ± 3.2 | 0.261 | 10.0 ± 4.3 | 10.8 ± 3.2 | 0.396 |
| APACHE II (score) | 21.9 ± 5.9 | 22.9 ± 4.9 | 0.369 | 21.7 ± 6.4 | 22.9 ± 4.9 | 0.362 |
| PaO2/FiO2 ratio | 103 ± 35 | 106 ± 31 | 0.696 | 107 ± 40 | 106 ± 31 | 0.918 |
| Temp (°C) | 37.23 ± 0.87 | 37.34 ± 0.97 | 0.570 | 37.54 ± 0.84 | 37.34 ± 0.97 | 0.361 |
| RR (bpm/min) | 26 ± 15 | 22 ± 6 | 0.136 | 24 ± 6 | 22 ± 6 | 0.138 |
| HR (bpm/min) | 113 ± 18 | 117 ± 24 | 0.404 | 116 ± 18 | 117 ± 24 | 0.717 |
Abbreviations: PSM, Propensity score matching; DLIST Group, double-lumen irrigation-suction tube Group; CHD, coronary artery atherosclerotic heart disease; CKD, chronic kidney disease; ICU, intensive care unit; APACHE II, Acute Physiological and Chronic Health Evaluation II; SOFA, Sequential Organ Failure Assessment; ECMO, extracorporeal membrane oxygenation; CRRT, continuous renal replacement therapy; Temp, Celsius temperature scale; HR, heart rate; RR, respiratory rate. Continuous variables were expressed as mean ± SD and compared using Welch t test; categorical variables were presented as frequencies (percentages) and analyzed using Pearson chi-square test.
Table 2.
Baseline characteristics of patients in the standard care group and DLIST group before and after propensity score matching (PSM).
| Characteristic | Before PSM | P value | After PSM | P value | ||
|---|---|---|---|---|---|---|
| Standard Ccare group (n=56) | DLIST group (n=35) | Standard care group (n=35) | DLIST group (n=35) | |||
| WBC day 1 (109/L) | 14.4 ± 6.0 | 16.7 ± 12.1 | 0.316 | 14 ± 6 | 17 ± 12 | 0.310 |
| WBC day 3 (109/L) | 13 ± 6 | 15 ± 7 | 0.308 | 14 ± 6 | 15 ± 7 | 0.706 |
| WBC day 7 (109/L) | 15 ± 8 | 18 ± 9 | 0.175 | 15 ± 9 | 18 ± 9 | 0.307 |
| WBC day 14 (109/L) | 13 ± 5 | 15 ± 7 | 0.109 | 13 ± 6 | 15 ± 7 | 0.257 |
| Neu day 1 (109/L) | 84 ± 13 | 84 ± 6 | 0.848 | 85 ± 7 | 84 ± 6 | 0.598 |
| Neu day 3 (109/L) | 84.6 ± 12.9 | 84.9 ± 6.7 | 0.885 | 85.6 ± 6.5 | 84.9 ± 6.7 | 0.686 |
| Neu day 7 (109/L) | 84.1 ± 6.8 | 86.0 ± 5.2 | 0.131 | 84.8 ± 6.3 | 86.0 ± 5.2 | 0.398 |
| Neu day 14 (109//L) | 80 ± 14 | 84 ± 8 | 0.047 | 79 ± 16 | 84 ± 8 | 0.097 |
| Lym day 1 (109/L) | 1.00 ± 0.91 | 1.11 ± 0.60 | 0.458 | 1.19 ± 1.05 | 1.11± 0.60 | 0.717 |
| Lym day 3 (109/L) | 0.84 ± 0.55 | 0.96 ± 0.56 | 0.320 | 0.94 ± 0.54 | 0.96 ± 0.56 | 0.863 |
| Lym day 7 (109/L) | 1.06 ± 0.71 | 1.02 ± 0.59 | 0.772 | 1.07 ± 0.70 | 1.02 ± 0.59 | 0.726 |
| Lym day 14 (109/L) | 0.84 ± 0.44 | 0.85 ± 0.36 | 0.913 | 0.85 ± 0.48 | 0.85 ± 0.36 | 0.955 |
| Plt day 1 (109/L) | 183 ± 122 | 199 ± 95 | 0.492 | 190 ± 133 | 199 ± 95 | 0.751 |
| Plt day 3 (109/L) | 163 ± 135 | 174 ± 83 | 0.648 | 172 ± 155 | 174 ± 83 | 0.962 |
| Plt day 7 (109/L) | 213 ± 134 | 213 ± 98 | 0.991 | 208 ± 149 | 213 ± 98 | 0.886 |
| Plt day 14 (109/L) | 255 ± 141 | 266 ± 141 | 0.720 | 207 ± 115 | 266 ± 141 | 0.060 |
| HB day 1 (g/L) | 99 ± 35 | 85 ± 21 | 0.016 | 94 ± 36 | 85 ± 21 | 0.202 |
| HB day 3 (g/L) | 86 ± 28 | 74 ± 16 | 0.017 | 82 ± 27 | 74 ± 16 | 0.152 |
| HB day 7 (g/L) | 80 ± 20 | 73 ± 13 | 0.040 | 77 ± 21 | 73 ± 13 | 0.292 |
| HB day 14 (g/L) | 74 ± 12 | 74 ± 10 | 0.878 | 73 ± 13 | 74 ± 10 | 0.769 |
| PCT day 1 (ng/L) | 25 ± 27 | 17 ± 23 | 0.123 | 28 ± 28 | 17 ± 23 | 0.074 |
| PCT day 3 (ng/L) | 23 ± 22 | 16 ± 16 | 0.088 | 25 ± 22 | 16 ± 16 | 0.067 |
| PCT day 7 (ng/L) | 11 ± 15 | 8 ± 9 | 0.320 | 12 ± 17 | 8 ± 9 | 0.223 |
| PCT day 14 (ng/L) | 4.52 ± 2.73 | 4.64 ± 3.99 | 0.871 | 5.04 ± 2.92 | 4.64 ± 3.99 | 0.639 |
| IL-6 day 1 (ng/L) | 294 ± 263 | 271 ± 261 | 0.683 | 320 ± 307 | 271 ± 261 | 0.481 |
| IL-6 day 3 (ng/L) | 246 ± 185 | 232 ± 114 | 0.638 | 278 ± 210 | 232 ± 114 | 0.251 |
| IL-6 day 7 (ng/L) | 211 ± 167 | 232 ± 247 | 0.664 | 234 ± 188 | 232 ± 247 | 0.974 |
| IL-6 day 14 (ng/L) | 109 ± 51 | 130 ± 112 | 0.303 | 106 ± 38 | 130 ± 112 | 0.237 |
Abbreviations: PSM, Propensity score matching; DLIST Group, double-lumen irrigation-suction tube Group; WBC, white blood cell; NEU, neutrophilic granulocyte; Lym, lymphocyte count; PLT, platelet; HB, hemoglobin; PCT, procalcitonin; IL-6, interleukin-6; IMV, invasive mechanical ventilation. Continuous variables were expressed as mean ± SD and compared using Welch t test; categorical variables were presented as frequencies (percentages) and analyzed using Pearson chi-square test.
Table 3.
Outcomes of patients in the standard care group and double-lumen irrigation-suction tube (DLIST) group before and after propensity score matching (PSM).
| Characteristic | Before PSM | P value | After PSM | P value | ||
|---|---|---|---|---|---|---|
| Standard care group (n=56) | DLIST group (n=35) | Standard care group (n=35) | DLIST group (n=35) | |||
| 28-day ventilator-free | 9 ± 9 | 12 ± 8 | 0.1411 | 8 ± 9 | 12 ± 8 | 0.0411 |
| 28-day ICU stay | 21 ± 9 | 25 ± 6 | 0.0061 | 20 ± 9 | 25 ± 6 | 0.0071 |
| 60-day ICU stay | 36 ± 24 | 49 ± 19 | 0.0051 | 33 ± 23 | 49 ± 19 | 0.0031 |
| 90-day ICU stay | 50 ± 38 | 69 ± 32 | 0.0101 | 45 ± 38 | 69 ± 32 | 0.0051 |
| ICU stay (days) | 16 ± 13 | 22 ± 18 | 0.0791 | 16 ± 15 | 22 ± 18 | 0.1441 |
| Hospital stay (days) | 25 ± 18 | 60 ± 41 | < 0.0011 | 22 ± 17 | 60 ± 41 | < 0.0011 |
| 28-day all-cause mortality (n,%) | 27 (48.2%) | 7 (20.0%) | 0.0072 | 19 (54.3%) | 7 (20.0%) | 0.0032 |
| 60-day all-cause mortality (n,%) | 30 (53.6%) | 10 (28.6%) | 0.0192 | 21 (60.0%) | 10 (28.6%) | 0.0082 |
| 90-day all-cause mortality (n,%) | 30 (53.6%) | 12 (34.3%) | 0.0732 | 21 (60.0%) | 12 (34.3%) | 0.0312 |
Abbreviations: PSM, propensity score matching; DLIST Group, double-lumen irrigation-suction tube Group; ICU, intensive care unit. Continuous variables were expressed as mean ± SD and compared using Welch t test; categorical variables were presented as frequencies (percentages) and analyzed with Pearson chi-square test.
Table 4.
Standardized mean differences before and after propensity score matching.
| Before matching | After matching | |
|---|---|---|
| Age | −1.200 | −0.152 |
| Male | 0.303 | −0.076 |
| Hypertension | −0.477 | −0.246 |
| Diabetes | −0.036 | −0.231 |
| CHD | 0.009 | 0.00 |
| CKD | 0.008 | 0.120 |
| ECMO | 0.115 | 0.00 |
| CRRT | 0.083 | −0.120 |
| SAP symptom | ||
| Onset | −0.237 | −0.24 |
| SOFA | 0.273 | 0.241 |
| APACHE II | 0.211 | 0.255 |
| T | 0.118 | −0.206 |
| RR | −0.567 | −0.377 |
| HR | 0.165 | 0.078 |
| PF | 0.088 | −0.029 |
| WBC day 1 | 0.184 | 0.191 |
| Neu day 1 | −0.062 | −0.130 |
| Lym day 1 | 0.196 | −0.124 |
| Plt day 1 | 0.166 | 0.093 |
| HB day 1 | −0.679 | −0.429 |
| PCT day 1 | −0.363 | −0.484 |
| IL-6 day 1 | −0.089 | −0.185 |
Abbreviations: CHD, coronary artery atherosclerotic heart disease; CKD, chronic kidney disease; ICU, intensive care unit; APACHE II, Acute Physiological and Chronic Health Evaluation II; SOFA, Sequential Organ Failure Assessment; ECMO, extracorporeal membrane oxygenation; CRRT, continuous renal replacement therapy; Temp, Celsius temperature scale; HR, heart rate; RR, respiratory rate; WBC, white blood cell; NEU, neutrophilic granulocyte; Lym, lymphocyte count; PLT, platelet; HB, hemoglobin; PCT, procalcitonin; IL-6, interleukin-6.
Figure 2.
(A) Kernel density distribution of age in unmatched (left panel, unadjusted sample) and propensity score-matched cohorts (right panel, adjusted sample).
Red curve = standard care group (treatment = 0), cyan curve = DLIST group (treatment = 1). The degree of overlap between the density curves was markedly improved after propensity score matching, indicating a balanced age distribution between the 2 groups after matching. (B) Covariate balance assessment for age and propensity score distance before and after propensity score matching. Red circle = unmatched sample; blue triangle = matched sample. Left axis: absolute standardized mean differences (ASMD); right axis: variance ratio. ASMD and the variance ratio of age decreased substantially after matching.
Laboratory and Clinical Outcomes After PSM
Laboratory results showed no significant inter-group differences in white cell counts, platelets, procalcitonin, or interleukin-6. The DLIST group had lower lactate levels on day 7 (2.68 ± 1.55 vs 3.92 ± 3.23, P = 0.045) and higher calcium on day 14 (1.11 ± 0.11 vs 1.03 ± 0.19, P = 0.039), suggesting possible metabolic and inflammatory modulation. The DLIST group vs the standard care group had 28-day (20.0% vs 54.3%, P = 0.003), 60-day (28.6% vs 60.0%, P = 0.008), and 90-day (34.3% vs 60.0%, P = 0.031) all-cause mortality (Table 5). The PaO2/FiO2 ratio over time did not differ significantly between the standard care and DLIST groups at any time point, either before or after PSM (all P > 0.05) (Table 4). Although the total hospital stay was longer in the DLIST group (60 ± 41 vs 22 ± 17 days, P < 0.001), this likely reflects improved survival rather than prolonged morbidity.
Table 5.
Changes in PaO2/FiO2 ratio over time in the standard care group and DLIST group before and after propensity score matching (PSM).
| Characteristic | Before PSM | P value | After PSM | P value | ||
|---|---|---|---|---|---|---|
| Standard care group (n=56) | DLIST group (n=35) | Standard care group (n=35) | DLIST group (n=35) | |||
| PaO2/FiO2 ratio day 1 | 103.1 ± 35.2 | 106.4 ± 31.2 | 0.6961 | 106.9 ± 40.5 | 106.4 ± 31.4 | 0.918 |
| PaO2/FiO2 ratio day 3 | 151.8 ± 46.2 | 147.8 ± 60.0 | 0.728 | 154.9 ± 52.0 | 147.8 ± 60.0 | 0.601 |
| PaO2/FiO2 ratio day 7 | 194.1 ± 68.9 | 176.9 ± 64.0 | 0.239 | 191.2 ± 72.2 | 177.0 ± 64.0 | 0.383 |
| PaO2/FiO2 ratio day 14 | 195.0 ± 49.3 | 207.4 ± 75.4 | 0.345 | 189.3 ± 51.0 | 207.4 ± 75.4 | 0.245 |
| PaO2/FiO2 ratio day 7-1 | 48.6 (17.7, 81.9) | 29.6 (7.6, 64.4) | 0.234 | 40.9 (10.3, 76.8) | 29.6 (7.6, 64.4) | 0.431 |
| PaO2/FiO2 ratio day 7-3 | 37.5 (10.5, 72.7) | 27.6 (8.6, 49.6) | 0.385 | 23.4 (9.9, 54.6) | 27.6 (8.6, 49.6) | 0.810 |
| PaO2/FiO2 ratio day 14-1 | 57.8 (25.5, 83.9) | 51.30 (41.85, 8.6) | 0.692 | 52.9 (23.1, 82.0) | 51.3 (41.8, 8.6) | 0.441 |
| PaO2/FiO2 ratio day 14-3 | 50.0 (24.0, 70.2) | 63.8 (31.9, 104.8) | 0.164 | 49.0 (−5.1, 71.2) | 63.8 (31.9, 104.8) | 0.121 |
| PaO2/FiO2 ratio day 14-7 | 7.8 (−15.0, 39.8) | 20.0 (−7.0, 47.1) | 0.186 | 7.70 (−17.2, 28.5) | 20.0 (−7.0, 47.1) | 0.171 |
Abbreviations: PSM, propensity score matching; DLIST Group, double-lumen irrigation-suction tube group. Continuous variables were expressed as mean ± SD or median (IQR) and compared using Welch t test or the Mann-Whitney U test; categorical variables were presented as frequencies (percentages) and analyzed with Pearson chi-square test. Serial measurements were analyzed as separate between-group comparisons.
Survival Analysis and Cox Regression
Survival analysis (Figure 3) was performed for the standard care and DLIST groups. The DLIST group showed significantly higher survival rates at all 3 time points compared with the standard care group: 28-day (log-rank P = 0.003, hazard ratio [HR] = 0.292; 95%CI, 0.122–0.695), 60-day (log-rank P = 0.005; HR = 0.354; 95% CI, 0.166–0.753), and 90-day (log-rank P = 0.013; HR = 0.418; 95% CI, 0.205–0.852). The median survival time in the standard care group was 24 days, whereas the median survival time was not reached in the DLIST group during follow-up, indicating that fewer than 50% of patients in the DLIST group had died by the end of follow-up. To identify factors independently associated with 90-day mortality, we constructed a multivariable model including age, sex, SAP symptom onset days, SOFA score, APACHE II score, and DLIST intervention. Univariable Cox regression (Figure 4) identified DLIST (HR = 0.42; 95% CI, 0.20–0.85; P = 0.016), age (HR = 1.03; 95% CI, 1.01–1.06; P = 0.025), onset time of SAP (HR = 1.04; 95% CI; 1.01–1.09; P = 0.042), and SOFA score (HR = 1.10; 95% CI, 1.01–1.20; P = 0.028) as significant predictors of mortality. After adjustment in the multivariable model for age, sex, SOFA score, APACHE II score, and onset time of SAP, only DLIST (HR = 0.29; 95% CI, 0.13–0.62; P = 0.002) and onset time of SAP (HR = 1.12; 95% CI, 1.05–1.19; P < 0.001) remained independently associated with 90-day mortality (Table 6).
Figure 3.
Kaplan-Meier survival curves for 28-, 60-, and 90-day all-cause mortality after propensity score matching. Survival analysis was performed for the standard care group (red line) and DLIST group (blue line). The DLIST group showed significantly higher survival rates at all 3 time points compared with the standard care group: 28-day (log-rank P = 0.003; HR = 0.292; 95% CI, 0.122–0.695), 60-day (log-rank P = 0.005; HR = 0.354; 95% CI, 0.166–0.753), and 90-day (log-rank P = 0.013; HR = 0.418; 95% CI, 0.205–0.852). The median survival time in the standard care group was 24 days, whereas the median survival time was not reached in the DLIST group during follow-up.
Figure 4.
Forest plot of univariable and multivariable Cox regression for 90-day all-cause mortality in patients with severe acute pancreatitis complicated by acute respiratory distress syndrome. Green markers represent univariable Cox regression results, and red markers represent multivariable adjusted results. Dots indicate hazard ratios (HR), horizontal bars show 95% CIs. The vertical red dashed line at HR=1.0 is the threshold for statistical significance. Abbreviations: HR, hazard ratio; DLIST Group, double-lumen irrigation-suction tube Group; SAP, Severe acute pancreatitis; APACHE II, Acute Physiological and Chronic Health Evaluation II; SOFA, Sequential Organ Failure Assessment.
Table 6.
Univariable and multivariable Cox regression for 90-day all-cause mortality in patients with severe acute pancreatitis complicated by acute respiratory distress syndrome.
| Variables | Univariable | Multivariable | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| β | S.E | Z | P | HR (95% CI) | β | S.E | Z | P | HR (95% CI) | |
| Standard care (%) | 1.00 (Reference) | 1.00 (Reference) | ||||||||
| DLIST (%) | −0.87 | 0.36 | −2.4 | 0.016 | 0.42 (0.20 ~ 0.85) | −1.25 | 0.4 | −3.16 | 0.002 | 0.29 (0.13 ~ 0.62) |
| Female (%) | 1.00 (Reference) | 1.00 (Reference) | ||||||||
| Male (%) | −0.3 | 0.45 | −0.66 | 0.509 | 0.74 (0.31 ~ 1.80) | −0.71 | 0.5 | −1.42 | 0.156 | 0.49 (0.18 ~ 1.31) |
| Age (years) | 0.03 | 0.01 | 2.24 | 0.025 | 1.03 (1.01 ~ 1.06) | 0.02 | 0.02 | 1.37 | 0.17 | 1.02 (0.99 ~ 1.06) |
| SAP symptom onset (days) | 0.04 | 0.02 | 2.03 | 0.042 | 1.04 (1.01 ~ 1.09) | 0.11 | 0.03 | 3.73 | < .001 | 1.12 (1.05 ~ 1.19) |
| SOFA (score) | 0.1 | 0.04 | 2.19 | 0.028 | 1.10 (1.01 ~ 1.20) | 0.11 | 0.06 | 1.72 | 0.085 | 1.11 (0.99 ~ 1.26) |
| APACHE II (score) | 0.05 | 0.03 | 1.63 | 0.102 | 1.05 (0.99 ~ 1.12) | 0.05 | 0.04 | 1.27 | 0.205 | 1.06 (0.97 ~ 1.15) |
Abbreviations: HR, hazard ratio; DLIST Group, double-lumen irrigation-suction tube group; SAP, severe acute pancreatitis; APACHE II, Acute Physiological and Chronic Health Evaluation II; SOFA, Sequential Organ Failure Assessment.
Discussion
In this propensity-matched retrospective analysis, intervention with dual-catheter peritoneal lavage was associated with a substantial reduction in mortality among patients with SAP and ARDS. The DLIST group showed significantly lower all-cause mortality at 28 days (20.0% vs 54.3%, P = 0.003), 60 days (28.6% vs 60.0%, P = 0.008), and 90 days (34.3% vs 60.0%, P = 0.031), and more ventilator-free days at 28 days (12 ± 8 vs 8 ± 9 days, P = 0.041) compared with the standard care group.
Advantages of Our Study
Our findings align with the pathophysiological rationale that removal of pancreatic enzymes and inflammatory cytokines from the peritoneal cavity may attenuate lung injury and subsequent ARDS. The dual-catheter system enables continuous irrigation, which may be more effective than intermittent lavage in clearing mediators. The significant increase in ventilator-free days further suggests that lavage may facilitate earlier respiratory recovery, possibly by reducing alveolar damage.
The novelty of our study lies in its specific focus on the SAP-ARDS subgroup and the evaluation of an intervention window (within 14-days of SAP symptom onset time). Previous studies, such as the POINTER trial [20], have focused on patients with infected necrosis and compared immediate versus delayed drainage; its patient population was less severe (only 20% had organ failure at baseline), and the intervention timing was later (median 24 days). The trial did not show a benefit from immediate drainage. In contrast, our study included only patients with established organ failure (ARDS) and evaluated intervention within 14 days of SAP symptom onset time. This distinction is critical, as the pathophysiological drivers differ between early sterile inflammation and later infected necrosis. The multicenter randomized controlled TIMING trial led by Ke et al investigated the efficacy of early versus delayed percutaneous catheter drainage in patients with necrotizing pancreatitis and persistent organ failure, with participant screening performed on day 7 after disease onset based on predefined organ failure criteria [21]. Distinct differences in baseline disease severity and clinical characteristics were identified between the TIMING cohort and our study population. The patients enrolled in the TIMING trial exhibited significantly milder illness and earlier hospital admission, with an approximate median SOFA score of 4, APACHE II score of 15, and a 3-day interval from AP onset to admission. In contrast, our cohort presented with critically severe conditions, including median SOFA score of 10, APACHE II score of 21, and a prolonged 9-day interval from disease onset to hospitalization. Furthermore, the TIMING trial focused solely on optimizing drainage timing, while our study integrated both intervention timing and continuous lavage-drainage strategy as core therapeutic variables. The TIMING trial demonstrated no significant intergroup differences in primary composite outcomes, all-cause mortality, and 21-day organ failure-free days. Collectively, the discrepancies in baseline disease severity, admission latency, and therapeutic intervention strategies may explain the heterogeneous clinical outcomes across different studies. Further prospective studies with rigorously standardized continuous lavage-drainage protocols and strictly controlled intervention timing are warranted to validate the prognostic efficacy of this comprehensive therapeutic strategy in SAP populations. Similarly, Kohli et al [22] compared large-volume lavage with dependent drainage in SAP and found a trend toward decreased mortality (18.8% vs 28.8%, P = 0.370), consistent with our findings of improved outcomes associated with DLIST and a significant reduction in APACHE II scores (3.5 ± 3.405 vs 1.16 ± 3.811 P=0.012) in the lavage group. However, in the study of Kohli et al, the patients differed from those in our study, with moderate SAP observed in 19 (32%) patients, and SAP in 41 (68%). At admission, the APACHE II score in their study was 10.98 ± 3.934. Notably, despite longer total hospital stays, patients undergoing lavage in both studies had lower mortality, a pattern that likely reflects the successful rescue of critically ill patients who otherwise would have died early, rather than prolonged hospitalization due to complications.
Safety Profile and Complications of DLIST
Potential complications of DLIST include bleeding, catheter occlusion, and catheter-related infection. Bleeding is mostly induced by puncture injury or inflammatory vascular erosion, while catheter occlusion frequently results from accumulated necrotic tissue, blood clots, and dense exudates within the abdominal cavity. Prolonged catheter indwelling increases the risks of puncture-site infection and secondary intra-abdominal infection. All patients with SAP in our center received multidisciplinary integrated management, with daily joint ward rounds by intensive care and surgical specialists for real-time condition evaluation. Two episodes of severe intra-abdominal bleeding were observed in the DLIST group, which could not be definitely attributed to catheter-associated adverse events or progressive complications of SAP itself. Meanwhile, 1 case of massive intra-abdominal bleeding was documented in the standard treatment group, indicating that hemorrhage was more likely associated with the inherent progression of severe pancreatitis. Neither catheter occlusion nor catheter-related infection occurred in either group. Also, nurses continuously monitored lavage and drainage patency, and physicians routinely inspected puncture sites and performed timely dressing changes during daily ward rounds, which effectively reduced catheter-related infectious risks.
Limitations
This study has several limitations. First, this is a single-center, retrospective observational study; the sample size is small, especially after PSM, with only 35 patients per group. This project is a single-center, real-world cohort of patients with severe SAP-ARDS. The incidence of SAP combined with ARDS itself is low; cases meeting inclusion and exclusion criteria are scarce, and cases that retain complete DLIST intervention and follow-up data are scarce, which is the objective reason for the insufficient sample size after matching. Although PSM was used to balance only age, unmeasured residual confounding and potential selection bias remained, limiting causal inference. Second, the timing of DLIST initiation varied among patients, resulting in inconsistent exposure duration and possible temporal confounding. Third, from a methodological perspective, DLIST was an elective intervention after admission; patients had to survive until treatment initiation, leading to an inherent immortal-time bias. Only conventional Cox regression was used in this study, we did not apply a time-dependent Cox model. Therefore, immortal-time bias could not be eliminated, which may compromise the robustness of the association estimates. Fourth, the sample size and endpoint events after PSM matching were relatively limited, with events per variable (EPV) slightly lower than the conventional statistical threshold. Fifth, the multivariable Cox model carried potential overfitting risk in the small matched cohort, which was mitigated by dual-cohort validation using both the original and PSM-matched cohorts. A total of 6 covariates were incorporated into the multivariable Cox model. In the original cohort, 42 endpoint events were observed in the 90-day cohort (EPV=7.0), while 33 events occurred in the PSM-matched cohort (EPV=5.5), with the EPV of the matched cohort marginally below the conventional recommended cutoff value.
Conclusions
This exploratory retrospective cohort study indicates that early DLIST within 14 days of SAP symptom onset may be associated with better short-term clinical outcomes among patients with SAP complicated by ARDS; however, prospective multicenter randomized trials are necessary to verify causality, identify the optimal timing, screen appropriate candidates, and formulate standardized procedures.
Data Availability
The data used in this analysis may be made available from the corresponding author upon reasonable request.
Acknowledgments
During the preparation of this manuscript, the authors used AI-based language tools (including ChatGPT-5.2) for English language editing, grammar checking, and improving the clarity and readability of the text. These tools were used solely for language refinement and formatting purposes. All scientific content, including study design, data collection, statistical analysis, interpretation of results, and conclusions, was developed entirely by the authors.
Footnotes
Financial support: This study was funded by the Noncommunicable Chronic Diseases-National Science and Technology Major Project (No.2024ZD0526002 to J.P), the Guangxi Health Commission Self-funded Scientific Research Project (No. Z-A20250064 to X.J.T), and the Key R&D plan of Qingxiu District, Guangxi Zhuang Autonomous Region (No. 2020032 to S.L.X)
Conflict of interest: None declared
Publisher’s note: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher
Ethics Approval and Consent to Participate: This study was approved by the Ethics Committee of the People’s Hospital of Guangxi Zhuang Autonomous Region (approval number: KY-ZC-2025-049). The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Given that this study was retrospective and adopted strict data anonymization procedures to protect patient privacy, the requirement for individual patient informed consent was waived by the above ethics committee.
Declaration of Figures’ Authenticity: All figures submitted have been created by the authors who confirm that the images are original with no duplication and have not been previously published in whole or in part.
References
- 1.Xiao AY, Tan ML, Wu LM, et al. Global incidence and mortality of pancreatic diseases: A systematic review, meta-analysis, and meta-regression of population-based cohort studies. Lancet Gastroenterol Hepatol. 2016;1(1):45–55. doi: 10.1016/S2468-1253(16)30004-8. [DOI] [PubMed] [Google Scholar]
- 2.Barreto SG, Habtezion A, Gukovskaya A, et al. Critical thresholds: key to unlocking the door to the prevention and specific treatments for acute pancreatitis. Gut. 2021;70(1):194–203. doi: 10.1136/gutjnl-2020-322163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Banks PA, Bollen TL, Dervenis C, et al. Classification of acute pancreatitis-2012: Revision of the Atlanta classification and definitions by international consensus. Gut. 2013;62(1):102–11. doi: 10.1136/gutjnl-2012-302779. [DOI] [PubMed] [Google Scholar]
- 4.Schepers NJ, Bakker OJ, Besselink MG, et al. Impact of characteristics of organ failure and infected necrosis on mortality in necrotising pancreatitis. Gut. 2019;68(6):1044–51. doi: 10.1136/gutjnl-2017-314657. [DOI] [PubMed] [Google Scholar]
- 5.Garg PK, Singh VP. Organ failure due to systemic injury in acute pancreatitis. Gastroenterology. 2019;156(7):2008–23. doi: 10.1053/j.gastro.2018.12.041. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Matta B, Gougol A, Gao X, et al. Worldwide variations in demographics, management, and outcomes of acute pancreatitis. Clin Gastroenterol Hepatol. 2020;18(7):1567–75e2. doi: 10.1016/j.cgh.2019.11.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Ke L, Zhou J, Mao W, et al. Chinese Acute Pancreatitis Clinical Trials Group (CAPCTG) Immune enhancement in patients with predicted severe acute necrotising pancreatitis: A multicentre double-blind randomised controlled trial. Intensive Care Med. 2022;48(7):899–909. doi: 10.1007/s00134-022-06745-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Finkenstedt A, Jaber S, Joannidis M. Ten tips to manage severe acute pancreatitis in an intensive care unit. Intensive Care Med. 2023;49(9):1127–30. doi: 10.1007/s00134-023-07121-9. [DOI] [PubMed] [Google Scholar]
- 9.Zhang Y, Yu WQ, Zhang J, et al. Efficacy of early percutaneous catheter drainage in acute pancreatitis of varying severity associated with sterile acute inflammatory pancreatic fluid collection. Pancreas. 2020;49(9):1246–54. doi: 10.1097/MPA.0000000000001666. [DOI] [PubMed] [Google Scholar]
- 10.Hu Q, Hu Y, Tan C, et al. Acute pancreatitis: Mechanisms and therapeutic approaches. Signal Transduct Target Ther. 2026;11(1):15. doi: 10.1038/s41392-025-02394-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.van Santvoort HC, Besselink MG, Bakker OJ, et al. A step-up approach or open necrosectomy for necrotizing pancreatitis. New Engl J Med. 2010;362(16):1491–502. doi: 10.1056/NEJMoa0908821. [DOI] [PubMed] [Google Scholar]
- 12.van Brunschot S, van Grinsven J, van Santvoort HC, et al. Endoscopic or surgical step-up approach for infected necrotising pancreatitis: A multicentre randomised trial. Lancet. 2018;391(10115):51–58. doi: 10.1016/S0140-6736(17)32404-2. [DOI] [PubMed] [Google Scholar]
- 13.Niu D-G, Li W-Q, Huang Q, et al. Open necrosectomy combined with continuous positive drainage and prophylactic diverting loop ileostomy for late infected pancreatic necrosis: A retrospective cohort study. BMC Gastroenterol. 2020;20(1):212. doi: 10.1186/s12876-020-01343-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Becker V, Huber W, Meining A, et al. Infected necrosis in severe pancreatitis – Combined nonsurgical multi-drainage with directed transabdominal high-volume lavage in critically ill patients. Pancreatology. 2009;9(3):280–86. doi: 10.1159/000212093. [DOI] [PubMed] [Google Scholar]
- 15.Matthay MA, Arabi Y, Arroliga AC, et al. A new global definition of acute respiratory distress syndrome. Am J Respir Crit Care Med. 2024;209(1):37–47. doi: 10.1164/rccm.202303-0558WS. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.International Association of Pancreatology Revised Guidelines on Acute Pancreatitis 2025: Supported and Endorsed by the American Pancreatic Association, European Pancreatic Club, Indian Pancreas Club, and Japan Pancreas Society. Pancreatology. 2025;25(6):770–814. doi: 10.1016/j.pan.2025.04.020. [DOI] [PubMed] [Google Scholar]
- 17.Spanier BWM, Nio Y, van der Hulst RWM, et al. Practice and yield of early CT scan in acute pancreatitis: A Dutch Observational Multicenter Study. Pancreatology. 2010;10(2–3):222–28. doi: 10.1159/000243731. [DOI] [PubMed] [Google Scholar]
- 18.Liu W-h, Wang T, Yan H-t, et al. Predictors of percutaneous catheter drainage (PCD) after abdominal paracentesis drainage (APD) in patients with moderately severe or severe acute pancreatitis along with fluid collections. PLoS One. 2015;10(2):e0115348. doi: 10.1371/journal.pone.0115348. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Gupta P, Koshi S, Samanta J, et al. Kissing catheter technique for percutaneous catheter drainage of necrotic pancreatic collections in acute pancreatitis. Exp Ther Med. 2020;20(3):2311–16. doi: 10.3892/etm.2020.8897. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Boxhoorn L, van Dijk SM, van Grinsven J, et al. Immediate versus postponed intervention for infected necrotizing pancreatitis. New Engl J Med. 2021;385(15):1372–81. doi: 10.1056/NEJMoa2100826. [DOI] [PubMed] [Google Scholar]
- 21.Ke L, Li G, Mao W, et al. Chinese Acute Pancreatitis Clinical Trials Group (CAPCTG) Early versus delayed catheter drainage for patients with necrotizing pancreatitis and early persistent organ failure (TIMING): A multicenter randomized controlled trial. Intensive Care Med. 2025;51(8):1431–41. doi: 10.1007/s00134-025-08020-x. [DOI] [PubMed] [Google Scholar]
- 22.Kohli P, Gupta V, Kochhar R, et al. Lavage through percutaneous catheter drains in severe acute pancreatitis: Does it help?A randomized control trial. Pancreatology. 2019;19(7):929–34. doi: 10.1016/j.pan.2019.09.003. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data used in this analysis may be made available from the corresponding author upon reasonable request.