Abstract
Background
Peritoneal dialysis-related peritonitis (PDRP) is a common complication of peritoneal dialysis that can lead to inadequate dialysis, technical failure, or death. The distinct clinical characteristics and outcomes of gram-negative, gram-positive, and culture-negative PDRP remain unclear, limiting the development of optimal management strategies and prognostic assessment for these subtypes.
Methods
This multicenter retrospective study included 156 hospitalized patients with PDRP from three centers between January 2020 and June 2025. Demographic, clinical, laboratory, and treatment data were collected and analyzed.
Results
In the early phase of PDRP, patients with gram-negative infections showed more severe inflammation than those with gram-positive or culture-negative infections, including higher peritoneal dialysate white blood cell counts on day 1, 3, and 5 (P < 0.05); higher C-reactive protein (142.4 vs. 95.1 vs. 77.1 mg/L; P < 0.05) and procalcitonin (8.9 vs. 1.7 vs. 1.2 ng/mL; P < 0.05) levels, altered iron metabolism (higher serum ferritin: 425 vs. 388 vs. 223 ng/mL, P < 0.05; lower serum iron: 4.6 vs. 7.7 vs. 6.2 ng/mL, P < 0.05); and consistently higher abdominal pain scores from days 2 to 12 (4.5 vs. 3.8 vs. 4.0 to 1.0 vs. 0.4 vs. 0.3, P < 0.05). Mortality did not differ significantly among the three groups. However, treatment failure, defined as catheter removal or death, was significantly more frequent in the gram-negative group than in the other two groups (22.7% vs. 10.9% vs. 0%; P < 0.05). Multivariable analysis revealed that hypoalbuminemia was the only factor associated with mortality, whereas gram-negative infection, elevated C-reactive protein, and hypoalbuminemia were associated with treatment failure.
Conclusions
Gram-negative infections are associated with greater inflammation, more severe pain, and higher rates of treatment failure. Although mortality did not differ among the three groups, treatment failure was related to infection type, inflammatory burden, and nutritional status. Further prospective studies are required to confirm these findings.
Keywords: Peritoneal dialysis-related peritonitis, Gram-negative, Gram-positive, Culture-negative, Abdominal pain, Treatment failure
Background
Peritoneal dialysis-related peritonitis (PDRP) is the most common complication of peritoneal dialysis (PD) and may cause substantial morbidity, including pain, gastrointestinal dysfunction, peritoneal membrane injury, adhesions, treatment failure, and death [1]. Once PDRP develops, prompt and effective treatment is essential to reduce the risk of treatment failure.
The clinical features and outcomes of PDRP vary substantially according to the causative organism. Gram-positive cocci, particularly Staphylococcus aureus, are frequently implicated [2], whereas gram-negative peritonitis (GNP) is consistently associated with more severe outcomes, including higher rates of hospitalization, catheter removal, and mortality [3, 4]. Culture-negative peritonitis is also common yet remains poorly characterized. It is often associated with prior antibiotic use or technical factors and generally appears to have a better prognosis than GNP [5, 6]. However, few studies have directly compared gram-negative, gram-positive, and culture-negative peritonitis within the same cohort, limiting our understanding of their distinct clinical profiles.
Recent International Society for Peritoneal Dialysis (ISPD) guidelines emphasize assessing patients’ overall clinical status rather than relying on isolated local indicators [7]. However, in routine practice, management often focuses on the causative organism and antibiotic selection, whereas broader clinical parameters, such as abdominal pain, nutritional status, and systemic inflammation, are evaluated less consistently.
Accordingly, this multicenter retrospective study investigated the clinical characteristics of PDRP across different infection types, with particular attention to the pattern and clinical relevance of abdominal pain. Clarifying these features may support a more integrated assessment of PDRP and help inform future patient-care strategies.
Methods
Patients
This retrospective, episode-based study included patients with PDRP who were hospitalized at Shanghai Sixth People’s Hospital, Shanghai Fengxian District Central Hospital, and Lushan People’s Hospital between January 2020 and June 2025. Each hospitalization was counted as a separate episode. Patients were routinely admitted to these three participating centers for inpatient management if they had symptoms suggestive of PD-related peritonitis, such as abdominal pain, fever, or cloudy effluent, and a peritoneal dialysate white blood cell count (WBC) > 100/µL, regardless of disease severity. PDRP was diagnosed according to the 2022 ISPD criteria. The inclusion criteria were as follows: (1) age ≥ 18 years, regardless of sex; (2) maintenance PD for over 3 months with a stable dialysis regimen; (3) confirmed PDRP diagnosis; and (4) complete laboratory and clinical data. The exclusion criteria were as follows: (1) incomplete medical records; (2) absence of peritoneal dialysate culture; (3) catheter-related peritonitis, pre-PD peritonitis, or PD catheter insertion–related peritonitis; and (4) noninfectious peritonitis, including eosinophilic or chemical peritonitis induced by agents such as icodextrin.
This retrospective study was approved by the Ethics Committee of Shanghai Sixth People’s Hospital (2016 − 113), and written informed consent was obtained from all participants.
Research methods
Demographic and clinical data were collected, including (1) baseline characteristics: sex, age, body weight, primary renal disease, dialysis duration, dialysis modality, peritoneal transport status, previous peritonitis, and exit-site care; (2) PDRP-related findings, including abdominal pain, fever, blood pressure, peritoneal dialysate culture results, and dialysate WBC on days 1, 3, and 5; (3) treatment-related variables, including antibiotic regimen (intraperitoneal or intravenous [IV]), hemodialysis use, probiotics administration, clinical severity assessment, and outcomes including cure, catheter removal or death; and (4) laboratory parameters before and after treatment, including hemoglobin, blood WBC count, blood lymphocyte count, blood neutrophil count, procalcitonin (PCT), C-reactive protein (CRP), serum albumin, serum globulin, total protein, albumin/globulin ratio, prealbumin, alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (AKP), serum creatinine, uric acid, parathyroid hormone (PTH), serum calcium, phosphate, iron, total iron binding capacity (TIBC), transferrin saturation (TSAT), and serum ferritin (SF). Fever was defined as a tympanic temperature of > 38 °C. Baseline mean arterial pressure (MAP) was calculated as diastolic blood pressure plus one-third of the pulse pressure (systolic minus diastolic pressure).
During hospitalization, all PD procedures were performed manually. Treatment protocols included continuous ambulatory peritoneal dialysis (CAPD), consisting of 3–4 daily exchanges with an overnight dwell for the final exchange, and day ambulatory peritoneal dialysis (DAPD), comprising 3–4 daily exchanges without overnight retention of the final exchange. Commercially available 1.5% or 2.5% low-calcium PD solutions (2 L/bag; Baxter International, Inc., Deerfield, IL, USA) were used. Peritoneal membrane transport characteristics were assessed using the standard peritoneal equilibrium test (PET) performed within 3 months of hospitalization. The 4-h dialysate-to-plasma creatinine ratio (D/P Cr) was classified as follows [8]: low, 0.34–0.49; low-average, 0.50–0.64; high-average, 0.65–0.81; and high, 0.82–1.03. Residual renal function and Kt/V data were collected within 3 months prior to the peritonitis episode and calculated using standard formulas [9].
Abdominal pain was assessed using the visual analog scale (VAS) [10]: 0, no pain; 1–2, mild pain; 3–4, moderate pain; 5–6, significant pain; 7–8, severe pain; 9–10, excruciating pain. Scores of 1–3 were assessed daily, 4–6 every 8 h, and 7–10 every 4 h; average values were calculated across repeated assessments. Once the pain score decreased to ≤ 3, assessment returned to once daily. In routine practice, intraperitoneal lidocaine (2%) was used for moderate pain (VAS 4–6), and indomethacin suppositories were used for significant pain and severe pain (VAS 7–10). Intraperitoneal lidocaine was administered as needed at 5 mL per dose, no more than twice a day.
The primary outcome was treatment failure, defined as catheter removal or PDRP-related death [11]. Pain trajectories were analyzed as a secondary outcome. Pain scores, death, and catheter removal were assessed during the same hospitalization. Pain was evaluated daily from admission until discharge, catheter removal, or death, whichever occurred first, for up to 14 days. Pain scores after day 14, catheter removal, or death were excluded. Longitudinal data were analyzed using a linear mixed-effects model, with time treated as a categorical variable from days 1 to 14 or until the occurrence of an event. Missing data due to early discharge, death, or catheter removal were handled under the missing-at-random assumption.
Peritoneal dialysate sample collection methods: (1) For the initial diagnostic sample, because the time at which patients first noticed cloudy effluent could vary, the first diagnostic sample was collected after a minimum 4‑h dwell, according to the sampling requirements; (2) For follow‑up assessments on day 3 and 5, samples were obtained from the overnight dwell in hospitalized patients, providing a standardized dwell time (typically 10–12 h). Dialysate samples were collected aseptically at all three centers. The injection port of the dialysis bag was disinfected with iodine, and 10–20 mL of effluent was aspirated with a sterile syringe when peritoneal symptoms were first suspected. Each sample was inoculated into aerobic and anaerobic blood culture bottles (one bottle each) and incubated using the BACTEC blood culture system (BD BACTEC™, Becton Dickinson, Franklin Lakes, NJ, USA). Species identification was then performed using standard microbiological techniques, including the BD Phoenix™ automated system, in accordance with each center’s laboratory practice. Preliminary positive culture results were reported within 48 h, and final negative results were reported after 5 days of incubation in accordance with standard protocols. Episodes were categorized as culture-positive or culture-negative. Samples were generally collected before antibiotic initiation.
The initial empirical antibiotic therapy comprised a first-generation cephalosporin or vancomycin, combined with a third-generation cephalosporin or an aminoglycoside. Antibiotics were administered primarily via the intraperitoneal route, with IV therapy used selectively at the clinician’s discretion. Oral fluconazole prophylaxis was administered according to local practice. In China, antibiotics are classified as nonrestricted, restricted, or special-use agents [12]. Empirical antibiotics were selected based on the patients’ clinical condition, institutional antimicrobial stewardship principles, and ISPD guideline recommendations. Clinical response was assessed daily. Within the first 3 days of observation, empirical therapy could be adjusted earlier based on the basis of the preliminary 48-h culture results, if clinically indicated, without waiting for antimicrobial susceptibility testing. Treatment decisions were primarily guided by the clinical response. If necessary, antibiotics were further adjusted according to the day‑5 culture and susceptibility results in consultation with pharmacists.
Definition
According to the 2022 ISPD guidelines [13], PDRP was diagnosed when at least two of the following criteria were met: (1) abdominal pain and/or cloudy ascites; (2) dialysate WBC > 0.1 × 109/L or > 100/µL after a dwell time of at least 2 h, with > 50% polymorphonuclear neutrophils; and (3) a positive dialysate culture. Complete relief was defined as resolution of the clinical symptoms of peritonitis, clear peritoneal effluent, and a dialysate WBC count < 0.1 × 109/L. PDRP-related death was defined as death within 30 days of peritonitis onset or death due to peritonitis during hospitalization. Peritonitis-associated catheter removal was defined as the removal of the PD catheter as part of the treatment for an active episode of peritonitis. Catheter-related peritonitis was defined as peritonitis occurring within 3 months of catheter infection, either exit-site or tunnel infection, with the same organism identified at the exit site or from a tunnel collection and in effluent or another sterile site in the context of antibiotic exposure. Pre-PD peritonitis was defined as peritonitis occurring after PD catheter insertion and before PD initiation. PD catheter insertion–related peritonitis was defined as peritonitis occurring within 30 days of PD catheter insertion [12].
Statistical methods
Data and figures were analyzed and figures were generated using SPSS version 26.0 (IBM Corp., Armonk, NY, USA) and R version 4.0.2 (R Foundation for Statistical Computing, Vienna, Austria). Categorical variables are presented as frequency (percentage), whereas continuous variables are expressed as mean ± standard deviation (x̄ ± s) for normally distributed data or median (interquartile range) for non-normally distributed data. Data preprocessing included data verification and assessment of normality testing using the Kolmogorov–Smirnov test. Normally distributed variables were compared using analysis of variance (ANOVA), whereas non-normally distributed variables were compared using the Kruskal–Wallis test. Categorical variables were compared using the χ2 test or Fisher’s exact test, as appropriate.
To longitudinally compare differences in pain trajectories among patients with different types of infection, a linear mixed effects model was employed to analyze repeatedly measured pain scores, with shaded bands representing 95% confidence intervals (CIs). The number of patients with available pain assessments at each time point was displayed below the x-axis.
Spearman’s rank correlation coefficient was used to assess associations between VAS pain scores, and peritoneal dialysate WBC counts on days 1, 3, and 5, both overall and within each infection subgroup. Univariate and multivariable logistic regression analyses were performed to evaluate the associations among biochemical parameters, infection type, and clinical outcomes. Results are reported as odds ratios (ORs) with 95% CIs. Statistical significance was defined as a two-tailed P-value of < 0.05.
Results
Baseline characteristics of patients with PDRP
In total, 162 PD-related peritonitis episodes were screened. After excluding cases with incomplete medical records (n = 1) or fungal peritonitis (n = 5), 156 episodes were included in the final analysis: 44 gram-negative, 92 gram-positive, and 20 culture-negative cases (Fig. 1). Among the gram-negative isolates, Escherichia coli was the most common (n = 16), followed by Klebsiella spp. (n = 7), Enterobacteriaceae (n = 6), Acinetobacter spp. (n = 4), Pseudomonas aeruginosa (n = 3), and other gram-negative species (n = 8). Among the gram-positive isolates, coagulase-negative staphylococci were the most common (n = 48), followed by Streptococcus spp. (n = 24), Staphylococcus aureus (n = 6), Enterococcus spp. (n = 6), and other gram-positive rods (n = 8). Age, body weight, dialysis duration, sex, diabetes, hypertension, history of peritonitis, dialysis mode, peritoneal transport type, Kt/V, and MAP did not differ significantly among the three groups. However, the proportion of patients presenting with fever differed significantly (P < 0.05; Table 1).
Fig. 1.
Flow diagram showing peritoneal dialysis-related peritonitis in different bacterial infections. PDRP, peritoneal dialysis-related peritonitis
Table 1.
General information of patients with peritoneal dialysis-related peritonitis and different infection types
| Variable | Total (n = 156) |
Gram-negative (n = 44) |
Gram-positive (n = 92) |
Culture-negative (n = 20) |
P value |
|---|---|---|---|---|---|
| Age, years | 63 (53, 70) | 66 (54, 71) | 61 (52, 67) | 66 (56, 71) | 0.161# |
| Body weight (kg) | 64.8 ± 11.9 | 63.8 ± 12.1 | 65.4 ± 10.4 | 64.6 ± 17.2 | 0.760* |
| Dialysis duration, months | 42.1 (20.7, 71.9) | 29.9 (18.9, 68.5) | 45.5 (25.1, 76.9) | 41.9 (16.9, 88.9) | 0.579# |
| Male, n (%) | 98 (62.8) | 29 (65.9) | 59 (64.1) | 10 (50.0) | 0.437& |
| Diabetes, n (%) | 53 (33.9) | 12 (27.3) | 35 (38.0) | 6 (30.0) | 0.427& |
| Hypertension, n (%) | 82 (52.6) | 25 (56.8) | 47 (51.1) | 10 (50.0) | 0.797& |
| Previous peritonitis episodes, n (%) | 54 (34.6) | 10 (22.7) | 34 (37.0) | 10 (50.0) | 0.252& |
| Exit-site infection, n (%) | 3 (3.3) | 0 | 3 (3.3) | 0 | 0.703& |
| Dialysis mode, n (%) | 0.676& | ||||
| CAPD | 148 (94.9) | 41 (93.2) | 87 (94.6) | 20 (100.0) | |
| DAPD | 8 (5.1) | 3 (6.8) | 5 (5.4) | 0 | |
| Peritoneal transport type, n (%) | 0.277& | ||||
| Low | 32 (20.5) | 11 (25.0) | 17 (18.5) | 4 (20.0) | |
| Low average | 54 (34.6) | 16 (36.4) | 34 (37.0) | 4 (20.0) | |
| High average | 51 (32.7) | 14 (31.8) | 29 (31.5) | 8 (40.0) | |
| High | 19 (12.2) | 3 (6.8) | 12 (13.0) | 4 (20.0) | |
| Kt/V | 1.60 ± 0.38 | 1.57 ± 0.32 | 1.61 ± 0.40 | 1.63 ± 0.38 | 0.121* |
| 1st day ultrafiltration volume (mL) | 470 (55,1115) | 310 (55,730) | 522 (110,1115) | 585 (150,1050) | 0.071# |
| Fever, n (%) | 26 (16.7) | 14 (31.8) | 9 (9.8) | 3 (15.0) | 0.005& |
| MAP (mmHg) | 92 (79, 103) | 87 (73, 103) | 93 (80, 106) | 94 (85, 103) | 0.491# |
CAPD, continuous ambulatory peritoneal dialysis; DAPD, day ambulatory peritoneal dialysis; MAP, mean arterial pressure
# Kruskal–Wallis test was used for non-normally distributed variables; * Analysis of variance (ANOVA) was used for normally distributed variables; & χ2 test or Fisher’s exact test was used for categorical variables
Baseline dialysate WBC, blood lymphocyte counts and proportion, blood neutrophil proportion, CRP, PCT, serum iron, TIBC, and SF also differed significantly among the groups (P < 0.05). Patients with gram-negative infections had higher dialysate WBC counts, greater inflammatory activity, higher ferritin levels, and lower serum iron levels (Table 2).
Table 2.
Baseline indicators of patients with peritoneal dialysis-related peritonitis and different infection types
| Indicator | Gram-negative (n = 44) |
Gram-positive (n = 92) |
Culture-negative (n = 20) |
P value |
|---|---|---|---|---|
| Peritoneal dialysate WBC (×106/L) | ||||
| Day 1 | 4177 (1088,11266) | 1656 (729,4585) | 983 (307,1885) | 0.002# |
| Day 3 | 478 (59,2878) | 93 (34,404) | 163 (38,375) | 0.002# |
| Day 5 | 159 (49,1009) | 36 (12,107) | 73 (23,199) | < 0.001# |
| Hemoglobin (g/L) | 102.9 ± 17.4 | 102.2 ± 19.6 | 111.7 ± 22.9 | 0.145* |
| Blood WBC (×109/L) | 10.4 ± 4.5 | 9.2 ± 4.6 | 9.3 ± 2.6 | 0.300* |
| Blood lymphocyte count (×109/L) | 0.8 ± 0.4 | 1.1 ± 0.6 | 1.1 ± 0.5 | 0.019* |
| Blood lymphocyte proportion (%) | 8.6 ± 4.3 | 13.5 ± 8.2 | 12.7 ± 5.3 | < 0.001* |
| Blood neutrophil count (×109/L) | 9.0 ± 4.5 | 7.5 ± 4.5 | 7.3 ± 2.5 | 0.125* |
| Blood neutrophil proportion (%) | 83.6 ± 7.1 | 77.2 ± 10.3 | 75.1 ± 16.0 | 0.001* |
| CRP (mg/L) | 142.4 (81.1,196.4) | 95.1 (45.6,147.8) | 77.1 (49.8,97.9) | < 0.001# |
| Procalcitonin (ng/mL) | 8.9 (1.9,29.3) | 1.7 (0.7,12.3) | 1.2 (0.4,9.9) | 0.001# |
| Albumin (g/L) | 28.1 ± 6.6 | 27.8 ± 6.1 | 28.9 ± 6.7 | 0.793* |
| Globulin (g/L) | 29.8 ± 5.8 | 29.8 ± 5.9 | 29.7 ± 6.7 | 0.996* |
| Total protein (g/L) | 57.8 ± 10.6 | 57.5 ± 9.3 | 58.6 ± 11.2 | 0.904* |
| Albumin/globulin ratio | 1.0 ± 0.2 | 1.0 ± 0.3 | 1.1 ± 0.2 | 0.371* |
| Prealbumin (mg/L) | 189.3 ± 83.6 | 199.0 ± 77.6 | 214.0 ± 57.1 | 0.492* |
| ALT (U/L) | 14 (10,19) | 13 (9,17) | 13 (10,16) | 0.662# |
| AST (U/L) | 19 (11,25) | 16 (11,23) | 19 (13,23) | 0.899# |
| AKP (U/L) | 79 (63,109) | 82 (61,110) | 75 (65,101) | 0.861# |
| Serum creatinine (µmol/L) | 852 ± 298 | 857 ± 252 | 756 ± 272 | 0.306* |
| Uric acid (µmol/L) | 375 ± 111 | 350 ± 99 | 337 ± 99 | 0.292* |
| Serum calcium (mmol/L) | 2.14 ± 0.22 | 2.16 ± 0.25 | 2.16 ± 0.24 | 0.912* |
| Serum phosphorus (mmo/L) | 1.62 ± 0.55 | 1.53 ± 0.43 | 1.33 ± 0.34 | 0.071* |
| PTH (pg/mL) | 182 (106,343) | 202 (99,394) | 142 (86,244) | 0.199# |
| Serum iron (ng/mL) | 4.6 (3.1,7.5) | 7.7 (4.3,10.9) | 6.2 (4.6,11.1) | 0.008# |
| TIBC (µmol/L) | 29.1 ± 9.1 | 31.9 ± 11.6 | 39.3 ± 11.3 | 0.003* |
| TSAT (%) | 19.5 ± 12.4 | 24.4 ± 12.8 | 20.3 ± 11.7 | 0.121* |
| SF (ng/mL) | 425 (288,650) | 388 (192,654) | 223 (109,418) | 0.043# |
WBC, white blood cell count; CRP, C-reactive protein; ALT, alanine aminotransferase; AST, aspartate aminotransferase; AKP, alkaline phosphatase; PTH, parathyroid hormone; TIBC, total iron-binding capacity; TSAT, transferrin saturation; SF, serum ferritin
# Kruskal–Wallis test was used for non-normally distributed variables; * Analysis of variance (ANOVA) was used for normally distributed variables
Changes in abdominal pain among patients with PDRP
Patients received intraperitoneally 2% lidocaine when pain scores were 4–6 and indomethacin suppositories when scores exceeded 7. Pain trajectory analysis revealed that pain scores decreased over time in all three infection groups (Fig. 2). From day 2 through day 12, pain scores were significantly higher in the gram-negative group than in the gram-positive and culture-negative groups (4.5 vs. 3.8 vs. 4.0 to 1.0 vs. 0.4 vs. 0.3, P < 0.05 for both comparisons). No significant differences were observed between the gram-positive and culture-negative groups (Table 3).
Fig. 2.
Pain trajectories in patients with peritoneal dialysis-related peritonitis according to infection type. Lines represent least-squares mean VAS pain scores estimated using the linear mixed-effects model, and shaded bands represent 95% confidence intervals. Censoring events due to catheter removal or death are indicated. Pain scores after catheter removal or death were not included in the longitudinal analysis. VAS, visual analog scale
Table 3.
Pain scores for patients with peritoneal dialysis-related peritonitis with different infection types
| Day | Gram-negative (n = 44) |
Gram-positive (n = 92) |
Culture-negative (n = 20) |
P value Culture-negative vs. Gram-positive | P value Gram-negative vs. Gram-positive |
|---|---|---|---|---|---|
| 1 | 5.2 (4.8, 5.6) | 4.8 (4.6, 5.1) | 4.6 (4.1, 5.2) | 0.579 | 0.119 |
| 2 | 4.5 (4.1, 4.9) | 3.8 (3.6, 4.1) | 4.0 (3.4, 4.6) | 0.629 | 0.005 |
| 3 | 4.2 (3.8, 4.6) | 3.1 (2.9, 3.4) | 3.4 (2.8, 4.0) | 0.424 | < 0.001 |
| 4 | 3.8 (3.4, 4.2) | 2.7 (2.4, 3.0) | 2.8 (2.3, 3.4) | 0.658 | < 0.001 |
| 5 | 3.5 (3.1, 3.9) | 2.2 (1.9, 2.5) | 2.2 (1.6, 2.8) | 0.937 | < 0.001 |
| 6 | 2.9 (2.5, 3.3) | 1.8 (1.5, 2.1) | 1.8 (1.2, 2.4) | 0.992 | < 0.001 |
| 7 | 2.6 (2.2, 3.0) | 1.4 (1.2, 1.7) | 1.6 (0.9, 2.2) | 0.765 | < 0.001 |
| 8 | 2.1 (1.7, 2.6) | 1.1 (0.8, 1.4) | 1.3 (0.6, 1.9) | 0.655 | < 0.001 |
| 9 | 1.7 (1.3, 2.1) | 0.9 (0.6, 1.2) | 1.1 (0.5, 1.8) | 0.587 | 0.003 |
| 10 | 1.4 (1.0, 1.9) | 0.8 (0.5, 1.1) | 0.8 (0.1, 1.5) | 0.994 | 0.019 |
| 11 | 1.3 (0.9, 1.8) | 0.6 (0.3, 0.9) | 0.6 (-0.2, 1.3) | 0.880 | 0.008 |
| 12 | 1.0 (0.6, 1.4) | 0.4 (0.1, 0.8) | 0.3 (-0.4, 1.0) | 0.788 | 0.040 |
| 13 | 0.7 (0.2, 1.1) | 0.3 (-0.0, 0.7) | -0.1 (-0.8, 0.7) | 0.330 | 0.222 |
| 14 | 0.4 (-0.1, 0.8) | 0.2 (-0.2, 0.6) | -0.1 (-0.9, 0.7) | 0.509 | 0.524 |
Data are presented as least-squares means with 95% confidence intervals (CIs)
P-values are calculated using pairwise comparisons, with the gram-positive group as a reference in the linear mixed-effects model
Additional correlation analyses revealed that VAS pain scores were positively associated with dialysate WBC counts on days 1, 3, and 5, supporting the clinical relevance of abdominal pain monitoring in patients with PDRP. These findings suggest that pain severity may partially reflect the degree of local peritoneal inflammation (ρ = 0.581–0.744, all P < 0.001). This association was particularly strong in patients with gram-negative peritonitis (ρ = 0.904), consistent with the more severe inflammatory phenotype observed in this subgroup (Table 4).
Table 4.
Correlations between dialysate white blood cell count and visual analog scale pain score on days 1, 3, and 5
| Time | All patients (n = 156) |
Gram-negative (n = 44) |
Gram-positive (n = 92) |
Culture-negative (n = 20) |
||||
|---|---|---|---|---|---|---|---|---|
| ρ | P | ρ | P | ρ | P | ρ | P | |
| Day 1 | 0.581 | < 0.001 | 0.677 | < 0.001 | 0.462 | < 0.001 | 0.665 | 0.001 |
| Day 3 | 0.744 | < 0.001 | 0.837 | < 0.001 | 0.647 | < 0.001 | 0.524 | 0.018 |
| Day 5 | 0.701 | < 0.001 | 0.904 | < 0.001 | 0.485 | < 0.001 | 0.792 | < 0.001 |
ρ, Spearman’s correlation coefficient
Factors associated with mortality and treatment failure in patients with PDRP
Mortality did not differ significantly among the three groups, whereas treatment failure differed significantly (P < 0.05; Table 5). Catheter removal was attributed to refractory peritonitis (n = 9), recurrent peritonitis (n = 2), and repeat episodes (n = 1).
Table 5.
Treatment outcomes in patients with peritoneal dialysis-related peritonitis and different infection types
| Variable | Gram-negative (n = 44) |
Gram-positive (n = 92) |
Culture-negative (n = 20) |
P value |
|---|---|---|---|---|
| Hospital stay, days | 15.0 (9.3–20.0) | 9.5 (7.0–14.0) | 9.5 (7.0–14.8) | 0.005# |
| Outcome, n (%) | ||||
| Treatment failure | 10 (22.7) | 10 (10.9) | 0 | 0.011& |
| Death | 4 (9.1) | 4 (4.3) | 0 | 0.183& |
| Treatment, n (%) | ||||
| Temporary hemodialysis | 10 (22.7) | 13 (14.1) | 0 | 0.015& |
| Intravenous antibiotic use | 35 (79.5) | 45 (48.9) | 6 (30.0) | < 0.001& |
| Intraperitoneal antibiotic change | 20 (45.5) | 30 (32.6) | 5 (25.0) | 0.201& |
| Probiotics | 19 (43.2) | 40 (43.5) | 5 (25.0) | 0.296& |
# Kruskal–Wallis test was used for non-normally distributed variables; & χ2 test or Fisher’s exact test was used for categorical variables
Multivariable logistic regression models were used to identify factors associated with mortality and treatment failure. Due to the limited number of outcome events (n = 20), the final models were adjusted for age, sex, and dialysis duration to reduce overfitting. Candidate variables were selected based on their clinical relevance and potential to predict peritonitis outcomes, including gram-negative infection, blood WBC, CRP, PCT, TIBC, and albumin. Hypoalbuminemia was the only factor associated with mortality (OR: 0.79, 95% CI: 0.63–0.95; P = 0.020). Treatment failure was associated with gram-negative infection (OR: 3.23, 95% CI: 1.14–9.24; P = 0.026), higher CRP levels (OR: 1.01, 95% CI: 1.00–1.01; P = 0.008), and lower albumin levels (OR: 0.89, 95% CI: 0.81–0.97; P = 0.012; Table 6).
Table 6.
Multivariable analysis of factors associated with mortality and treatment failure
| Outcome | Variable | OR (95%CI) | P |
|---|---|---|---|
| Mortality | Albumin | 0.79 (0.63–0.95) | 0.020 |
| Treatment failure | C-reactive protein | 1.01 (1.00–1.01) | 0.008 |
| Albumin | 0.89 (0.81–0.97) | 0.012 | |
| Gram-negative infection | 3.23 (1.14–9.24) | 0.026 |
Models were adjusted for age, sex, and dialysis duration
Discussion
Comparative analysis of the gram-negative, gram-positive, and culture-negative PDRP subgroups revealed distinct clinical profiles. Patients with gram-negative infections had more pronounced local and systemic inflammation at onset, reflected by higher peritoneal dialysate leukocyte counts, higher blood neutrophil percentages, and increased CRP and procalcitonin levels. These findings are consistent with previous reports suggesting that gram-negative PDRP is generally more severe, whereas gram-positive and culture-negative episodes often exhibit milder features [14–18]. Gram-negative infections were also associated with lower serum free iron and higher stored iron levels in the early phase, potentially reflecting a stronger inflammatory response [19, 20]. Overall, these findings emphasize the clinical severity of gram-negative PDRP and the need for heightened clinical awareness.
Abdominal pain is a key symptom of peritonitis that often reflects the severity of systemic inflammation and influences clinical decision-making, including the need for hospitalization [12]. Despite its significance, pain management has received limited attention in the context of PDRP, with most studies focusing on acute abdominal conditions or gastrointestinal disorders [21–23]. In the present study, patients with gram-negative infections reported consistently higher abdominal pain scores from days 2 to 12 than those with gram-positive or culture-negative infections. Pain intensity generally decreased alongside clinical improvement and timely analgesic intervention, suggesting that serial pain assessment may help reflect disease progression and overall clinical status. This study provides an initial graphical description of abdominal pain patterns in patients with PDRP and highlights the potential value of pain monitoring in clinical assessment. These preliminary observations provide insights into symptom presentation and support the incorporation of pain assessment into holistic, goal-directed care for patients with PDRP.
Mortality was independently associated only with hypoalbuminemia, suggesting that host nutritional status may be a stronger determinant of death than infection-related factors in this population. In contrast, treatment failure—a composite outcome including catheter removal—was associated with gram-negative infection, systemic inflammation reflected by CRP, and hypoalbuminemia. This suggests that gram-negative infections may contribute to adverse outcomes through multiple pathways and that a composite endpoint may be more sensitive for detecting clinically meaningful differences when number of events is limited. Preventing treatment failure in PDRP may therefore require early control of infection before local inflammation progresses to systemic involvement. Worsening systemic status, reflected by inflammation or malnutrition, often indicates increased risk of adverse outcomes. Gram-negative bacteria may serve as an early warning signal; however, as the disease progresses, clinical attention should shift from initial microbiological classification to continuous assessment of the patient’s overall condition. Early intervention, timely treatment, and individualized risk assessment may be essential for preventing treatment failure, consistent with the goal-directed management approach advocated by the ISPD [24–26].
However, data on antimicrobial resistance, including extended-spectrum beta-lactamase (ESBL), carbapenem-resistant Enterobacteriaceae (CRE), and methicillin-resistant Staphylococcus aureus (MRSA), were not systematically available in this retrospective study. Given that resistant organisms are associated with higher rates of catheter removal and treatment failure in PDRP, the independent association between gram-negative infection and treatment failure should be interpreted cautiously. This association may partly reflect underlying resistance patterns rather than gram-negative status itself. Antibiotic management also varied considerably after initial guideline-based therapy, with subsequent adjustments guided by clinical judgment, culture results, and local practice. These real-world variations, including differences in route, dose, duration, and de-escalation strategy, could not be reliably captured or analyzed. Therefore, differences in antibiotic use may have contributed to the observed association between gram-negative infections and treatment failure.
Understanding culture-negative PDRP remains limited. Negative bacterial cultures may result from prior antibiotic exposure, specimen processing issues, laboratory limitations, or atypical pathogens [13]. Some culture-negative cases may represent gram-positive peritonitis masked by prior antibiotic use, a common limitation of retrospective studies. However, direct pairwise comparisons were not performed, and the number of culture-negative samples was small (n = 20). These findings should therefore be considered preliminary and exploratory. They suggest that culture-negative peritonitis may not always be milder and may share some clinical features with gram-positive infections; however, further validation is needed [27]. Studies directly comparing gram-positive, gram-negative, and culture-negative peritonitis in patients undergoing PD remain limited. This three-way comparison provides a more integrated view of their distinct clinical characteristics and may inform future research and management strategies.
At the three centers in the present study, all patients with suspected or confirmed PD-related peritonitis were routinely hospitalized, regardless of disease severity. This differs from practice in some countries, where only patients with severe diseases are hospitalized, and should be considered when comparing these findings with those from other settings. Inpatient care enables continuous monitoring by dedicated PD nurses and timely physician-led treatment adjustments, which may improve pain management and overall care. This may also partially explain the higher-than-guideline use of IV antibiotics in patients with gram-negative peritonitis. In our experience, patients with PDRP generally prefer to continue PD, whereas those with gram-negative infections often present with more severe systemic features, including elevated WBC counts and CRP levels. Hospitalization allows for intensive management, and IV antibiotics are initially administered to achieve rapid therapeutic levels and ensure reliable drug delivery. Therefore, IV antibiotics were used according to clinical severity and institutional practice, differing from ISPD guidance; this has been acknowledged as a limitation. In patients with early clinical deterioration, catheter removal was performed without waiting until day 5, which is consistent with the ISPD recommendation that catheter removal should be performed earlier if the patient’s clinical condition is deteriorating. The individualized decision-making in our practice—taking into account symptom onset, prior antibiotic exposure, and clinical trajectory—is therefore consistent with the guideline’s emphasis on integrating clinical judgment rather than relying on a rigid time cutoff.
This study had some limitations that should be considered. The generalizability of these findings is limited by the relatively small sample size and retrospective design. However, PD-related peritonitis is generally uncommon, with the ISPD recommending an overall rate of no more than 0.40 episodes per patient-year at risk [12]. The small sample size and limited number of outcome events prevented adjustment for center effects and other potential confounders. Detailed antimicrobial susceptibility profiles were not systematically documented, and data on preculture antibiotic exposure, specimen transport times, and center-level culture-negativity or contamination rates were not recorded or audited. Although the three centers belonged to the same hospital group and shared clinical protocols, treatment varied considerably across patients and centers, leaving residual confounding that could not be fully adjusted. Antibiotic escalation and de-escalation practices were heterogeneous and could not be reliably analyzed. Additionally, dwell time was standardized for follow-up samples but differed for the initial diagnostic sample, which should be considered when interpreting changes across days. Due to the retrospective design, the frequency of VAS monitoring and the use of analgesic strategies could not be verified. In addition, pain assessments and analgesic use were not precisely time-aligned, preventing adjustment for this potential confounder in the pain trajectory analysis. Only three patients (< 2% of all episodes) experienced more than one peritonitis episode during the study period; thus, repeated episodes were not statistically adjusted for. Although the linear mixed-effects model incorporated all available pain measurements, later time points may have been affected by informative censoring due to discharge, catheter removal, or death. Therefore, pain trajectory results should be interpreted as exploratory.
Conclusions
This study characterized inflammatory and nutritional status as well as abdominal pain patterns in patients with gram-negative, gram-positive, and culture-negative PDRP. Gram-negative infections appear to be associated with stronger inflammatory responses and greater pain severity. Although gram-negative infection was not associated with mortality, it was associated with treatment failure as well as elevated CRP and hypoalbuminemia. These findings have clinical value, but their potential to guide management warrants further investigation through prospective studies.
Acknowledgements
We sincerely thank Professor Niansong Wang for his guidance and support of this research. We would like to thank Editage (www.editage.cn) for the English language editing.
Abbreviations
- AKP
Alkaline phosphatase
- ALT
Alanine aminotransferase
- ANOVA
Analysis of variance
- AST
Aspartate aminotransferase
- CAPD
Continuous ambulatory peritoneal dialysis
- CI
Confidence interval
- CRE
Carbapenem-resistant enterobacteriaceae
- CRP
C-reactive protein
- DAPD
Day ambulatory peritoneal dialysis
- ESBL
Extended-spectrum beta-lactamase
- GNP
Gram-negative peritonitis
- ISPD
International society for peritoneal dialysis
- IV
Intravenous
- MAP
Mean arterial pressure
- MRSA
Methicillin-resistant Staphylococcus aureus
- OR
Odds ratio
- PCT
Procalcitonin
- PD
Peritoneal dialysis
- PDRP
Peritoneal dialysis-related peritonitis
- PET
Peritoneal equilibrium test
- PTH
Parathyroid hormone
- SF
Serum ferritin
- TIBC
Total iron binding capacity
- TSAT
Transferrin saturation
- VAS
Visual analog scale
- WBC
White blood cell count
Author contributions
Yumei Liu: data curation, formal analysis, investigation, methodology, and writing of the original draft. Zejin Zhang: data curation, formal analysis, investigation, methodology, and writing the original draft. Yuqi Liang: formal analysis, investigation, methodology, writing, review, and editing. Xiaoling Zha: data curation, formal analysis, investigation, methodology, writing, review, and editing. Tao Xu: conceptualization, project administration, resources, supervision, validation, writing, review, and editing. All authors reviewed the manuscript.
Funding
None.
Data availability
The data underlying this article will be shared upon reasonable request to the corresponding author.
Declarations
Ethics approval and consent to participate
All procedures were performed in accordance with the ethical standards of the Ethics Committee of Shanghai Sixth People’s Hospital and the 1964 Declaration of Helsinki and its later amendments or comparable ethicalstandards. This retrospective study was approved by the Ethics Committee of the Shanghai Sixth People’s Hospital (2016-113). Written informed consent was obtained from all participants.
Consent for publication
Not applicable. This study used aggregated, anonymized data from medical records. No individual participant data are presented.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Zejin Zhang and Yuqi Liang are co-first authors.
References
- 1.Lin W, Jiang C, Yu H, Wang L, Li J, Liu X, et al. The effects of Fushen granule on the composition and function of the gut microbiota during peritoneal dialysis-related peritonitis. Phytomedicine. 2021;86:153561. [DOI] [PubMed] [Google Scholar]
- 2.Szeto CC, Chow KM, Kwan BC, Law MC, Chung KY, Yu S, et al. Staphylococcus aureus peritonitis complicates peritoneal dialysis: review of 245 consecutive cases. Clin J Am Soc Nephrol. 2007;2:245–51. [DOI] [PubMed] [Google Scholar]
- 3.Tang W, Cho Y, Hawley CM, Badve SV, Johnson DW. The role of monitoring gentamicin levels in patients with gram-negative peritoneal dialysis-associated peritonitis. Perit Dial Int. 2014;34:219–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Szeto CC, Chow KM. Gram-negative peritonitis–the Achilles heel of peritoneal dialysis? Perit Dial Int. 2007;27(Suppl 2):S267–71. [PubMed] [Google Scholar]
- 5.Kocyigit I, Unal A, Karademir D, Bahcebasi S, Sipahioglu MH, Tokgoz B, et al. Improvement in culture-negative peritoneal dialysis-related peritonitis: a single center’s experience. Perit Dial Int. 2012;32:476–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Htay H, Cho Y, Pascoe EM, Hawley C, Clayton PA, Borlace M, et al. Multicentre registry data analysis comparing outcomes of culture-negative peritonitis and different subtypes of culture-positive peritonitis in peritoneal dialysis patients. Perit Dial Int. 2020;40:47–56. [DOI] [PubMed] [Google Scholar]
- 7.Brown EA, Blake PG, Boudville N, Davies S, de Arteaga J, Dong J, et al. International Society for peritoneal dialysis practice recommendations: prescribing high-quality goal-directed peritoneal dialysis. Perit Dial Int. 2020;40:244–53. [DOI] [PubMed] [Google Scholar]
- 8.van Biesen W, Heimburger O, Krediet R, Rippe B, La Milia V, Covic A, et al. Evaluation of peritoneal membrane characteristics: clinical advice for prescription management by the ERBP working group. Nephrol Dial Transpl. 2010;25:2052–62. [DOI] [PubMed] [Google Scholar]
- 9.van Olden RW, Krediet RT, Struijk DG, Arisz L. Measurement of residual renal function in patients treated with continuous ambulatory peritoneal dialysis. J Am Soc Nephrol. 1996;7:745–50. [DOI] [PubMed] [Google Scholar]
- 10.Price DD, McGrath PA, Rafii A, Buckingham B. The validation of visual analogue scales as ratio scale measures for chronic and experimental pain. Pain. 1983;17:45–56. [DOI] [PubMed] [Google Scholar]
- 11.Muthucumarana K, Howson P, Crawford D, Burrows S, Swaminathan R, Irish A. The relationship between presentation and the time of initial administration of antibiotics with outcomes of peritonitis in peritoneal dialysis patients: the PROMPT study. Kidney Int Rep. 2016;1:65–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Xiao Y, Li L. China’s national plan to combat antimicrobial resistance. Lancet Infect Dis. 2016;16:1216–8. [DOI] [PubMed] [Google Scholar]
- 13.Li PK, Chow KM, Cho Y, Fan S, Figueiredo AE, Harris T, et al. ISPD peritonitis guideline recommendations: 2022 update on prevention and treatment. Perit Dial Int. 2022;42:110–53. [DOI] [PubMed] [Google Scholar]
- 14.Sun Y, Chen X, Ni J, Yu J. Peritoneal dialysis peritonitis due to Neisseria: clinicopathological features of 10 patients with a review of the literature. Ren Fail. 2025;47:2466820. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lin WH, Tseng CC, Wu AB, Chang YT, Kuo TH, Chao JY, et al. Clinical and microbiological characteristics of peritoneal dialysis-related peritonitis caused by Escherichia coli in southern Taiwan. Eur J Clin Microbiol Infect Dis. 2018;37:1699–707. [DOI] [PubMed] [Google Scholar]
- 16.Lin WH, Tseng CC, Wu AB, Yang DC, Cheng SW, Wang MC, et al. Clinical and microbiological characteristics of peritoneal dialysis-related peritonitis caused by Klebsiella pneumoniae in southern Taiwan. J Microbiol Immunol Infect. 2015;48:276–83. [DOI] [PubMed] [Google Scholar]
- 17.Zeng Y, Jiang L, Lu Y, Wang Z, Song K, Shen H, et al. Peritoneal dialysis-related peritonitis caused by gram-negative organisms: ten-years experience in a single center. Ren Fail. 2021;43:993–1003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Troidle L, Gorban-Brennan N, Kliger A, Finkelstein F. Differing outcomes of gram-positive and gram-negative peritonitis. Am J Kidney Dis. 1998;32:623–8. [DOI] [PubMed] [Google Scholar]
- 19.Martins AC, Almeida JI, Lima IS, Kapitão AS, Gozzelino R. Iron metabolism and the inflammatory response. IUBMB Life. 2017;69:442–50. [DOI] [PubMed] [Google Scholar]
- 20.Radić J, Bašić-Jukić N, Vujičić B, Klarić D, Radulović G, Jakić M, et al. Anemia is correlated with malnutrition and inflammation in Croatian peritoneal dialysis patients: A multicenter nationwide study. Perit Dial Int. 2017;37:472–5. [DOI] [PubMed] [Google Scholar]
- 21.Sinopoulou V, Gordon M, Akobeng AK, Gasparetto M, Sammaan M, Vasiliou J, et al. Interventions for the management of abdominal pain in Crohn’s disease and inflammatory bowel disease. Cochrane Database Syst Rev. 2021;11:CD013531. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Sinopoulou V, Gordon M, Dovey TM, Akobeng AK. Interventions for the management of abdominal pain in ulcerative colitis. Cochrane Database Syst Rev. 2021;7:CD013589. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Donnelly F, Feo R, Jangland E, Muntlin Athlin Å. The management of patients with acute abdominal pain in the emergency department: a qualitative study of nurse perceptions. Australas Emerg Care. 2019;22:97–102. [DOI] [PubMed] [Google Scholar]
- 24.Mao Y, Xiao D, Deng S, Xue S. Development of a clinical risk score system for peritoneal dialysis-associated peritonitis treatment failure. BMC Nephrol. 2023;24:229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Yu J, Zhu L, Ni J, Tong M, Wang H. Technique failure in peritoneal dialysis-related peritonitis: risk factors and patient survival. Ren Fail. 2023;45:2205536. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Da Luz LG, Ankawi G, Digvijay K, Rosner MH, Ronco C. Technique failure in peritoneal dialysis: etiologies and risk assessment. Blood Purif. 2021;50:42–9. [DOI] [PubMed] [Google Scholar]
- 27.Tanratananon D, Deekae S, Raksasuk S, Srithongkul T. Evaluation of different methods to improve culture-negative peritoneal dialysis-related peritonitis: a single-center study. Ann Med Surg (Lond). 2021;63:102139. [DOI] [PMC free article] [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 underlying this article will be shared upon reasonable request to the corresponding author.