Peritoneal Dialysis–Associated Peritonitis

PD Equipment and Training.

At least four randomized, controlled trials support the use of prophylactic antibiotics before PD catheter insertion (6,10). Intravenous vancomycin, cefazolin, gentamicin, and cefuroxime have been tested (10). The optimal choice of antibiotic, however, is not well defined, and should be determined by the local spectrum of antibiotic resistance. Besides prophylactic antibiotics, other aspects of catheter insertion practice, including the method of catheter placement (mini-laparotomy, laparoscopy, or peritoneoscopy), site of skin incision (midline or lateral), catheter design (e.g., extended, presternal, or upper abdominal catheter), configuration (straight or swan-neck, single or double cuff), and the direction of exit site do not significantly affect the peritonitis rate (11,12). Nonetheless, a large, observational study suggests that the double-cuff catheter is associated with a reduction in peritonitis caused by Staphylococcus aureus (13).

Disconnect PD systems with a “flush before fill” design are consistently associated with a lower peritonitis rate than the traditional spike systems, and are the standard of continuous ambulatory peritoneal dialysis (CAPD) practice nowadays (11,14). There is no significant difference in peritonitis rate between various disconnect systems (Y-set, double-bag, or luer lock) (11,14), or between CAPD and machine-assisted automated PD (15,16). It is uncertain whether the choice of dialysis solution (conventional glucose-based solutions or biocompatible solutions with neutral pH and low glucose-degradation product) leads to any differences in peritonitis occurrence (17).

Training and Nursing Practice.

A good PD training program would logically minimize the peritonitis rate. It is generally accepted that PD training should be conducted by nursing staff with the appropriate qualifications and experience, and the latest ISPD recommendations for teaching PD patients and their caregivers should be followed (18,19). However, published data are limited, and the critical elements of a training program that determine the peritonitis rate remain undefined. The ongoing Targeted Education Approach to Improve Peritoneal Dialysis Outcomes Trial, to be completed in 2023 (20), will help to clarify the benefit of comprehensive PD training programs.

After PD training is completed, a home visit by PD nurse is valuable in detecting unforeseen practical problems with home dialysis (6). However, the benefit of home visit on peritonitis risk has not been formally tested. In addition to the initial training, retraining should be considered after peritonitis or catheter infection episodes; any change in dexterity, vision, or mental acuity; change in supplier or connection system; prolonged hospitalization; or interruption of PD because of other reasons (6). Early studies suggest that a continuous quality improvement (CQI) program in the PD center may help to reduce peritonitis rates (6,21). Nationwide CQI programs have been found to sustainably reduce peritonitis rates (22). A detailed description on the organization of CQI programs is beyond the scope of this review. Nonetheless, a multidisciplinary team that runs CQI programs should meet and review performance metrics regularly (6).

Exit Site and Catheter Infections.

Exit site and catheter tunnel infections are an important risk factor of PD-associated peritonitis (23). Their early detection and prompt antibiotic treatment are logical steps to minimize the risk of subsequent peritonitis (6). The proper care of catheter exit site plays a pivotal role in prevention. Daily topical application of antibiotic cream or ointment to the catheter exit site is recommended (6), and mupirocin cream or ointment should be the agent of choice (24). Daily application of mupirocin cream or ointment to the skin around the exit site reduces the rate of S. aureus exit site infection and probably decreases the rate of peritonitis (24,25). Intranasal mupirocin is effective for reducing S. aureus exit site infection, but not peritonitis (26). Excessive amounts of topical mupirocin directly applied onto the polyurethane or silicone catheter surface can cause catheter erosion (27). Patients must be educated about the proper method of application.

Topical gentamicin is a reasonable alternative to mupirocin for exit site care (28), but the evidence seems less robust. Gentamicin offers an advantage over mupirocin in centers with a high rate of exit site infection by Gram-negative organisms, but the possibility of gentamicin resistance, which affects the choice of antibiotic for peritonitis treatment, is a definite concern. Other alternative strategies, such as topical antibacterial honey (29) or triple ointment (polymyxin, bacitracin, and neomycin) (30), have been tested, but none is shown to be superior than topical mupirocin. In general, regular systemic antibiotic prophylaxis is not advisable. Although intermittent oral rifampicin reduces the rate of S. aureus peritonitis (31), rifampicin resistance, adverse effects, and drug interactions are all serious concerns.

Other Modifiable Risk Factors.

Many other modifiable risk factors for PD peritonitis have been reported (8), but their absolute risk (e.g., cirrhosis, polycystic kidney disease, left ventricular assist device, neutropenia during chemotherapy) are not well defined, and interventions to only very few have been proved to reduce peritonitis risk. Peritonitis often follows invasive endoscopic procedures (e.g., colonoscopy, hysteroscopy) in patients on PD (32). Prophylactic systematic antibiotic before colonoscopy or invasive gynecologic procedures should be considered (6). Although the optimal antibiotic regimen is unknown, intravenous ampicillin with or without aminoglycoside or metronidazole is most commonly used (10). The efficacy of prophylactic antibiotic given intraperitoneally before other invasive procedures is not proved. Prophylactic antibiotics should also be considered after wet contamination or other breaches in technique (5), but there is no widely accepted regimen (6). Although it is a common practice to change the extension tubings after touch contamination, published evidence is limited. Constipation, enteritis, and hypokalemia are associated with an increased risk of peritonitis by enteric organisms (6,8), and these conditions deserve treatment on their own right.

Secondary Prevention.

Most fungal peritonitis episodes are preceded by the use of systemic antibiotics (6,33). Randomized, controlled trials and a systematic review show that the use of either oral nystatin or fluconazole during antibiotic therapy reduces the risk of secondary fungal (especially Candida) peritonitis (6,10). In countries where nystatin is available, it should be the preferred choice because it has no systematic effect or drug interactions. Antifungal prophylaxis may also reduce the risk of fungal peritonitis when a patient on PD receives systemic antibiotics for nonperitonitis infections (10), but this practice does not seem to be widely adopted.

After each episode of peritonitis, a root cause analysis should be performed to determine the etiology and possible interventions to prevent further episodes (6). For example, exchange technique should be reviewed after peritonitis episodes caused by touch contamination, and replacement of PD catheter should be considered after relapsing or repeat peritonitis episodes (6). The key measures for the prevention of PD-associated peritonitis are summarized in Table 1.

Diagnosis.

The diagnosis of PD-associated peritonitis requires any two of the following features: (1) clinical features consistent with peritonitis, i.e., abdominal pain or cloudy dialysis effluent; (2) dialysis effluent white cell count >100/μl (after a dwell time of at least 2 hours), with >50% neutrophils; and (3) positive dialysis effluent culture (6). However, prompt clinical diagnosis and early initiation of antibiotic therapy are key to successful treatment. Therefore, patients presenting with cloudy effluent should be presumed to have peritonitis and treated as such until the diagnosis is confirmed or excluded (6). Whenever peritonitis is suspected, PD effluent should be tested for cell count, differential, Gram stain, and bacterial culture (6). Blood culture bottle kits are the preferred technique for bacterial culture (6). If immediate delivery of the inoculated culture bottles to the laboratory is not possible, they should be incubated at 37°C. Other effluent concentration techniques may further increase the yield, but are cumbersome to use. There is insufficient evidence for other novel laboratory techniques (e.g., reagent strip or molecular-based tests) (6).

Empirical Antibiotic Therapy.

Once the appropriate microbiologic specimens have been obtained, empirical antibiotic therapy should be started (6). No single antibiotic regimen has been proved to be superior than the others, and the choice should be center-specific (34). The basic principle is to provide adequate coverage of both Gram-positive and Gram-negative organisms, including Pseudomonas species. The current recommendations are vancomycin or first-generation cephalosporin for Gram-positive organism coverage, and third-generation cephalosporin or aminoglycoside for Gram-negative organism coverage (6). The choice of vancomycin versus first-generation cephalosporin should depend on the prevalence of methicillin-resistant organisms in each center.

Intraperitoneal administration of antibiotics is the preferred route unless there are features of systemic sepsis (6). When there is a foreseeable delay in administering intraperitoneal antibiotics, however, the systemic route should be used as a temporary measure so as to ensure a prompt treatment (35). Vancomycin, aminoglycosides, and cephalosporin can be mixed in the same dialysis solution bag (36). However, vancomycin and ceftazidime are incompatible if combined in the same syringe for injection (6). The recommended dosages of antibiotics are summarized in the latest ISPD recommendations (6), but many of them are on the basis of clinical experience rather than pharmacokinetic studies. The dosage of many antibiotics needs to be adjusted for patients with substantial residual kidney function (4,6). A fixed generic dosage for all patients may explain the observation that residual kidney function is associated with treatment failure (37).

Intraperitoneal antibiotics can be given as continuous (in each exchange) or intermittent dosing (6). Intermittent dosing is often possible because many antibiotics have substantial systemic absorption during peritonitis, which permit reentry into the peritoneal cavity in subsequent PD cycles. When given intermittently, the antibiotic-containing PD solution should dwell for at least 6 hours to allow adequate absorption. For β-lactams, both continuous and intermittent intraperitoneal dosing are reasonable options, but continuous dosing has a theoretical advantage because the bactericidal activity is time-dependent (i.e., the reduction in bacterial density is proportional to the time above minimal inhibitory concentration), and should be the preferred regimen (6). However, intermittent dosing is often effective and may be the only feasible regimen when the patient requires helpers or health care visitors to administer the antibiotics, or in patients on automated PD who could not be converted to CAPD temporarily (6).

Unlike β-lactams, intraperitoneal vancomycin is more commonly administered intermittently every 4–5 days. The serum vancomycin level should be kept >15 μg/ml to maintain efficacy (38). Intraperitoneal aminoglycoside is also preferably administered as daily intermittent dosing (6). Short-term aminoglycoside therapy does not accelerate the loss of residual kidney function (39), but prolonged or repeated exposure is associated with vestibular toxicity (40) and should be avoided.

Patients on automated PD who develop peritonitis may switch temporarily to CAPD, so as to facilitate intraperitoneal antibiotics therapy, but conversion is not always feasible for pragmatic reasons (6). For patients who remain on automated PD, the intermittent intraperitoneal dosing should be given in the day dwell (6). Unfortunately, there is a substantial knowledge gap regarding the antibiotic dosing for the treatment of peritonitis in automated PD. Because extrapolation of pharmacokinetic data from CAPD to automated PD may result in significant underdosing in patients on automated PD (6), a higher daily dose is often required.

Adjunctive Measures.

Most patients with PD-associated peritonitis could be managed as outpatients. The decision of hospital admission depends on the clinical severity, hemodynamic status, and often practical considerations of treatment. Antifungal prophylaxis, preferably oral nystatin, should be given along with antibiotic therapy (6). Intraperitoneal heparin is usually added when the PD effluent is cloudy, so as to prevent catheter occlusion by fibrin. In addition, careful blood glucose monitoring is advisable in patients with diabetes because glucose absorption from the PD solution may be increased during peritonitis. Peritoneal protein loss is also increased during peritonitis and malnutrition may develop quickly.

Subsequent Management.

Once the PD effluent Gram stain or culture results are known, antibiotic therapy should be adjusted (6). In general, if Gram-positive organisms are identified, antibiotic coverage for Gram-negative bacteria (i.e., aminoglycoside or third-generation cephalosporin) could be stopped, and vice versa once sensitivities are available. PD effluent leukocyte counts and bacterial culture should be performed again 2–3 days after antibiotic therapy, especially when there is no clinical improvement. PD effluent leukocyte count >1090/µl on day 3 may predict treatment failure (41).

The current ISPD recommendations provide a detailed description on the treatment of peritonitis episodes caused by specific organisms (6). In essence, if the clinical response is satisfactory, peritonitis caused by coagulase-negative staphylococci, streptococci, or culture-negative episodes should be treated for 2 weeks (6). For culture-negative episodes, it remains controversial whether the antibiotic for Gram-negative coverage should be discontinued. The current recommendations state that if aminoglycoside is used as the empirical Gram-negative coverage, it should be stopped to minimize the risk of ototoxicity from repeated exposure (6), although a small study has suggested that N-acetylcysteine may prevent aminoglycoside-related ototoxicity (42).

For the treatment of peritonitis episodes caused by S. aureus, enterococci, Corynebacterium species, Gram-negative bacilli (Pseudomonas or non-Pseudomonas species), and polymicrobial peritonitis, effective antibiotics should be continued for 3 weeks. Because enterococci have intrinsic resistance to cephalosporin, and ampicillin is rapidly inactivated when given intraperitoneally (43), enterococcal peritonitis should be treated with intraperitoneal vancomycin unless there is vancomycin resistance (6). Unlike other bacterial causes, Pseudomonas peritonitis should be treated with two effective antibiotics with different mechanisms of action (e.g., gentamicin or oral ciprofloxacin with ceftazidime or cefepime) (6,44,45). If multiple enteric organisms are identified from PD effluent and when there is no prompt clinical response to empirical antibiotics, surgical evaluation should be obtained immediately, and metronidazole should be used with vancomycin and either an aminoglycoside or ceftazidime (6). In contrast, if multiple Gram-positive organisms are identified from the PD effluent, antibiotic treatment alone is usually effective (46). Standard antituberculous chemotherapy is highly effective for peritonitis caused by Mycobacterium tuberculosis. The treatment regimen for nontuberculous mycobacterial peritonitis is not well defined, but catheter removal is usually needed.

Severe Episodes.

The indications of PD catheter removal are summarized in Table 2. Specifically, refractory peritonitis episode is now defined as failure of the effluent to clear after 5 days of appropriate antibiotics (6), whereas relapsing peritonitis refers to the episode that occur within 4 weeks of completion of therapy of a prior episode with the same organism or being culture negative (6). Recurrent peritonitis refers to an episode that occurs within 4 weeks of completion of therapy of a prior episode but with a different organism (6), whereas repeat peritonitis refers to an episode that occurs >4 weeks after completion of therapy of a prior episode with the same organism (6).

After catheter removal for fungal or refractory peritonitis, effective antibiotics should be continued for another 2 weeks (6,47). Insertion of a new PD catheter and return to PD is sometimes possible (47,48), but should be performed at least 2 weeks after catheter removal and complete resolution of peritoneal symptoms (6). PD catheter should also be removed for refractory exit site or tunnel infections (6). If there is no concomitant peritonitis (or after PD effluent has cleared up from the concomitant episode), a new PD catheter could be inserted simultaneously and PD could be continued (7).