Infection remains the number one issue in morbidity and mortality in patients on peritoneal dialysis (PD). It remains the leading cause of technique failure and transfer to hemodialysis. Some studies estimate that peritonitis accounts for 20–25% of modality failures. In addition, exit site and tunnel infections occur in about 1 in every 24 patient months while on PD and many times lead to a disruption in dialysis. A strong home dialysis program highlights infection prevention as a main part of their quality maintenance program. In data reported by the United States Renal Data System (USRDS), collected from information received by Medicare, the incident number of all dialysis patients in 2013 was 117,162, with a prevalence of 661,638. Of this group of incident patients, only 9% initiate dialysis with PD [1]. The reported all-cause hospitalization frequency and total hospitalization days have decreased over last decade. The number of infection related hospital days per patient year in PD patients has also improved, decreasing by 24.8% compared to 2005. However, the rate for infection related hospitalization in peritoneal dialysis patients continues to exceed the rates for cardiovascular related hospitalizations. With recognition of peritonitis as a major factor in morbidity in patients transitioning to PD, greater emphasis has been placed on finding techniques to improve rates of infection.

Historically, the classic presentation for peritonitis among PD patients typically includes cloudy dialysate and increasing abdominal pain. The patient may or may not have accompanying fever or other symptoms of systemic infection such as rigors and hypotension. Peritonitis is frequently thought of as the result of a break in sterile technique or touch contamination. This commonly leads to dialysate contamination by gram positive skin organisms. Studies have demonstrated simple hand washing and hand drying prior to connection minimizes risk of infection [2]. Risk of infection is most when the PD system is open, during an exchange or due to leak. Use of prophylactic antibiotics following such events reduces risk of future peritonitis, although overall incidence of infection is low [3].

Touch contamination is only one of a number of ways that peritoneal fluid can be seeded. Secondary seeding of the dialysate through translocation of bacteria across bowel mucosa can also occur as a complication of elective gastrointestinal invasive procedures and through bowel related issues including gastroenteritis, diarrhea or constipation, or bowel ischemia [4]. Resultant microbiology is reflective of the source. Peritoneal fluid cultures in these cases are typically polymicrobial or grow gram negative organisms. Secondary seeding has also been described through gynecologic procedures. Although much more rare, hematogenous spread is also a concern, through dental procedures or other systemic infections. Another major focus of attention is the relationship between catheter related infections and the progression to peritonitis. Prevention of exit site infections and tunneled infections have been associated with improved rates of peritonitis [5].

The visualization by the patient of a cloudy dialysate bag, accompanied by complaints of abdominal pain, is a standard presentation for peritonitis. The classic physical exam may show fever and abdominal rebound tendernesss. Initial workup typically includes CBC with differential, PD effluent cell count and culture. Measurement of dialysate cell count typically will demonstrate WBCs of greater than 100 cells/mm³, 50% of which are neutrophils. However, cloudy effluent can be the result of non-infectious etiologies as well. The differential for cloudy effluent may include chemical peritonitis, allergic peritonitis, and hemoperitoneum, but both clinical presentation and dialysate cell count differential help differentiate the presence of an infectious etiology. When the percentage of eosinophils predominates, this may suggest more non-infectious etiologies or more allergic type issues. When lymphocytes are the predominant cell on cell count, fungal or mycobacterial infection need to be considered. However the cell count differential itself does not accurately predict the type of infection present.

In recent decades, the ability to rely on visual recognition of a cloudy bag has decreased due to the shift in preference by practitioners and patients from chronic ambulatory dialysis (CAPD) to automated peritoneal dialysis (APD). Patients may present with early subjective symptoms of abdominal discomfort without any history of a change in appearance of dialysis fluid. Repeated examination of dialysis fluid and collection of dialysate for cell count may become more critical in the diagnosis of peritonitis and may become abnormal later in the course of infection. Collection of cell count from the patient on APD, who does not have a last fill, may result in a WBC count below 100 cells/mm³. In these patients, a dry abdomen during the day or much shorter contact time between dialysate and peritoneum, may result in low WBC cell count measurements. Recognition of these issues and allowing test dialysate fluid to dwell an appropriate length of time in the abdomen before collecting a cell count sample, will avoid false negative results. In APD patients, the appropriate clinical presentation, accompanied by a dialysate cell count that is mostly neutrophils, would be sufficient to support empiric antibiotic therapy for presumed peritonitis in these patients.

Awareness of clinical risk factors that increase a peritoneal dialysis patient’s likelihood of developing acute peritonitis contributes to development of interventions to lower peritonitis rates. Intervention to avoid severe constipation and education of patients regarding early notification of their home dialysis unit in the event of a break in technique, are important preventative measures. Although the incidence of secondary peritonitis is rare from invasive gastrointestinal, dental and gynecologic invasive procedures, the benefit of prophylactic antibiotics warrant their use pre-procedure [4, 6]. Typically, based upon a 1994 review by Strippoli, most centers administer intravenous aminoglycoside plus ampicillin prior to the colonoscopy. In dental procedures, typically, oral amoxacillin prophylxis is sufficient. The importance of educating patients regarding notification of their home dialysis unit prior to undergoing such elective procedures should not be overlooked.

Multiple studies have looked at a variety of catheter related characteristics and surgical techniques to assess if there may be a benefit to one type of catheter or one surgical implantation technique. Development of the double cuffed catheter, exit sit location, direction of the tip of the catheter and design of the catheter portion within the abdomen all were conceived to improve infection risk. To date, data has not strongly demonstrated a particular catheter to be superior in terms of infection prevention [7]. Laproscopic techniques with or without burying of the catheter prior to externalization and use have also not been demonstrated to lead to improved infectious outcomes, thus the International Society of Peritoneal Dialysis (ISPD) Guidelines do not recommend a specific type of peritoneal catheter as better suited to prevention of peritonitis [8]. However, data does support pre-operative administration of intravenous antibiotics prior to catheter placement. A randomized trial by Gadallah et al., showed that, in a 14 day perioperative period, single dose vancomycin pre-operatively was superior to cefazolin. A smaller benefit was also seen with the use of the first generation cephalosporin, as compared to no prophylactic antibiotic use [9]. Vancomycin use remains controversial, with ongoing concerns for development of antibiotic resistant organisms. ISPD guidelines suggest the careful consideration of cost/benefit by each center of the use of prophylactic vancomycin [8].

Identification of higher risk patients may allow for more guided surveillance and early prevention techniques in these patient groups. Kumar et al. found that neither race nor socioeconomic status predicted technique survival [10] however Nessim et al. examined data in over 4000 patients from the Baxter POET database [11]. The predictors of peritonitis in these incident patients from Canada included black race, female patients with diabetes and those who transferred from hemodialysis to peritoneal dialysis. There was no modality effect, with similar infection risk identified between CAPD and APD patients. The data is contradictory on whether CAPD increases risk of infection. There has been increased concern with the recognition that there are many more potential points/times of contact and possible contamination that result from increased frequency of connection/disconnection. Past suggestion of increased infection risk with CAPD in studies appears to be related to early connection/disconnection technology. More current studies have not found a consistent difference in infection rates between the two PD techniques [12, 13].

Although difficult to measure and standardize, the foundation of a center’s peritonitis prevention program remains the experience of the program and the effectiveness of the peritoneal dialysis nurse trainer. The years of experience of the training professional inversely correlates with the risk of peritonitis in patients trained, demonstrating the need for continuing support of the trainer nurse [14]. Ongoing reinforcement of technique and re-education of patients also appears to be important in preserving low rates of infection. The length of time spent in training is not well defined and there are no randomized trials evaluating this question. A recent Brazilian prospective cohort study of over 2000 incident PD patients found a association between greater than 15 h of training and decreased peritonitis rates. As well, centers with larger number of patients and greater length of experience had decreased peritonitis rates [15]. The ISPD Guidelines recommend that any PD training center track peritonitis episodes and maintaining ongoing continuing education for patient and staff. They recommend retraining when an acute event as hospitalization or infection occurs [16].

Exit site infections are an independent risk factor for peritonitis [17]. The presence of erythema, swelling and pain on exam with purulent drainage are findings suggestive of an exit site infection. In clinical settings where there is the suggestion of active exit site infection, empiric antibiotic treatment is recommended by the ISPD guidelines [18]. Oral administration of antibiotics is generally sufficient; any empiric antibiotic therapy should provide coverage for Staphylococcal aureus organisms [19]. The presence of Staphylococcal or Pseudomonal infection is suggestive of more extensive catheter involvement and has high risk to extend into a tunnel infection. Physical exam demonstrating swelling, erythema and tenderness along the track of the catheter is supportive tunnel involvement by infection. Use of ultrasound may be helpful to define fluid collections along the track [20, 21]. Peritoneal catheter removal is recommended when there is ongoing exit site or tunnel infection which is non-responsive to therapy, or there is an exit site or tunnel infection leading to peritonitis. Some data suggests if there infection is limited to the catheter, a single surgical procedure to remove and re-implant the peritoneal catheter may be successful [22]. This may allow the patient to avoid a switch to temporary hemodialysis. Infections with pseudomonas are particularly problematic and often lead to catheter removal. Some novel techniques have been described to salvage the existing catheter in patient who have an exit site infection refractory to antibiotics. “Cuff shaving” by unroofing the subcutaneous cuff, shaving it off, and rerouting the catheter to an alternate exit site has been reported successful in 87.5% of children in one study from Japan. This technique was also successful in 13 adults with chronic tunnel infection [23]. Wu et al. described 26 catheters in 23 patients in which the entire subcutaneous tubing was replaced from just above the internal cuff with no interruption in dialysis [24].