Preventing peritonitis in PD patients

Peritonitis remains a serious complication of peritoneal dialysis. Peritonitis is still a frequent cause of technique failure, and can be associated with death of the patient. Therefore, it is important to establish best demonstrated practices to reduce peritonitis rates to low levels.

Much remains to be learned about preventing peritonitis. The International Society for Peritoneal Dialysis (ISPD) 2005 guidelines on peritoneal dialysis (PD)-related infections included sections on prevention.1 However, because the data are rather limited, the ISPD Standards and Guidelines Committee subsequently felt this would be better presented as a position paper.2 This will soon be published in Peritoneal Dialysis International and will also be freely available to all on the ISPD website (ispd.org). In addition, Guidelines on Prevention and Treatment of Peritonitis in Children are under review and are likely to be available toward the end of 2011 or early 2012 both on the ISPD website and as a publication in Peritoneal Dialysis International. The present paper represents my personal views on preventing peritonitis, but I refer the reader to these important resources.

One of the most important aspects of prevention of peritonitis in PD patients is following the peritonitis rates in a program. This is surprisingly infrequently done; a survey by ISPD nurses of PD programs around the world found that more than 50% did not know the peritonitis rates of their own program.3 A center is not able to evaluate the problem if peritonitis rates are not followed closely, at a minimum yearly, but preferably quarterly. One approach is to have the home training nurse keep track of the peritonitis in a longitudinal fashion by calculating this monthly, and have meetings of at least the physicians caring for the peritoneal dialysis with the nurses to examine the infectious complications in the PD programs and strategize approaches to prevent further episodes approximately quarterly.

The nephrology world should work on standardizing the method of expressing infection rates. While the traditional approach has been to express the rate as months between episodes, a preferable method is to calculate as peritonitis episodes per year at risk. This is done by adding together the sum of the days on peritoneal dialysis of all the patients in a program and then converting this to dialysis years. This then is the denominator of the formula, and the peritonitis episodes are the numerator. For example, if during a single month a center has 25 patients on the entire month (30 days each) and one patient who started during the month (starting the time at risk from the first day of training) and was on PD for 10 days during the month, the time at risk for that month is 25 × 30 + 10 days, or 760 days. Conversion to dialysis years is performed by dividing 760 days by 365 days/year and results in a time at risk of 2.08 years for that 1 month for that particular program. Table I outlines time to be included and excluded in the calculation. If there was one episode of peritonitis during this month, the center's peritonitis rate is 1/2.08 or 0.48 episodes per year at risk for that particular month. This can then be compared with the previous rates of the center, and trends can be considered.

In addition to calculating the total peritonitis rate, a program needs to examine the organism-specific peritonitis episodes and calculate a rate for each organism. For example, if a program has in 1 year a total of 24 years at risk, with four episodes of peritonitis during the year, one of which was coagulase-negative Staphylococcus, the total rate of peritonitis is 0.16 episodes per year at risk (four episodes divided by 24 years), and the rate of coagulase-negative Staphylococcus peritonitis is 0.04 episodes per year at risk (one episode divided by 24 years at risk). This is an acceptable rate for coagulase-negative Staphyloccus and suggests that the training process is appropriate. This can also be benchmarked against published literature. In all adult PD patients in Australia between 2003 and 2006, the total peritonitis rate was 0.6 episodes per year at risk, and that due to coagulase-negative Staphylococcus was 0.16 episodes per year at risk.4 In both programs, about 25% of the episodes are due to coagulase-negative Staphylococcus, but the small program has a strikingly lower rate both overall and for coagulase-negative Staphylococcus. This illustrates how misleading it is to use percentages to examine peritonitis by organisms.

Table II gives an example of the calculated organism-specific rates for a single center in Pittsburgh with data collected as part of the University of Pittsburgh Peritoneal Registry. This program has very low rates of coagulase-negative Staphylococcus, Staphylococcus aureus, and Pseudomonas aeruginosa peritonitis. The low rates of peritonitis due to coagulase-negative Staphylococcus can be ascribed to the training methods, while the low rates of S. aureus and P. aeruginosa can be ascribed to the use of gentamicin cream for routine exit-site care. A close examination of the organism-specific rates indicates that our program needs to focus on reducing the rate of enteric peritonitis, as these predominate. In this way the PD team for each program can examine the data and determine problematic areas that require attention.

As part of the continuous quality improvement (CQI) approach toward peritonitis, for every episode of peritonitis an effort should be made to determine causality. Type of organism is an important clue to causality, as is information gathered from the episode. Sometimes the patient is aware of a contamination episode not properly addressed at the time that led to peritonitis. The organisms in these cases are common environmental and skin contaminants. Alternatively, the patient may have an exit-site infection with an organism of the same type. Generally these are S. aureus or P. aeruginosa but can also be other Gram-negative bacilli or diphtheroids. Peritonitis due to enteric organisms in a patient with no known contamination and with a history of either constipation or diarrhea suggests a gastrointestinal source. Some causes of peritonitis are listed in Table III.

Protocols for preventing peritonitis are critical for the success of any PD program. A list is provided in Table IV. A protocol for placing peritoneal catheters is the starting point.5 The approach should include identification prior to surgery of the ideal placement of the exit site and preparation for surgery with cleansing of the skin. Some would de-colonize those with nasal S. aureus carriage using intranasal mupirocin twice daily for 5 days, but this approach has been poorly studied. The expertise of the operator placing the catheter is important for achieving a non-traumatic tunnel and a round, tight exit site for rapid healing. Most programs recommend keeping the dressing in place with dressing changes by the PD nurse until healing is well under way, at which point chronic exit-site care is done. Routine exit-site care by the patient can begin at that point.

Exit-site colonization, especially with S. aureus or P. aeruginosa, can lead to exit-site and tunnel infection and subsequently peritonitis caused by the same organism. These are generally severe episodes and not infrequently the catheter must be removed. Most exit-site infections can be prevented by using proper exit-site care, including use of exit-site antibiotic cream as part of routine care. Either mupirocin or gentamicin can be used.6 Judgment needs to be used in the decision to remove PD catheters for exit-site or tunnel infections. For example, as shown in Table II, one episode of Streptococcus tunnel infection occurred in our program in 2010; this was in an immunocompromised patient, developed rapidly and was severe. The decision was made to rapidly remove the catheter within a few days of presentation and therefore prevent peritonitis from this organism.

Training of the patient needs to be structured and cover all the important material including recognition of contamination, knowledge of how to do the connection without contamination, and recognition of peritonitis. The patient should be tested for achievement of knowledge at the end of training. Length of training should be individualized to each patient, and is probably best done by a one-on-one approach by a single nurse who has been trained to teach patients how to do PD. A video, “Training the Trainer,” is freely available at www.ispd.org and was sponsored by the ISPD.

Retraining is an important area not well studied but is likely important to correct the adoption of poor technique. After the initial training is completed, the patient should be seen within 1–2 weeks and evaluated for any problems with the procedure; ultrafiltration and volume status should be assessed. At this point adjustments in the prescription can be made, but in addition, this is an appropriate time for the nurse and physician to reinforce the correct procedure and identify potential problems. Retraining should be ongoing and can be done as part of the monthly visit, although home visits to assess the environment are also useful.

Prevention of enteric peritonitis is poorly studied, but some data suggest that hypokalemia and constipation might lead to transmural migration of bacteria across the bowel wall with resultant peritonitis due to enteric organisms.7, 8 While it has not been proved that hypokalemia correction reduces the risk of peritonitis, such correction is easily accomplished with dietary intervention or a small dose of potassium supplement. Bowel hygiene with attention to prevention of constipation is also a good approach. Procedures such as colonoscopy can lead to peritonitis, and therefore the abdomen should be empty of dialysis fluid during these procedures.9 I recommend prophylactic antibiotics prior to the procedure to prevent this complication. Such an approach decreases the risk of peritonitis from about 5% to close to 0%.9

Clearing of the effluent in treatment of a peritonitis episode does not mean that the organism will not cause another episode. In the Australian study of coagulase-negative Staphylococcus, 17% (n = 158 episodes) relapsed (recurred within 4 weeks) and another 194 had a repeat episode of coagulase-negative Staphylococcus, often in the second month after treatment.4 Further study needs to be done to rigorously evaluate approaches to prevent such additional episodes, perhaps by using alternative antibiotics, longer courses of therapy, higher doses, or adding rifampin. In our program after one second episode closely linked to a first episode, the PD catheter is generally replaced.

Preventing peritonitis requires adequate training of the nurses and physicians in the program. An interesting example of an outreach program is from southern China.10 This center of excellence undertook to establish satellite programs for PD in a province of China, making this cheaper and desirable form of dialysis more widely available. The approach is outlined in Table V. This model resulted in very rapid growth of PD, 1-year mortality of 83%, 1-year technique survival of 93%, a peritonitis rate at satellite programs of 0.26 episodes per year at risk, and that of coagulase-negative Staphylococcus of 0.03 episodes per year at risk. This model, which uses a center of excellence to train and monitor the satellite units, could be adapted in many other areas of the world as one to emulate.

To summarize, with close attention to peritonitis episodes and calculation of organism-specific rates, a program can determine possible causality and develop approaches to prevent further episodes. This should be an iterative process. Rates of 0.17 episodes per year at risk are reported in Japanese children11 and rates of 0.22 episodes per year at risk in a nationwide survey in Japanese adults.12 The low rates are ascribed to good sanitation, the connectology used, and a good education process.12 These results are in contrast to the rates of peritonitis reported for children in Australia (0.72 episodes per year at risk with a rate of 0.17 episodes per year at risk for coagulase-negative Staphylococcus).13 These widely different peritonitis rates probably mostly represent variances in protocols and training failures. The Australian and New Zealand nephrologists, who are leading the way in examining PD-related infections on a national basis, are calling for action to improve outcomes in PD patients, which includes lowering infectious complications.14 Low rates of peritonitis are achievable.