Peritonitis and Exit-Site Infection
Cheuk-Chun Szeto
Philip K.-T. Li
David J. Leehey
I. PERITONITIS
A. Incidence. Peritonitis remains the Achilles’ heel of peritoneal dialysis (PD). Peritonitis is a “contributing factor” to 16% of deaths on PD. Furthermore, it is the most common cause of treatment failure, accounting for nearly 30% of the cases. The overall incidence of peritonitis in continuous ambulatory peritoneal dialysis (CAPD) patients during the 1980s and early 1990s averaged 1.1-1.3 episodes per patient-year in the United States. With improved patient training, PD delivery systems, and prophylactic measures, the rate of peritonitis has fallen worldwide. Many centers now report a peritonitis rate of 0.2 to 0.6 episodes per patient-year at risk, or 1 episode per 20-60 patient-months of PD (Piraino, 2011). The introduction of Y-set and double-bag disconnect systems has substantially reduced the incidence of peritonitis, particularly episodes caused by gram-positive organisms (Monteon, 1998; Li, 2002). The same flush-before-fill methodology used in CAPD Y sets can also be used effectively in automated peritoneal dialysis (APD). Peritonitis rates with APD and CAPD are not generally different. Patients on APD going “dry” during the day (i.e., no daytime dwell) may have a decreased risk of infection compared with those with a day dwell. The current International Society for Peritoneal Dialysis recommendations on PD-related infections (Piraino, 2011) state that every program should monitor infection rates, ideally once a month, but, at a minimum, on a yearly basis.
B. Pathogenesis
1. Pathways of infection
a. Intraluminal. Peritonitis occurs most often because of errors in technique in making or breaking a transfer set-to-bag or catheter-to-transfer set connection. This allows bacteria to gain access to the peritoneal cavity via the catheter lumen. Typically, the organisms involved are coagulase-negative staphylococci or diphtheroids.
b. Periluminal. Bacteria present on the skin surface can enter the peritoneal cavity via the peritoneal catheter tract.
Typically, the organisms involved are Staphylococcus aureus or Pseudomonas aeruginosa.
Typically, the organisms involved are Staphylococcus aureus or Pseudomonas aeruginosa.
c. Bowel source. Bacteria of intestinal origin can enter the peritoneal cavity by migrating across the bowel wall. This is the usual mechanism of peritonitis episodes associated with diarrheal states and/or instrumentation of the colon, and may also be seen with strangulated hernia. Typical organisms involved are Escherichia coli and Klebsiella sp.
d. Hematogenous. Less commonly, peritonitis is due to bacteria that have seeded the peritoneum from a distant site by way of the bloodstream. Typical organisms here are streptococci and staphylococci.
e. Transvaginal. This is uncommon, but ascending infection may occur from the vagina via the uterine tubes into the peritoneum. Some Candida peritonitis may occur by this route.
2. Role of host defenses. The peritoneal leukocytes are critical in combating bacteria that have entered the peritoneal space by any of the routes already mentioned. A number of factors are now known to alter their efficacy in phagocytizing and killing invading bacteria.
a. Dialysis solution pH and osmolality. Standard PD solutions have a pH around 5 and an osmolality ranging from 1.3 to 1.8 times that of normal plasma, depending on the glucose concentration used. These unphysiologic conditions may inhibit the ability of peritoneal leukocytes to phagocytize and kill bacteria. High osmolality, low pH, and the presence of the lactate anion combine to cause inhibition of superoxide. There is now some evidence that newer normal pH, “biocompatible” solutions may reduce the peritonitis rate, but this has not been a consistent finding among published studies (Cho, 2014).
b. Peritoneal dialysis solution calcium levels. The antimicrobial actions of peritoneal macrophages are enhanced by both calcium and cholecalciferol. Use of active vitamin D has been reported to reduce the rate of peritonitis (Kerschbaum, 2013). Use of a 1.25-mM (2.5-mEq/L) calcium concentration in PD solution has gained popularity as it may improve adynamic bone disease and reduce vascular calcification. An increased risk of Staphylococcus epidermidis peritonitis has been reported with the use of low-calcium dialysis solutions (Piraino, 1992), but no subsequent confirmatory reports have been published.
C. Etiology. Using appropriate culture techniques, an organism can be isolated from the peritoneal fluid in over 90% of cases in which symptoms and signs of peritonitis and an elevated peritoneal fluid neutrophil count are present. The responsible pathogen is usually a bacterium, but fungal peritonitis occurs occasionally (Table 27.1).
TABLE 27.1 Frequency of Organisms Isolated in Patients with Peritonitis | |||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| |||||||||||||||||||||||||||||||||
D. Diagnosis. At least two of the following three findings should be present: (a) symptoms and signs of peritoneal inflammation, (b) cloudy peritoneal fluid with an elevated peritoneal fluid cell count (>100/mcL) due predominantly (>50%) to neutrophils, and (c) demonstration of bacteria in the peritoneal effluent by Gram stain or culture.
1. Symptoms and signs. The most common symptom is abdominal pain, but this is sometimes very mild. Others are nausea, vomiting, and diarrhea (Table 27.2). Sometimes, especially in the elderly, the only symptoms are a relatively sudden loss of residual renal function and postural hypotension. On the other hand, abdominal pain can be present in dialysis patients owing to nonperitonitis-related abdominal causes; in those starting dialysis after a failed transplant in whom steroid treatment has been stopped, abdominal pain due to adrenal insufficiency should be considered.
2. Peritoneal fluid
a. Cloudiness of the fluid. The peritoneal fluid generally becomes cloudy when the cell count exceeds 50-100/mcL (50-100 × 106/L). In most patients, sudden onset of cloudy fluid with appropriate abdominal symptoms is sufficient evidence of peritonitis to warrant initiation of antimicrobial therapy. However, cloudy peritoneal fluid may be due to other factors (e.g., fibrin, blood, or, rarely, malignancy or chyle) rather than to an increase in the white blood cell (WBC) count. Occasionally, fluid drained after a prolonged dwell period (such as after the daytime dwell in APD patients) appears cloudy in the absence of peritonitis. Conversely, a relatively translucent peritoneal fluid does not completely exclude peritonitis. Cloudy fluid has been reported with the use of calcium channel blockers, presumably because they
increase triglyceride concentration in the peritoneal fluid (Ram, 2012).
increase triglyceride concentration in the peritoneal fluid (Ram, 2012).
TABLE 27.2 Symptoms and Signs of Peritonitis | ||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||||||||||
b. Importance of performing a differential count of peritoneal fluid cells. Peritonitis is usually associated with an increase in the absolute number and percentage of neutrophils in the peritoneal fluid. On some occasions, a high peritoneal fluid cell count causing cloudy fluid will be present owing to an increase in the number of peritoneal fluid monocytes or eosinophils (see what follows). Most such cases are not associated with peritonitis and do not require antimicrobial treatment. For this reason, one should perform a differential cell count on the peritoneal fluid sample. Prior to counting, the fluid is spun in a special centrifuge (e.g., Cytospin, Shandon, Inc., Pittsburgh, PA) and the sediment colored with Wright stain.
c. Obtaining the specimen
1. CAPD patients. After disconnecting the drain bag full of peritoneal effluent, the bag is inverted several times to mix its contents. A sample (7 mL) is aspirated from the port of the drain bag and transferred to a tube containing ethylenediamine tetraacetic acid (EDTA).
2. APD patients. A representative cell count can be obtained easily from the daytime dwell by first draining the abdomen and taking the sample from the drainage bag. In those who are “day dry,” there may be some residual fluid present in the abdomen at the time the patient is seen. In these cases, the peritoneal fluid sample can be obtained directly via the peritoneal catheter. After careful cleaning of the catheter with
povidone-iodine, a syringe is attached using meticulous sterile technique, and 2-3 mL of fluid in the catheter lumen is withdrawn and discarded. The peritoneal fluid sample (7 mL) is then withdrawn from the catheter using a second syringe. The sample is injected into a tube containing EDTA. If insufficient fluid is obtained in this manner, one can infuse 1 L or more of dialysis solution and drain the abdomen, obtaining a sample from the effluent. Although the absolute peritoneal fluid cell count will be lower in this diluted specimen, the differential count will be similar to that in a sample obtained directly via the catheter.
povidone-iodine, a syringe is attached using meticulous sterile technique, and 2-3 mL of fluid in the catheter lumen is withdrawn and discarded. The peritoneal fluid sample (7 mL) is then withdrawn from the catheter using a second syringe. The sample is injected into a tube containing EDTA. If insufficient fluid is obtained in this manner, one can infuse 1 L or more of dialysis solution and drain the abdomen, obtaining a sample from the effluent. Although the absolute peritoneal fluid cell count will be lower in this diluted specimen, the differential count will be similar to that in a sample obtained directly via the catheter.
3. Storage time. Morphologic identification of the various cell types can become quite difficult in effluent samples stored for more than 3-5 hours prior to injection into the EDTA-containing sample tube.
d. Peritoneal fluid cell counts in peritonitis. The absolute peritoneal fluid cell count in CAPD patients is usually <50 and often <10 cells/mcL. In “day dry” APD patients, the normal cell count may be much higher, especially in specimens taken directly via the catheter when the peritoneal fluid volume is small. Normally, the peritoneal white cells are mainly mononuclear (monocytes, macrophages, and occasional lymphocytes), and the percentage of neutrophils does not exceed 15%. A value >50% suggests peritonitis, while a value >35% should raise suspicion. The percentage of neutrophils is raised in fungal and even tuberculous peritonitis as well as in the more common bacterial peritonitis.
The percentage of neutrophils in the peritoneal fluid is occasionally elevated in the absence of peritonitis— in patients with infectious diarrhea or active colitis (or appendicitis or diverticulitis), in those with pelvic inflammatory disease, and in women who are menstruating or ovulating or who have recently had a pelvic examination.
e. Peritoneal fluid monocytosis. If there is persistent peritoneal fluid monocytosis or lymphocytosis, tuberculous peritonitis should be considered. Peritoneal fluid monocytosis may also occur in conjunction with peritoneal fluid eosinophilia.
f. Peritoneal fluid eosinophilia. The peritoneal fluid eosinophil count may become elevated in PD patients, causing cloudy fluid and leading to a suspicion of peritonitis (Humayun, 1981). Usually, the peritoneal fluid monocyte count is also elevated. Peritoneal fluid eosinophilia occurs most often soon after peritoneal catheter insertion. It may be seen in the sterile peritonitis that can occasionally occur in those patients who have initiated treatment with icodextrin PD solution. The irritant effect of peritoneal air (e.g., introduced at time of laparotomy) and possibly of plasticizers leached into the peritoneum
from PD solution containers and tubings is another suspected cause. In such cases, the eosinophilia most often resolves spontaneously within 2-6 weeks. Peritoneal fluid eosinophilia can also occur uncommonly during the treatment phase of peritonitis. There have been several case reports of its occurrence in association with fungal and parasitic infections of the peritoneum.
from PD solution containers and tubings is another suspected cause. In such cases, the eosinophilia most often resolves spontaneously within 2-6 weeks. Peritoneal fluid eosinophilia can also occur uncommonly during the treatment phase of peritonitis. There have been several case reports of its occurrence in association with fungal and parasitic infections of the peritoneum.
g. Culture of peritoneal fluid. The incidence of positive peritoneal fluid cultures in patients suspected of having peritonitis depends on culture technique. Culture-negative peritonitis should not be >20% of episodes.
1. Storage. Peritoneal fluid should be cultured promptly; however, infected fluid kept at room temperature or refrigerated for a period often grows pathogenic organisms on subsequent culture. If immediate delivery to the laboratory is not possible, the inoculated culture bottles should ideally be incubated at 37°C.
2. Sample volume. The volume of peritoneal fluid sent for culture should be at least 50 mL as larger volumes increase the likelihood of a positive culture.
3. Sample preparation. The aliquot is centrifuged (e.g., at 3,000 g for 15 minutes) to concentrate the organisms. The supernatant is decanted off, and the pellet resuspended in 3-5 mL of sterile saline and inoculated into standard blood culture media (aerobic and anaerobic). Rapid culture techniques (e.g., Septi-chek, BACTEC) may be utilized.
4. Yield of positive cultures. Seventy to ninety percent of dialysate samples taken from patients with clinical peritonitis yield positive cultures for a specific organism within 24-48 hours. More time may be needed for more fastidious organisms.
5. Improving culture yield. This may be done by hypotonic lysis. The centrifuged sediment is resuspended in 100 mL of sterile water to induce lysis of its cellular elements. This may lead to release of bacteria from neutrophils and increase the chance of a positive culture, even in patients who have already received antibiotics.
6. Incidence of false-positive results. With very sensitive culture techniques, about 7% of cultures may be positive in patients without clinical peritonitis. The significance of this is unclear.
h. Gram stain. Gram stain of the peritoneal fluid sediment is useful but positive in less than half of cases of cultureproven peritonitis. Gram stain is also useful for making the diagnosis of fungal peritonitis. Staining with fluorescent acridine orange dye has been reported to increase the visibility of bacterial organisms.
i. Necessity of performing blood cultures. Routine blood cultures are not necessary unless a patient appears septic or an acute surgical abdominal condition is suspected.
E. Treatment
1. Initial management
a. Choice of antimicrobial therapy. Empiric antibiotics must cover both gram-positive and gram-negative organisms. Vancomycin or a first-generation cephalosporin such as cefazolin or cephalothin is used in combination with an antibiotic such as ceftazidime or an aminoglycoside. In general, a center-specific selection of empiric therapy, dependent on the local history of sensitivities of organisms causing peritonitis, is recommended.
1. Gram-positive. First-generation cephalosporins (e.g., cefazolin) are often preferred to vancomycin because of the emergence of vancomycin-resistant organisms. Intraperitoneal (IP) cefazolin can be conveniently administered in a single daily dose of 15 mg/kg, although a 25% increase in dose is recommended in patients with substantial residual renal function (Manley, 1999). Alternatives to vancomycin include nafcillin and clindamycin. Vancomycin can be used as first-line treatment or reserved for patients harboring β-lactam-resistant organisms, especially methicillin-resistant S. aureus (MRSA), or with penicillin/cephalosporin allergy. Ciprofloxacin alone is not recommended for gram-positive infections.
2. Gram-negative or indeterminate. Gram stain is usually not diagnostic, and so gram-negative organisms need to be covered by a third-generation cephalosporin or aminoglycoside. In theory, aminoglycosides should be avoided if possible in patients with residual renal function because of their nephrotoxicity (Shemin, 1999), although short courses of aminoglycosides probably do not harm residual renal function (Lui, 2005). Aminoglycosides may be used in patients without residual renal function, although one still must be wary of otovestibular toxicity. Table 27.3 lists sample prescriptions based on the use of cefazolin in combination with ceftazidime.
b. Delivery methods and schedules for antimicrobial drugs
1. IP versus oral (PO) or intravenous (IV) antimicrobial therapy. IP administration of antibiotics is preferred to IV or PO dosing for treating peritonitis. IV antibiotics, however, should be used when there is clinical evidence of systemic sepsis.
2. The loading dose. A loading dose of antimicrobials is usually given IP when CAPD is the treatment modality (Table 27.4). If a patient appears toxic, an IV loading dose should be used. For aminoglycosides, the IV loading dose is usually gentamicin or tobramycin 1.5 mg/kg or amikacin 5 mg/kg. If a patient is in substantial pain and cannot tolerate the usual exchange
volume, the IP loading dose can be administered in a smaller volume of dialysis solution (e.g., 1 L). For APD patients, the loading dose can be given IV, but it can also be instilled via a peritoneal dwell that is then left in place for at least 4-6 hours.
volume, the IP loading dose can be administered in a smaller volume of dialysis solution (e.g., 1 L). For APD patients, the loading dose can be given IV, but it can also be instilled via a peritoneal dwell that is then left in place for at least 4-6 hours.
TABLE 27.3 Sample Prescriptions for Initial Treatment of Peritonitis with Unknown Organism Type | |||||
|---|---|---|---|---|---|
|
3. Maintenance antimicrobial dose. After the loading dose has been given, a CAPD or APD schedule is continued, with maintenance doses of antimicrobials added to each exchange (Table 27.4). Some centers switch APD patients to CAPD, but this is not routine. Maintenance antibiotics in CAPD patients can be administered as an intermittent dose once daily. For patients on an APD schedule, antibiotics can be administered conveniently in the daytime dwell. For patients on a day dry APD schedule, temporary conversion to a CAPD regimen may be considered because of ease of antibiotic administration, or, alternatively, a low-volume daytime dwell (e.g., 1 L) could be temporarily added. Because of increased cycler clearance of antibiotics, doses need to be higher in patients who remain on APD during the treatment of peritonitis (Manley and Bailie, 2002). (Examples are given in Table 27.5.)
4. Antimicrobial dosing guidelines. Suggested loading and maintenance doses for a number of antimicrobial drugs are listed in Table 27.4. For maintenance doses added to the dialysis solution, continuous and intermittent dosing of antibiotics are equally efficacious. For continuous dosing, the same dose of antibiotic is added to each dialysis solution bag. Alternatively,
a larger dose is added to one bag only, every 12 or 24 hours (or, in the case of vancomycin, every 4-5 days). A randomized trial in children showed that intermittent vancomycin was as effective as continuous vancomycin (Schaefer, 1999). Single daily dosing of aminoglycosides has several advantages, including ease of administration, increased efficacy, and potentially less toxicity. Increased bacterial killing rates associated with prolonged postantibiotic effect are obtained using once-daily dosing. However, trough concentrations of antibiotic (i.e., 24 hours after a dose) will be low, and the exact duration of the postantibiotic effect is unknown, which has led to some concern about the advisability of this type of regimen, especially in patients with residual renal function (Low, 1996).
a larger dose is added to one bag only, every 12 or 24 hours (or, in the case of vancomycin, every 4-5 days). A randomized trial in children showed that intermittent vancomycin was as effective as continuous vancomycin (Schaefer, 1999). Single daily dosing of aminoglycosides has several advantages, including ease of administration, increased efficacy, and potentially less toxicity. Increased bacterial killing rates associated with prolonged postantibiotic effect are obtained using once-daily dosing. However, trough concentrations of antibiotic (i.e., 24 hours after a dose) will be low, and the exact duration of the postantibiotic effect is unknown, which has led to some concern about the advisability of this type of regimen, especially in patients with residual renal function (Low, 1996).
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
