Campylobacter Peritonitis in Continuous Ambulatory Peritoneal Dialysis: Report of Eight Cases and a Review of the Literature Colin J. Wood, FRACP, Vennessa Fleming, BMed Sci, John Turnidge, FRACP, FRCPA, Napier Thomson, PhD, FRACP, and Robert C. Atkins MSc, FRACP • Continuous ambulatory peritoneal dialysis (CAPD)-associated peritonitis remains the most common complication of this form of renal replacement therapy and the potential causative organisms are many and varied. Campylobacter bacteria are unusual causes of peritonitis in CAPO patients and we report our experience with eight cases and review those cases reported in the literature. In many episodes, there is a strong association with acute enterocolitis, which may precede the onset of cloudy dialysate by many days. The method of spread of these organisms from the gastrointestinal tract to the peritoneal cavity remains speculative. Bacteremic transfer would appear the most likely route, although it is probable that no single mode of spread explains all episodes. The treatment of choice of Campylobacter peritonitis is with intraperitoneal aminoglycoside in combination with oral erythromycin. An awareness of the potential for Campylobacter to cause CAPO-associated peritonitis, particular1y when diarrhea is a prominent feature, has resulted in this organism becoming increasingly isolated in our unit. © 1992 by the National Kidney Foundation, Inc. INDEX WORDS: Continuous ambulatory peritoneal dialysis; peritonitis; Campylobacter.
C
ONTINUOUS AMBULATORY peritoneal dialysis (CAPD) is now widely accepted as effective treatment for end-stage renal failure, but peritonitis remains the most common complication. I Staphylococci, particularly coagulasenegative species, are the most frequently isolated organisms, accounting for approximately 50% of all reported peritonitis episodes. Gram-negative bacilli, although occurring less commonly, are isolated in approximately 20% of peritonitis episodes, but this proportion can vary widely.2 Campylobacters are micro-aerophilic curved or spiral gram-negative bacteria recognized in the 1970s as a common cause of acute enterocolitis in man. They occur in the intestinal flora of many wild and domestic animals and poultry and can be transmitted to humans by milk, water, and undercooked meat polluted by such animal carriers. Although a common cause of enterocolitis, they are rare causes of CAPD-associated peritonitis, with only 12 cases having been described in the literature previously (Table 1).3-13 We report our experience with eight episodes of Campylobaeter peritonitis and compare our cases with those previously published. MATERIALS AND METHODS
Patients Three hundred fifty-four patients have commenced CAPD in our center since 1978. During this period, there have been 1,214 episodes of peritonitis, with 49% due to staphylococcal species (coagulase-negative staphylococci, 39%) and 24% due to gram-negative bacteria. The diagnosis of peritonitis was made on the presence of turbid dialysate with subsequent dialysate microscopy con-
firming greater than 100 white blood cells per microliter of dialysate. The empirical treatment for CAPD peritonitis was as follows. After collection of the first bag of cloudy dialysate, vancomycin 30 mg/kg was administered peritoneally with the next bag. If gram-positive organisms were isolated, a second dose of vancomycin was administered intraperitonea1ly at day 5. If gram-negative organisms were identified by microscopy or culture, piperacillin, 3-g loading dose followed by I g intraperitoneally, was administered with each bag to continue for 5 days after clearing of the dialysate. If the gram-negative organism isolated subsequently proved resistant to piperacillin, then antibiotic therapy was changed to a drug to which the organism was susceptible.
Laboratory Diagnosis The methods used in our laboratory are a combination of methods described by Reller et al 14 and methods developed in-house. 15 The first bag drained following presentation (ie, onset of turbid dialysate, often in association with abdominal pain) is examined by microscopy and culture. Heparin, 500 U IL, is added at the bedside. In our cases of Campy/obacter peritonitis, delay in processing ranged from 40 minutes to 18 hours, with an average of 7 hours. Fifty milliliters of thoroughly mixed peritoneal dialysate fluid is withdrawn with a sterile needle and syringe through the dialysate bag port following disinfection with 70% alcohol. Of this, 40 mL is distributed in 20-mL aliquots between 50 mL of tryptone soya broth (Oxoid, Basingstoke, England) supplemented with 0.5 giL 4-aminobenzoic acid (BDH Chemicals, Poole, England), 0.5 giL sodium polyetholsulphonate (Liquoid, Roche Diagnostics, Basel, Switzerland) and
From the Department 0/ Nephrology, Monash Medical Centre, Clayton, Australia. Address reprint requests to Colin J. Wood, FRACP, Department o/Nephrology, Monash Medical Centre, 246 Clayton Rd, Clayton 3168, Australia. © 1992 by the National Kidney Foundation, Inc. 0272-6386/92/1903-0007$3.00/0
American Journal of Kidney Diseases, Vol XIX, No 3 (March). 1992: pp 257-263
257
258
WOOD ET AL Table 1. Other Reports ot Campy/obaeter Peritonitis
Reference (Age, Sex)
Year
Species
Diarrhea
Treatment
SulphamethoxazoleTrimethoprim IP Gentamicin IP Erythromycin 0 Cephalothin IP, 1M Tobramycln IP Cephalothin IP Cloxacllin IP Clindamycin IP Erythromycin 0 Cephazolin IP Tobramycin IP
(50, M)
1982
C jejuni
Yes'
Vernon and Dominguez (61, M) Waters et aI (68, F)
1982
NS
1985
C fetus ss intestinalis C jejuni
Yes
(62, M)
1985
C fetus ss fetus
No
Blaser et al (51, M) Kristinsson et al a. (64, F)
1986 1986
C jejuni C fetus ss fetus
NS
Pepersack et al
Wens et al
b.
Yes
(63, M)
Meigh et al
(58, F)
Papasian and Burdick (44, M) Van den Enden et al (48, F) Caruana et al
(49, M)
Goodman and Wise (63, F)
C jejuni
Yes Yes'
1988
C coli (stool) C fetus (peritoneal fluid) Ccoli
1990
C jejuni
Yes'
1990
Campy/obacter ?species
NS
1990
C jejuni
No
1988
Yes'
NS Cefuroxime IP (1 st episode) Gentamicin IP Erythromycin 0 Cefuroxime IP Erythromycin IP, 0 Vancomycin IP Netilmicin IP Erythromycin 0 Vancomycin Gentamicin Cefazolin IP Tobramycin IP Imipenem IV, IP Cephalothin IP Tobramycin IP Erythromycln Amoxycillin Flucloxacillin IV, IP Tobramycin IP Cephalothin IP Cephalexin 0
Outcome
Resolved
Resolved Resolved
Resolved but subsequent septicemia with same organism NS Resolved
Resolved Resolved
Died Resolved
Resolved
Resolved
Abbreviations: IP, intraperitoneally; 0 ,orally; 1M, intramuscularly; IV, intravenously; NS, not stated . • Diarrhea preceded by cloudy dialysate.
0.166 giL pyridoxal HCl (Calbiochem-Behring, La Jolla, CAl, and 80 mL ofthioglycollate USP broth (Oxoid) supplemented with 0.5 giL 4-aminobenzoic acid, 0.5 giL Liquoid, 0.3 giL agar bacteriological (agar no. I; Oxoid), I % vitamin K (3phytylmenadione; Sigma Chemicals, St Louis, MO) and 1% haemin (Sigma Chemicals). The tryptone soya broth (vented) and the thioglycollate broth (unvented) are incubated at 37°C for 7 days. Two drops of peritoneal dialysate fluid are stained with 0.1% toluidine blue in ethanol and leukocyte and erythrocyte counts per cubic millimeter are performed in an improved Neubauer counting chamber. The remaining peritoneal di-
alysate fluid (-10 mL) is centrifuged at 2,500 rpm for 5 minutes to increase the sensitivity of the smear results. The resuspended deposit (0.5 mL) is used to prepare a Gram stain (counterstain Safranin 0) and a pipette drop is cultured onto each of three plates: saponin-lysed horse blood agar with liquoid (Columbia base agar with 5% defibrinated horse blood and 0.05% saponin), MacConkey agar, and Sabouraud's dextrose agar. The saponin-lysed horse blood and MacConkey agar plates are incubated at 35°C, 7% CO2 for a minimum of 2 days; the Sabouraud's dextrose agar plate is incubated at 30°C, aerobically for 7 days. The broths are Gram-stained, subcultured onto routine media incubated at 35°C in CO 2
CAMPYLOBACTER PERITONITIS and anaerobically for 2 days if one or more of the following occurs: (I) there is an increase in turbidity after incubation; (2) a microorganism is isolated from the primary plates; and/ or (3) the patient is not responding to treatment. Campylobacters were identified using techniques described by Morris and Patton. 16 Susceptibility tests were performed by the disc diffusion method as described by Barrett et aI. 17 In only one of our cases was Campy/obaeter isolated from primary plates using our standard procedure. We were guided in this instance by a history of gastroenteritis and included an additional primary horse blood agar plate incubated in a "Campgas" atmosphere of 16.5% CO2 in N2 at 35°C for 2 days. Campy/obaeter species were isolated from tryptone soya broth in five of six and in thioglycolate in six of six instances where this organism was specifically sought. It was found that screening of the broth cultures with phase contrast microscopy at 48 hours was a useful technique to detect the possible presence of these organisms. This proved to be easier to read than a Gram stain of the same material.
RESULTS
Between August 1988 and July 1990,216 peritonitis episodes were documented. Campylabaeter species were isolated in eight of these episodes. In two episodes of Campylabaeter peritonitis, two morphologically distinct variants of the same species were isolated from the initial dialysate sample. Campylabaeter jejuni was the most common type of Campylabaeter isolated (n = 5) with Campylabaeter eali and Campylabaeter laridis (recently renamed as Campylabaeter lari 18 ) each isolated on one and two occasions, respectively. During the above-mentioned period, of the 208 non-Campylabaeter peritonitis episodes, diarrhea was reported as part of the symptomatology at presentation in 19%. In the vast majority of these cases, the severity of diarrhea was limited to no more than three to four loose bowel motions daily and was either coincident on the development of cloudy dialysate or noted no more than a few hours before presentation. In Campylabaeter peritonitis, diarrhea was documented in seven of the eight episodes (88%) and was severe (more than eight bowel motions daily) in four (50%). The clinical details of these eight patients presenting with Campylabaeter peritonitis are outlined in Table 2. Patients 1, 2, 3, and 4 all presented initially with severe acute enterocolitis, but clear dialysate. Clinical peritonitis (cloudy dialysate) developed between 1 and 10 days after the initial presentation (10 days, 2 days, 2 days, 1 day, respectively). Patient 2 did not have fecal
259
cultures performed and underwent immediate . removal of the peritoneal dialysis catheter following the onset of cloudy dialysate due to a persistent staphylococcal tract infection. Patients 1, 3, and 4 all had positive fecal cultures for Campylabaeter at presentation. Patients 5, 6, and 7 presented more typically with cloudy dialysate and abdominal pain, but diarrhea was not a prominent feature in these patients. Patients 5 and 7 did not have fecal cultures performed. Patient 6 had fecal cultures performed 2 weeks after the initial presentation. These cultures were negative. Patient 8 had a complicated medical course before his presentation with Campylabaeter peritonitis. He was investigated extensively due to marked malabsorption syndrome resulting in frequent loose bowel motions, fecal fat excretion of 165 g per 3-day collection, and weight loss of 30 kg over a 6-month period. No cause for his malabsorption was found, but this led to a significantly malnourished state. His presentation with Campylabaeter peritonitis was not associated with any change in bowel habit, but this was difficult to assess in the setting of his steatorrhea. His course was one of steady deterioration despite appropriate antibiotic therapy and removal of the peritoneal dialysis catheter. Blood cultures grew C jejuni (also isolated from the initial dialysate sample) and repeated fecal cultures were negative. He remained unwell following transfer to hemodialysis and died after cardiac arrest 19 days following the presentation with peritonitis. Postmortem examination showed evidence of acute myocardial infarction and generalized peritonitis. No bowel pathology was detected to explain either the malabsorption or a source of peritonitis. The susceptibility of isolated campylobacters to antibiotics is demonstrated in Table 3. All campylobacters were sensitive to gentamicin and, where tested, to erythromycin, chloramphenicol, and tetracycline. Variable sensitivities were demonstrated with piperacillin and ciprofloxacin. Other antibiotics were tested infrequently. Satisfactory clearing of the dialysate was achieved in all patients except patient 8, although patient 7 presented 3 weeks after the initial episode with a recurrence of cloudy dialysate and abdominal pain. Dialysate cultures on this occasion were negative. In patients 1, 3, and 4,
WOOD ET AL
260 Table 2. Case Reports
Patient
Age/Sex
Period on CAPD (mo)
56/M
Diarrhea
Organism
Fecal Culture
Yes·
e jejuni
Positive
2
70/F
29
Yes·
e jejuni
Not tested
3
68/F
8
Yes·
e jejuni
Positive
4
66/F
56
Yes·
e lari
Positive
Not tested
(Iaridis)
5
40/F
31
Yes
eeolil & II
6
49/F
38
No
e jejuni e lari
Negative (2 weeks postepisode) Not tested
e jejuni
Negative
1&11
7
62/M
9
Yes
(Iaridis)
8
43/M
28
Yest:
Specific Treatment
Outcome
None (Vancomycin IP) Gentamicin IV Tenkhoff removed Piperacillin IPt Gentamicin IP Erythromycin 0 Piperacillin IP Gentamicin IP Erythromycin 0 Gentamicin IP Erythromycin 0 Gentamicin IP Ciprofloxacin 0
Resolved
Gentamicin IP Erythromycin 0 Doxycycline 0 Gentamicin IP/IV Erythromycin 0 Chloramphenicol IV
Resolved Resolved
Resolved
Resolved Resolved
Resolved
Died
• Severe (>8 bowel motions daily). t Initial treatment when gram-negative organism identified, before complete identification. t: Malabsorption syndrome: fecal fat, 165 g per 3-day collection.
clearing of the dialysate was noted before the commencement of antibiotics to which the organism demonstrated in vitro sensitivity. Five patients were treated with intraperitoneal gentamicin and oral erythromycin; patient 6 received oral ciprofloxacin in combination with intraperitoneal gentamicin; and patient 1 received intraperitoneal vancomycin alone. Patient 2 received intravenous gentamicin following removal ofthe dialysis catheter. DISCUSSION
Diarrhea is relatively common in any patient presenting with CAPO peritonitis, but is usually mild and generally thought to be of little clinical relevance. Approximately 19% of peritonitis episodes in our unit are associated with diarrhea. This association is stronger with Campy/abaeter peritonitis (70% of all reported episodes) and the diarrhea can frequently be severe. Interestingly, of the common bacterial pathogens causing diarrhea in our community, only Campy/abaeter species have been isolated from dialysate in our
CAPO population. No Salmonella or Shigella species have been seen. Acute Campy/abaeter enterocolitis is most commonly due to C jejuni, although a similar presentation can occur with C eali and C lari. 19 C fetus generally causes a bacteremic illness without significant diarrheal disease and shows a predilection for endovascular surfaces in immunocompromised patients. 2o This clinical variability is also seen in Campy/abaeter peritonitis. Peritonitis due to C jejuni, C eali, or C lari is often associated with severe diarrhea, which can precede the onset of cloudy dialysate by several days. This finding is supported by other reports of Campy/abaeter peritonitis (Table 1)3,4,8-11 where although the severity of the diarrhea is not stated, it is again noted to commonly precede the onset of peritonitis. It is unusual for diarrhea to precede non-Campy/abaeter peritonitis by more than a few hours. Diarrhea is less likely in C fetus peritonitis. In the four previously reported cases, diarrhea was present on two occasions and in both cases was considered most likely due to an
CAMPYLOBACTER PERITONITIS
261 Table 3. Susceptibility of Isolates Susceptibility Test Result
Patient No. Species
Gentamicin' Chloramphenicol' Tetracycline' Erythromycin' Piperacillint Ciprofloxacint Imipenemt Cefotaximet Ceftriaxonet Aztreonamt Ampicillint
1
2
3
4
5
5
6
6
7
8
CJ
CJ
CJ
CL
CC
CC
CJ
CJ
CL
CJ
S S S S S S
S S S S I S S S
S S S S R
S S S S R R
S
S
S
S
S I S
S S S
S S S S
R
R
S S S S S
S
S
R
R R
S
Abbreviations: CJ, Campy/obaeter jejuni; CL, Campy/obaeter /ari (/aridis); CC, Campy/obaeter coli; S, susceptible; L, intermediate; R, resistant; - , not done. , Interpretive zone diameters as per Barrett et al. 16 t Interpretive zone diameters as per National Committee on Clinical Laboratory Standards. 28
alternative organism, ie, Clastridium dif.fieile or Campylabaeter eali which was isolated from the patients' feces but not the dialysate. 8 ,9 Possible modes of spread of Campylabaeter to the peritoneal cavity in our patients is only speculative. It can potentially arise from either outside contamination at the time of bag change procedure, transmural migration of organisms across the intestinal wall, or bacteremic spread of organisms. Campylabaeter contamination of the fingertips is possible during the period of fecal shedding; however, relatively simple methods of routine hygiene can eradicate this contamination?1 Poor technique and poor hygiene have not been a feature of our patients' course on peritoneal dialysis and, although not excluded, touch contamination with transluminal spread of organisms would seem less likely ~ Reports of spontaneous bacterial peritonitis occurring in cirrhotic patients due to organisms of enteric origin (including C jejuni and C eali) are generally quoted in support of transmural migration of organisms resulting in peritonitis.22-24 Transmural migration is most likely to occur when the bowel wall is acutely inflamed, but evidence supporting this mode of spread is largely indirect. Bacteremia appears to be an unusual manifestation of Campylabaeter infection. Less than 1% of patients have had documented positive
blood cultures. 25 This has been explained in part by the infrequency with which blood cultures are taken in patients with acute diarrhea even when febrile. Blood cultures are also performed infrequently in CAPD patients presenting with peritonitis. Our patient (no. 8) is the first report of C jejuni bacteremia occurring in association with CAPD peritonitis. C fetus bacteremia has been documented in a CAPD patient 3 weeks after a peritonitis episode. 6 However, in an ongoing surveillance program of significant blood culture isolates in our state (Victoria) in Australia, 11 isolates of Campylabaeter species (C jejuni, 5; C eali, 4, Campylabaeter speeies, 2) have been documented over a 2-year period, suggesting that bacteremia with Campylabaeter may be more common than previously thought (J. Forsyth, personal communication). Blaser et aC postulate that either increased virulence, eg, resistance to bactericidal effects of normal serum or a host immune defect such as may occur in chronic renal failure, is important in the development of extraintestinal Campylabaeter infection. Transmission to the peritoneal cavity by the bloodstream is more likely with C fetus given that this organism is resistant to the bactericidal activity of human serum, whereas C jejuni and C eali are often serum-sensitive. 26 Despite the speculation on the differing mechanisms of developing Campylabaeter peritonitis, it is
WOOD ET AL
262
probable that no single mode of spread explains all episodes. Spontaneous resolution of Campylabaeter peritonitis has not previously been reported; however, three patients in our series were noted to have clearing of the dialysate before the commencement of appropriate antibiotic. This may relate to the organism's sensitivity to host-defense mechanisms, but full treatment with appropriate antibiotics is still advocated. Many differing antibiotic regimens have been used to treat Campylabaeter peritonitis, but intraperitoneal aminoglycoside in combination with oral erythromycin would seem the most appropriate approach given the results of antibiotic sensitivity profiles. Sensitivity patterns on the organisms isolated in our patients have demonstrated sensitivity to aminoglycodies and erythromycin, whereas sensitivity to Ciprofioxin and cephalosporins was variable. Campylobacters are reported as generally resistant to cephalosporins. 27
A search for Campylabaeter should be undertaken if there is a history of diarrhea either before or at the time of presentation with CAPO-associated peritonitis. Inclusion of a blood agar plate in the initial cultures incubated in an atmosphere appropriate for Campylabaeter species is recommended in these circumstances. Routine screening of all broth cultures by phase contrast microscopy also appears to be sensitive in detecting possible Campylabaeter species and would be simple to include in the normal laboratory protocol. It is likely that campylobacters cause a small proportion of culture-negative CAPO peritonitis and therefore a prospective search using the above techniques would be worthwhile. Refinements in microbiological techniques coupled with an increased awareness of the potential for Campylabaeter species to cause peritonitis in CAPO patients has resulted in this entity being increasingly recognized.
REFERENCES 1. Tranaeus A, Heimburger 0, Lindholm B: Peritonitis during continuous ambulatory peritoneal dialysis. Risk factors, clinical severity and pathogenetic aspects. Perit Dial Bull 8: 253-263, 1988 2. Vas SI: Peritonitis, in Nolph K (ed): Peritoneal Dialysis. The Hague, Netherlands. Nihoff, 1985, pp 403-439 3. Pepersack F, D'Haene MD, Toussaint C, et al: Campylobaeter jejuni peritonitis complicating continuous ambulatory peritoneal dialysis. J Clin Microbiol 16:739-741, 1982 4. Vernon SE, Dominguez C: Campylobacter and peritoneal dialysis. Ann Intern Med 96:534, 1982 5. Waters MJ, Andrew JH, Wright CA, et al: Campylobaeter jejuni peritonitis-Treatment in a patient on CAPD. Dialysis 5:2-3, 1985 6. Wens R, Dratwa M, Potvliege C, et al: Campylobaeter fetus peritonitis followed by septicaemia in a patient on continuous ambulatory peritoneal dialysis. J Infect 10:249-251, 1985 7. Blaser MJ, Perez GP, Smith PF, et al: Extraintestinal Campylobaeter jejuni and Campylobaeter coli infections: Host factors and strain characteristics. J Infect Dis 153:552-559, 1986 8. Kristinsson KG, Spencer RC, Brown CB: Campylobacter peritonitis in continuous ambulatory peritoneal dialysis. J Infect 13: 199-204, 1986 9. Meigh JA, Wilkie ME, Ackrill P: Infection with Campylobaeter coli and Campylobaeter fetus in a patient on continuous ambulatory peritoneal dialysis. J Infect 17:189-191, 1988 10. Papasian CJ, Burdick C: Campylobacter peritonitis. Clin Microbiol Newslet 10:102, 1988 11. Van den Enden WJ, Cabuy A, Daelemans Verhaegen J, et al: Campylobaeter jejuni peritonitis during chronic am-
bulatory peritoneal dialysis (CAPD). Perit Dial Bull 10: 177178,1990 12. Caruana J, Smith KL, Hess CK, et al: Campylobacter peritonitis in a CAPD patient. Perit Dial Bull 10: 10 1-102, 1990 13. Goodman D, Wise K: Peritonitis caused by Campylobaeter jejuni and serologically confirmed in a patient being treated with continuous ambulatory peritoneal dialysis. J Infect 21:71-75, 1990 14. Reller LB, Murray PR, McLowry JD: Blood cultures II, in Washington JA (ed): Cumulative Techniques and Procedures in Clinical Microbiology. Washington, DC, American Society for Microbiology, 1982, pp I-II 15. Matthews BJ: A comparison of four solid media for microbial yield from continuous ambulatory peritoneal dialysis (CAPD) effluent. Aust Microbiologist 11:225, 1990 (abstr) 16. Morris GK, Patton CM: Campylobacter, Lennette EH, Balows A, Hausler WJ, et al (eds): Manual of Clinical Microbiology (ed 4). Washington, DC, American Society for Microbiology, 1985, pp 302-308 17. Barrett JE, Kaplan RL, Goodman U: Campylobacter: A method for routine antimicrobial susceptibility testing, Pearson AD, Skirrow MB, Row B, et al (eds): Campylobacter II. London, UK, Public Health Laboratory Service, 1983, pp 26-27 18. von Graevenitz A: Revised nomenclature of Campylobaeter laridis, Enterobaeter intermedium and "Flavobacterium branehiophila." Int J System Bacteriol 40:211, 1990 19. Tee W, Anderson B, Ross B, et al: Atypical campylobacter associated with gastroenteritis. J Clin Microbiol 25: 1248-1252, 1987
CAMPYLOBACTER PERITONITIS
20. Schmidt U, Chme! H, Kaminski Z, et al: The clinical spectrum of Campy/obaeter fetus infections. Reports of five cases and review of the literature. Q J Med 196:431-442, 1980 21. Coates D, Hutchinson DV, Bolton FS: Survival of thermophilic campylobacter on finger-tips and their elimination by washing and disinfection. Epidemiol Infect 99:265274, 1987 22. McNeil NI, Buttoo S, Ridgway GL: Spontaneous bacterial peritonitis due to Campy/obaeter jejuni. Postgrad Med J 60:487-488, 1984 23. Domingo P, Mirelis B, Gimeno A, et al: Spontaneous peritonitis caused by Campy/obaeter jejuni in a cirrhotic patient. Med Clin (Barc) 84:416, 1985 (letter) 24. Ho H, Zuckerman MJ, Polly SM: Spontaneous bac-
263 terial peritonitis due to Campylobaeter eoli. Gastroenterology 92:2024-2025, 1987 25. Blaser M: Campy/obaeter species, in Mandell GL, Douglas RG, Bennett JE (eds): Principles and Practice ofInfectious Diseases (ed 3). New York, NY, Churchill Livingstone, 1990, pp 1649-1658 26. Blaser MJ, Smith PF, Kohler PF: Susceptibility of campylobacter isolates to the bactericidal activity of human serum. J Infect Dis 151:227-235,1985 27. Wright EP, Knowles MA: Beta-lactamase production by Campy/obaeter jejuni. J Clin Pathol 33:904-905, 1980 28. National Committee for Clinical Laboratory Standards (NCCLS): Performance Standards for Antimicrobial Disk Susceptibility Tests (ed 4). Villanova, PA, NCCLS Document M2-A4,1990
C
ONTINUOUS AMBULATORY peritoneal dialysis (CAPD) is now widely accepted as effective treatment for end-stage renal failure, but peritonitis remains the most common complication. I Staphylococci, particularly coagulasenegative species, are the most frequently isolated organisms, accounting for approximately 50% of all reported peritonitis episodes. Gram-negative bacilli, although occurring less commonly, are isolated in approximately 20% of peritonitis episodes, but this proportion can vary widely.2 Campylobacters are micro-aerophilic curved or spiral gram-negative bacteria recognized in the 1970s as a common cause of acute enterocolitis in man. They occur in the intestinal flora of many wild and domestic animals and poultry and can be transmitted to humans by milk, water, and undercooked meat polluted by such animal carriers. Although a common cause of enterocolitis, they are rare causes of CAPD-associated peritonitis, with only 12 cases having been described in the literature previously (Table 1).3-13 We report our experience with eight episodes of Campylobaeter peritonitis and compare our cases with those previously published. MATERIALS AND METHODS
Patients Three hundred fifty-four patients have commenced CAPD in our center since 1978. During this period, there have been 1,214 episodes of peritonitis, with 49% due to staphylococcal species (coagulase-negative staphylococci, 39%) and 24% due to gram-negative bacteria. The diagnosis of peritonitis was made on the presence of turbid dialysate with subsequent dialysate microscopy con-
firming greater than 100 white blood cells per microliter of dialysate. The empirical treatment for CAPD peritonitis was as follows. After collection of the first bag of cloudy dialysate, vancomycin 30 mg/kg was administered peritoneally with the next bag. If gram-positive organisms were isolated, a second dose of vancomycin was administered intraperitonea1ly at day 5. If gram-negative organisms were identified by microscopy or culture, piperacillin, 3-g loading dose followed by I g intraperitoneally, was administered with each bag to continue for 5 days after clearing of the dialysate. If the gram-negative organism isolated subsequently proved resistant to piperacillin, then antibiotic therapy was changed to a drug to which the organism was susceptible.
Laboratory Diagnosis The methods used in our laboratory are a combination of methods described by Reller et al 14 and methods developed in-house. 15 The first bag drained following presentation (ie, onset of turbid dialysate, often in association with abdominal pain) is examined by microscopy and culture. Heparin, 500 U IL, is added at the bedside. In our cases of Campy/obacter peritonitis, delay in processing ranged from 40 minutes to 18 hours, with an average of 7 hours. Fifty milliliters of thoroughly mixed peritoneal dialysate fluid is withdrawn with a sterile needle and syringe through the dialysate bag port following disinfection with 70% alcohol. Of this, 40 mL is distributed in 20-mL aliquots between 50 mL of tryptone soya broth (Oxoid, Basingstoke, England) supplemented with 0.5 giL 4-aminobenzoic acid (BDH Chemicals, Poole, England), 0.5 giL sodium polyetholsulphonate (Liquoid, Roche Diagnostics, Basel, Switzerland) and
From the Department 0/ Nephrology, Monash Medical Centre, Clayton, Australia. Address reprint requests to Colin J. Wood, FRACP, Department o/Nephrology, Monash Medical Centre, 246 Clayton Rd, Clayton 3168, Australia. © 1992 by the National Kidney Foundation, Inc. 0272-6386/92/1903-0007$3.00/0
American Journal of Kidney Diseases, Vol XIX, No 3 (March). 1992: pp 257-263
257
258
WOOD ET AL Table 1. Other Reports ot Campy/obaeter Peritonitis
Reference (Age, Sex)
Year
Species
Diarrhea
Treatment
SulphamethoxazoleTrimethoprim IP Gentamicin IP Erythromycin 0 Cephalothin IP, 1M Tobramycln IP Cephalothin IP Cloxacllin IP Clindamycin IP Erythromycin 0 Cephazolin IP Tobramycin IP
(50, M)
1982
C jejuni
Yes'
Vernon and Dominguez (61, M) Waters et aI (68, F)
1982
NS
1985
C fetus ss intestinalis C jejuni
Yes
(62, M)
1985
C fetus ss fetus
No
Blaser et al (51, M) Kristinsson et al a. (64, F)
1986 1986
C jejuni C fetus ss fetus
NS
Pepersack et al
Wens et al
b.
Yes
(63, M)
Meigh et al
(58, F)
Papasian and Burdick (44, M) Van den Enden et al (48, F) Caruana et al
(49, M)
Goodman and Wise (63, F)
C jejuni
Yes Yes'
1988
C coli (stool) C fetus (peritoneal fluid) Ccoli
1990
C jejuni
Yes'
1990
Campy/obacter ?species
NS
1990
C jejuni
No
1988
Yes'
NS Cefuroxime IP (1 st episode) Gentamicin IP Erythromycin 0 Cefuroxime IP Erythromycin IP, 0 Vancomycin IP Netilmicin IP Erythromycin 0 Vancomycin Gentamicin Cefazolin IP Tobramycin IP Imipenem IV, IP Cephalothin IP Tobramycin IP Erythromycln Amoxycillin Flucloxacillin IV, IP Tobramycin IP Cephalothin IP Cephalexin 0
Outcome
Resolved
Resolved Resolved
Resolved but subsequent septicemia with same organism NS Resolved
Resolved Resolved
Died Resolved
Resolved
Resolved
Abbreviations: IP, intraperitoneally; 0 ,orally; 1M, intramuscularly; IV, intravenously; NS, not stated . • Diarrhea preceded by cloudy dialysate.
0.166 giL pyridoxal HCl (Calbiochem-Behring, La Jolla, CAl, and 80 mL ofthioglycollate USP broth (Oxoid) supplemented with 0.5 giL 4-aminobenzoic acid, 0.5 giL Liquoid, 0.3 giL agar bacteriological (agar no. I; Oxoid), I % vitamin K (3phytylmenadione; Sigma Chemicals, St Louis, MO) and 1% haemin (Sigma Chemicals). The tryptone soya broth (vented) and the thioglycollate broth (unvented) are incubated at 37°C for 7 days. Two drops of peritoneal dialysate fluid are stained with 0.1% toluidine blue in ethanol and leukocyte and erythrocyte counts per cubic millimeter are performed in an improved Neubauer counting chamber. The remaining peritoneal di-
alysate fluid (-10 mL) is centrifuged at 2,500 rpm for 5 minutes to increase the sensitivity of the smear results. The resuspended deposit (0.5 mL) is used to prepare a Gram stain (counterstain Safranin 0) and a pipette drop is cultured onto each of three plates: saponin-lysed horse blood agar with liquoid (Columbia base agar with 5% defibrinated horse blood and 0.05% saponin), MacConkey agar, and Sabouraud's dextrose agar. The saponin-lysed horse blood and MacConkey agar plates are incubated at 35°C, 7% CO2 for a minimum of 2 days; the Sabouraud's dextrose agar plate is incubated at 30°C, aerobically for 7 days. The broths are Gram-stained, subcultured onto routine media incubated at 35°C in CO 2
CAMPYLOBACTER PERITONITIS and anaerobically for 2 days if one or more of the following occurs: (I) there is an increase in turbidity after incubation; (2) a microorganism is isolated from the primary plates; and/ or (3) the patient is not responding to treatment. Campylobacters were identified using techniques described by Morris and Patton. 16 Susceptibility tests were performed by the disc diffusion method as described by Barrett et aI. 17 In only one of our cases was Campy/obaeter isolated from primary plates using our standard procedure. We were guided in this instance by a history of gastroenteritis and included an additional primary horse blood agar plate incubated in a "Campgas" atmosphere of 16.5% CO2 in N2 at 35°C for 2 days. Campy/obaeter species were isolated from tryptone soya broth in five of six and in thioglycolate in six of six instances where this organism was specifically sought. It was found that screening of the broth cultures with phase contrast microscopy at 48 hours was a useful technique to detect the possible presence of these organisms. This proved to be easier to read than a Gram stain of the same material.
RESULTS
Between August 1988 and July 1990,216 peritonitis episodes were documented. Campylabaeter species were isolated in eight of these episodes. In two episodes of Campylabaeter peritonitis, two morphologically distinct variants of the same species were isolated from the initial dialysate sample. Campylabaeter jejuni was the most common type of Campylabaeter isolated (n = 5) with Campylabaeter eali and Campylabaeter laridis (recently renamed as Campylabaeter lari 18 ) each isolated on one and two occasions, respectively. During the above-mentioned period, of the 208 non-Campylabaeter peritonitis episodes, diarrhea was reported as part of the symptomatology at presentation in 19%. In the vast majority of these cases, the severity of diarrhea was limited to no more than three to four loose bowel motions daily and was either coincident on the development of cloudy dialysate or noted no more than a few hours before presentation. In Campylabaeter peritonitis, diarrhea was documented in seven of the eight episodes (88%) and was severe (more than eight bowel motions daily) in four (50%). The clinical details of these eight patients presenting with Campylabaeter peritonitis are outlined in Table 2. Patients 1, 2, 3, and 4 all presented initially with severe acute enterocolitis, but clear dialysate. Clinical peritonitis (cloudy dialysate) developed between 1 and 10 days after the initial presentation (10 days, 2 days, 2 days, 1 day, respectively). Patient 2 did not have fecal
259
cultures performed and underwent immediate . removal of the peritoneal dialysis catheter following the onset of cloudy dialysate due to a persistent staphylococcal tract infection. Patients 1, 3, and 4 all had positive fecal cultures for Campylabaeter at presentation. Patients 5, 6, and 7 presented more typically with cloudy dialysate and abdominal pain, but diarrhea was not a prominent feature in these patients. Patients 5 and 7 did not have fecal cultures performed. Patient 6 had fecal cultures performed 2 weeks after the initial presentation. These cultures were negative. Patient 8 had a complicated medical course before his presentation with Campylabaeter peritonitis. He was investigated extensively due to marked malabsorption syndrome resulting in frequent loose bowel motions, fecal fat excretion of 165 g per 3-day collection, and weight loss of 30 kg over a 6-month period. No cause for his malabsorption was found, but this led to a significantly malnourished state. His presentation with Campylabaeter peritonitis was not associated with any change in bowel habit, but this was difficult to assess in the setting of his steatorrhea. His course was one of steady deterioration despite appropriate antibiotic therapy and removal of the peritoneal dialysis catheter. Blood cultures grew C jejuni (also isolated from the initial dialysate sample) and repeated fecal cultures were negative. He remained unwell following transfer to hemodialysis and died after cardiac arrest 19 days following the presentation with peritonitis. Postmortem examination showed evidence of acute myocardial infarction and generalized peritonitis. No bowel pathology was detected to explain either the malabsorption or a source of peritonitis. The susceptibility of isolated campylobacters to antibiotics is demonstrated in Table 3. All campylobacters were sensitive to gentamicin and, where tested, to erythromycin, chloramphenicol, and tetracycline. Variable sensitivities were demonstrated with piperacillin and ciprofloxacin. Other antibiotics were tested infrequently. Satisfactory clearing of the dialysate was achieved in all patients except patient 8, although patient 7 presented 3 weeks after the initial episode with a recurrence of cloudy dialysate and abdominal pain. Dialysate cultures on this occasion were negative. In patients 1, 3, and 4,
WOOD ET AL
260 Table 2. Case Reports
Patient
Age/Sex
Period on CAPD (mo)
56/M
Diarrhea
Organism
Fecal Culture
Yes·
e jejuni
Positive
2
70/F
29
Yes·
e jejuni
Not tested
3
68/F
8
Yes·
e jejuni
Positive
4
66/F
56
Yes·
e lari
Positive
Not tested
(Iaridis)
5
40/F
31
Yes
eeolil & II
6
49/F
38
No
e jejuni e lari
Negative (2 weeks postepisode) Not tested
e jejuni
Negative
1&11
7
62/M
9
Yes
(Iaridis)
8
43/M
28
Yest:
Specific Treatment
Outcome
None (Vancomycin IP) Gentamicin IV Tenkhoff removed Piperacillin IPt Gentamicin IP Erythromycin 0 Piperacillin IP Gentamicin IP Erythromycin 0 Gentamicin IP Erythromycin 0 Gentamicin IP Ciprofloxacin 0
Resolved
Gentamicin IP Erythromycin 0 Doxycycline 0 Gentamicin IP/IV Erythromycin 0 Chloramphenicol IV
Resolved Resolved
Resolved
Resolved Resolved
Resolved
Died
• Severe (>8 bowel motions daily). t Initial treatment when gram-negative organism identified, before complete identification. t: Malabsorption syndrome: fecal fat, 165 g per 3-day collection.
clearing of the dialysate was noted before the commencement of antibiotics to which the organism demonstrated in vitro sensitivity. Five patients were treated with intraperitoneal gentamicin and oral erythromycin; patient 6 received oral ciprofloxacin in combination with intraperitoneal gentamicin; and patient 1 received intraperitoneal vancomycin alone. Patient 2 received intravenous gentamicin following removal ofthe dialysis catheter. DISCUSSION
Diarrhea is relatively common in any patient presenting with CAPO peritonitis, but is usually mild and generally thought to be of little clinical relevance. Approximately 19% of peritonitis episodes in our unit are associated with diarrhea. This association is stronger with Campy/abaeter peritonitis (70% of all reported episodes) and the diarrhea can frequently be severe. Interestingly, of the common bacterial pathogens causing diarrhea in our community, only Campy/abaeter species have been isolated from dialysate in our
CAPO population. No Salmonella or Shigella species have been seen. Acute Campy/abaeter enterocolitis is most commonly due to C jejuni, although a similar presentation can occur with C eali and C lari. 19 C fetus generally causes a bacteremic illness without significant diarrheal disease and shows a predilection for endovascular surfaces in immunocompromised patients. 2o This clinical variability is also seen in Campy/abaeter peritonitis. Peritonitis due to C jejuni, C eali, or C lari is often associated with severe diarrhea, which can precede the onset of cloudy dialysate by several days. This finding is supported by other reports of Campy/abaeter peritonitis (Table 1)3,4,8-11 where although the severity of the diarrhea is not stated, it is again noted to commonly precede the onset of peritonitis. It is unusual for diarrhea to precede non-Campy/abaeter peritonitis by more than a few hours. Diarrhea is less likely in C fetus peritonitis. In the four previously reported cases, diarrhea was present on two occasions and in both cases was considered most likely due to an
CAMPYLOBACTER PERITONITIS
261 Table 3. Susceptibility of Isolates Susceptibility Test Result
Patient No. Species
Gentamicin' Chloramphenicol' Tetracycline' Erythromycin' Piperacillint Ciprofloxacint Imipenemt Cefotaximet Ceftriaxonet Aztreonamt Ampicillint
1
2
3
4
5
5
6
6
7
8
CJ
CJ
CJ
CL
CC
CC
CJ
CJ
CL
CJ
S S S S S S
S S S S I S S S
S S S S R
S S S S R R
S
S
S
S
S I S
S S S
S S S S
R
R
S S S S S
S
S
R
R R
S
Abbreviations: CJ, Campy/obaeter jejuni; CL, Campy/obaeter /ari (/aridis); CC, Campy/obaeter coli; S, susceptible; L, intermediate; R, resistant; - , not done. , Interpretive zone diameters as per Barrett et al. 16 t Interpretive zone diameters as per National Committee on Clinical Laboratory Standards. 28
alternative organism, ie, Clastridium dif.fieile or Campylabaeter eali which was isolated from the patients' feces but not the dialysate. 8 ,9 Possible modes of spread of Campylabaeter to the peritoneal cavity in our patients is only speculative. It can potentially arise from either outside contamination at the time of bag change procedure, transmural migration of organisms across the intestinal wall, or bacteremic spread of organisms. Campylabaeter contamination of the fingertips is possible during the period of fecal shedding; however, relatively simple methods of routine hygiene can eradicate this contamination?1 Poor technique and poor hygiene have not been a feature of our patients' course on peritoneal dialysis and, although not excluded, touch contamination with transluminal spread of organisms would seem less likely ~ Reports of spontaneous bacterial peritonitis occurring in cirrhotic patients due to organisms of enteric origin (including C jejuni and C eali) are generally quoted in support of transmural migration of organisms resulting in peritonitis.22-24 Transmural migration is most likely to occur when the bowel wall is acutely inflamed, but evidence supporting this mode of spread is largely indirect. Bacteremia appears to be an unusual manifestation of Campylabaeter infection. Less than 1% of patients have had documented positive
blood cultures. 25 This has been explained in part by the infrequency with which blood cultures are taken in patients with acute diarrhea even when febrile. Blood cultures are also performed infrequently in CAPD patients presenting with peritonitis. Our patient (no. 8) is the first report of C jejuni bacteremia occurring in association with CAPD peritonitis. C fetus bacteremia has been documented in a CAPD patient 3 weeks after a peritonitis episode. 6 However, in an ongoing surveillance program of significant blood culture isolates in our state (Victoria) in Australia, 11 isolates of Campylabaeter species (C jejuni, 5; C eali, 4, Campylabaeter speeies, 2) have been documented over a 2-year period, suggesting that bacteremia with Campylabaeter may be more common than previously thought (J. Forsyth, personal communication). Blaser et aC postulate that either increased virulence, eg, resistance to bactericidal effects of normal serum or a host immune defect such as may occur in chronic renal failure, is important in the development of extraintestinal Campylabaeter infection. Transmission to the peritoneal cavity by the bloodstream is more likely with C fetus given that this organism is resistant to the bactericidal activity of human serum, whereas C jejuni and C eali are often serum-sensitive. 26 Despite the speculation on the differing mechanisms of developing Campylabaeter peritonitis, it is
WOOD ET AL
262
probable that no single mode of spread explains all episodes. Spontaneous resolution of Campylabaeter peritonitis has not previously been reported; however, three patients in our series were noted to have clearing of the dialysate before the commencement of appropriate antibiotic. This may relate to the organism's sensitivity to host-defense mechanisms, but full treatment with appropriate antibiotics is still advocated. Many differing antibiotic regimens have been used to treat Campylabaeter peritonitis, but intraperitoneal aminoglycoside in combination with oral erythromycin would seem the most appropriate approach given the results of antibiotic sensitivity profiles. Sensitivity patterns on the organisms isolated in our patients have demonstrated sensitivity to aminoglycodies and erythromycin, whereas sensitivity to Ciprofioxin and cephalosporins was variable. Campylobacters are reported as generally resistant to cephalosporins. 27
A search for Campylabaeter should be undertaken if there is a history of diarrhea either before or at the time of presentation with CAPO-associated peritonitis. Inclusion of a blood agar plate in the initial cultures incubated in an atmosphere appropriate for Campylabaeter species is recommended in these circumstances. Routine screening of all broth cultures by phase contrast microscopy also appears to be sensitive in detecting possible Campylabaeter species and would be simple to include in the normal laboratory protocol. It is likely that campylobacters cause a small proportion of culture-negative CAPO peritonitis and therefore a prospective search using the above techniques would be worthwhile. Refinements in microbiological techniques coupled with an increased awareness of the potential for Campylabaeter species to cause peritonitis in CAPO patients has resulted in this entity being increasingly recognized.
REFERENCES 1. Tranaeus A, Heimburger 0, Lindholm B: Peritonitis during continuous ambulatory peritoneal dialysis. Risk factors, clinical severity and pathogenetic aspects. Perit Dial Bull 8: 253-263, 1988 2. Vas SI: Peritonitis, in Nolph K (ed): Peritoneal Dialysis. The Hague, Netherlands. Nihoff, 1985, pp 403-439 3. Pepersack F, D'Haene MD, Toussaint C, et al: Campylobaeter jejuni peritonitis complicating continuous ambulatory peritoneal dialysis. J Clin Microbiol 16:739-741, 1982 4. Vernon SE, Dominguez C: Campylobacter and peritoneal dialysis. Ann Intern Med 96:534, 1982 5. Waters MJ, Andrew JH, Wright CA, et al: Campylobaeter jejuni peritonitis-Treatment in a patient on CAPD. Dialysis 5:2-3, 1985 6. Wens R, Dratwa M, Potvliege C, et al: Campylobaeter fetus peritonitis followed by septicaemia in a patient on continuous ambulatory peritoneal dialysis. J Infect 10:249-251, 1985 7. Blaser MJ, Perez GP, Smith PF, et al: Extraintestinal Campylobaeter jejuni and Campylobaeter coli infections: Host factors and strain characteristics. J Infect Dis 153:552-559, 1986 8. Kristinsson KG, Spencer RC, Brown CB: Campylobacter peritonitis in continuous ambulatory peritoneal dialysis. J Infect 13: 199-204, 1986 9. Meigh JA, Wilkie ME, Ackrill P: Infection with Campylobaeter coli and Campylobaeter fetus in a patient on continuous ambulatory peritoneal dialysis. J Infect 17:189-191, 1988 10. Papasian CJ, Burdick C: Campylobacter peritonitis. Clin Microbiol Newslet 10:102, 1988 11. Van den Enden WJ, Cabuy A, Daelemans Verhaegen J, et al: Campylobaeter jejuni peritonitis during chronic am-
bulatory peritoneal dialysis (CAPD). Perit Dial Bull 10: 177178,1990 12. Caruana J, Smith KL, Hess CK, et al: Campylobacter peritonitis in a CAPD patient. Perit Dial Bull 10: 10 1-102, 1990 13. Goodman D, Wise K: Peritonitis caused by Campylobaeter jejuni and serologically confirmed in a patient being treated with continuous ambulatory peritoneal dialysis. J Infect 21:71-75, 1990 14. Reller LB, Murray PR, McLowry JD: Blood cultures II, in Washington JA (ed): Cumulative Techniques and Procedures in Clinical Microbiology. Washington, DC, American Society for Microbiology, 1982, pp I-II 15. Matthews BJ: A comparison of four solid media for microbial yield from continuous ambulatory peritoneal dialysis (CAPD) effluent. Aust Microbiologist 11:225, 1990 (abstr) 16. Morris GK, Patton CM: Campylobacter, Lennette EH, Balows A, Hausler WJ, et al (eds): Manual of Clinical Microbiology (ed 4). Washington, DC, American Society for Microbiology, 1985, pp 302-308 17. Barrett JE, Kaplan RL, Goodman U: Campylobacter: A method for routine antimicrobial susceptibility testing, Pearson AD, Skirrow MB, Row B, et al (eds): Campylobacter II. London, UK, Public Health Laboratory Service, 1983, pp 26-27 18. von Graevenitz A: Revised nomenclature of Campylobaeter laridis, Enterobaeter intermedium and "Flavobacterium branehiophila." Int J System Bacteriol 40:211, 1990 19. Tee W, Anderson B, Ross B, et al: Atypical campylobacter associated with gastroenteritis. J Clin Microbiol 25: 1248-1252, 1987
CAMPYLOBACTER PERITONITIS
20. Schmidt U, Chme! H, Kaminski Z, et al: The clinical spectrum of Campy/obaeter fetus infections. Reports of five cases and review of the literature. Q J Med 196:431-442, 1980 21. Coates D, Hutchinson DV, Bolton FS: Survival of thermophilic campylobacter on finger-tips and their elimination by washing and disinfection. Epidemiol Infect 99:265274, 1987 22. McNeil NI, Buttoo S, Ridgway GL: Spontaneous bacterial peritonitis due to Campy/obaeter jejuni. Postgrad Med J 60:487-488, 1984 23. Domingo P, Mirelis B, Gimeno A, et al: Spontaneous peritonitis caused by Campy/obaeter jejuni in a cirrhotic patient. Med Clin (Barc) 84:416, 1985 (letter) 24. Ho H, Zuckerman MJ, Polly SM: Spontaneous bac-
263 terial peritonitis due to Campylobaeter eoli. Gastroenterology 92:2024-2025, 1987 25. Blaser M: Campy/obaeter species, in Mandell GL, Douglas RG, Bennett JE (eds): Principles and Practice ofInfectious Diseases (ed 3). New York, NY, Churchill Livingstone, 1990, pp 1649-1658 26. Blaser MJ, Smith PF, Kohler PF: Susceptibility of campylobacter isolates to the bactericidal activity of human serum. J Infect Dis 151:227-235,1985 27. Wright EP, Knowles MA: Beta-lactamase production by Campy/obaeter jejuni. J Clin Pathol 33:904-905, 1980 28. National Committee for Clinical Laboratory Standards (NCCLS): Performance Standards for Antimicrobial Disk Susceptibility Tests (ed 4). Villanova, PA, NCCLS Document M2-A4,1990
