Abdominal Catastrophe: Definition and Proposal for a New Approach

Lynda N. Newman,1 Carolyn P. Cacho,2 James A. Schulak,3 Miriam F. Weiss2
From: 1Department of Nursing, Division of Nephrology; 2Department of Internal Medicine; and 3Department of Surgery, University Hospitals of Cleveland, Cleveland, Ohio, U.S.A.



Abdominal catastrophe, defined as peritonitis from a visceral source, occurs in a significant number of patients treated by peritoneal dialysis. Peritonitis due to visceral injury is difficult to manage and is associated with high morbidity and mortality. Surgical intervention for both diagnosis and repair is definitive. However, no preventive strategy has been identified to date.
The experience at University Hospitals of Cleveland and the published experiences of many other centers demonstrate that the risk of this complication has not changed in parallel with the many other improvements in the technique and outcome of peritoneal dialysis.
We propose an approach to improve the understanding and outcome of this devastating complication. First, classification of peritonitis by source, not by organism, may lead to a more focused response to each episode of peritonitis. Second, the importance of antibiotic prophylaxis should be re-assessed in defined clinical settings that have a high likelihood of progressing to abdominal catastrophe. Third, optimal antibiotic regimens need to be devised and applied when visceral injury is highly suspected as a cause of peritonitis. Finally, the results of surgical interventions must be carefully studied.

Key words

Peritonitis, gram-negative peritonitis, enteric peritonitis, abdominal catastrophe, visceral injury

Introduction

Abdominal catastrophe is a serious illness with a high mortality rate. Visceral injury, causing leakage of enteric organisms into the peritoneal cavity, can result in peritonitis that ranges in severity from a self-limited and easily treated episode to an overwhelming disease process. The incidence, clinical course, and mortality rates of peritonitis of visceral origin have not changed over more than 20 years. The complication represents between 3.5% and 25% of published series of peritoneal dialysis–associated peritonitis, with a mortality rate of approximately 50% (1–6). Tables I and II summarize the current literature.
Abdominal catastrophe is difficult to diagnose. The initial laboratory presentation is not different from peritonitis of other etiologies (1,2,4,5). Often, the response to initial therapy provides the first clue to the severity of the episode. Peritonitis that fails to clear after 48 – 72 hours of appropriate antibiotic therapy is more likely to represent a severe and continuous contamination that may require surgical intervention (8). Many peritoneal dialysis units manage persistent peritonitis of any cause by catheter removal without further surgical exploration. The rationale for this approach is to allow normal peritoneal defenses to seal any source of bacterial contamination. Often, when catheter removal or exploratory laparotomy is contemplated, the initiating event is already 2–3 days old. By this time, even successful operative management may not reverse the morbid disease process (3,5,8). In addition, if visceral injury remains undiagnosed, abscess formation may ensue, causing inanition and a chronic inflammatory wasting illness (12).
This paper incorporates an abbreviated review of the literature on peritonitis due to visceral injury. We also compare abdominal catastrophe with all sources of peritonitis during two observation periods (1988–1996 and 1996–2000) in a large urban hospital-based peritoneal dialysis program. We present an argument in favor of a new approach to the classification of peritonitis.

table i Visceral injury resulting in peritonitis.        

 

Observation period Cases   Deaths

Reference

(years) [n (incidence)] Episodes/year of risk [n (% of reported events)]
Slingeneyer et al 1981 (7)
2.75 27 (10%) 15 (52%)
Spence et al 1985 (8)
2.2 10 (6.2%) 0 (0%)
Prowant et al 1986 (9)
8 — (10.5%)
Tranæus et al 1990 (10)
9 30 (23.2%) 0.17 1 (3%)
Germain et al 1992 (11)
10 12 (7.2%) 4 (33%)
Rotellar et al 1992 (12)
7 12
Tzamaloukas et al 1992 (1)
4 26 11 (42%)
Wellington and Rody 1993 (3)
7 7 4 (57%)
Wakeen et al 1994 (2)
14 15 (3.5%) 6 (40%)
Suh et al 1996 (13)
6 15 (7%) 0.46 2 (13%)
Harwell et al 1997 (5)
8.5 26 (7.3%) 0.11 13 (50%)
Caravaca et al 1998 (14)
6 14 (25%) 2 (14%)
Newman et al 2001 (6) 4 16 (8.4%) 0.10 7 (44%)

 

table ii Anatomic site of visceral injury.                    

 

      Perf Ischemic Bowel Surgical Intra-Ab   Total

Reference

Diver Chol Appen ulcer intestine obstruction procedure abscess Unknown cases

Slingeneyer et al 1981 (7) 9 1 9 1 4 3

9 1 9   1 4 3     27

Spence et al 1985 (8) 4 2 2 1 1

4

2 2   1 1       10
Germain et al 1992 (11) 5 1 1 1 4 5 1   1 1       4 12
Tzamaloukas et al 1992 (1) 4 3 5 7 4 3 4 3     5   7 4 3 26
Rotellar et al 1992 (12) 6 1

6

  1             7
Wellington and Rody 1993 (3) 5 2 5       2         7
Wakeen et al 1994 (2) 8 2 1 3 1 8 2 1   3 1       15
Suh et al 1996 (13) 10 1 3 1

10

1         3   1 15
Harwell et al 1997 (5) 6 3 4 1 10 1 1 6 3 4 1 10   1 1   26
Newman et al 2001 (6) 1 4 4 4 3 1 4   4 4 3       16
Diver = diverticulitis; Chol = cholecystitis; Appen = appendicitis; Perf = perforated; Intra-Ab = intra-abdominal.                    

 

 

Patients and Methods

Retrospective chart reviews were performed during two time periods: 1 January 1988 to 30 June 1996, and 1 July 1996 to 30 June 2000. Data were recorded on all patients who developed peritonitis while undergoing peritoneal dialysis at the University Hospitals of Cleveland (UHC). A more detailed description of methodology, statistical analysis, and outcomes has been described previously (5,6).
The episodes of peritonitis were grouped according to microbiology (Figure 1), but also according to these definitions of etiology (Table III):

  1. Technique failure (any contamination resulting from touch, failure to wear a mask, or technical error)
  2. Recurrent (any discrete episode involving the same organism and occurring within two weeks of completion of a course of antibiotic therapy)
  3. Prior antibiotic (any episode developing after a course of treatment of antibiotics for any reason)
  4. Exit/tunnel (an episode of peritonitis developing in the presence of a catheter infection with the same organism)
  5. Postoperative catheter placement (peritonitis developing after catheter placement, despite treatment with prophylactic antibiotics at the time of surgery)
  6. Diarrhea/urinary tract infection (UTI)/pelvic inflammation (peritonitis that developed contemporaneously with a documented urinary or pelvic infection, or a bout of diarrhea)
  7. Sepsis (peritonitis developing during the course of treatment for sepsis, as documented by blood cultures positive for the same organism found in the peritoneum)
  8. Unknown (no source identified)
  9. Enteric—anatomic diagnosis (episode of peritonitis demonstrated by surgery or histology to be caused by visceral injury)
  10. Presumed enteric (peritonitis due to a suspected, but undocumented, visceral injury)
  11. Surgical contamination (peritonitis developing after surgical contamination of the peritoneal cavity through an operative procedure—for example, colonoscopy with biopsy, vaginal hysterectomy)

figure 1 Microbiology of peritonitis during the two study periods. VRE = vancomycin-resistant enterococcus; CNS = coagulase-negative staphylococcus.

table iii Peritonitis episodes classified by etiology and risk per patient–year of treatment.        

 

1988–1996
1996–2000

 

Episodes
Episodes per
Episodes
Episodes per

Etiology

[n (%)]
patient–year at risk
[n (%)]
patient–year at risk

Technique failure

144 (40.7) 0.42 77 (40.3) 0.34

Recurrent

NC NC 12 (6.3) 0.05

Prior antibiotics

NC NC 7 (3.7) 0.03

Exit/tunnel infection

81 (22.9) 0.24 29 (15.2) 0.13

Postoperative catheter placement

NC NC 8 (4.2) 0.04

Diarrhea/UTI/pelvic

16 (4.5) 0.05 10 (5.2) 0.04

Sepsis

4 (1.1) 0.01 9 (4.7) 0.04

Unknown

50 (14.1) 0.15 2 (1.0) 0.01

Enteric (anatomic diagnosis)a

39 (11.0) 0.11 23 (12.0) 0.10

Presumed enteric

NC NC 7 (3.7) 0.03

Surgical contaminationa

20 (5.6) 0.06 7 (3.7) 0.03

Total

354 (100) 1.03 191 (100) 0.84
a In 1988–1996, this etiology includes postoperative catheter placement.
NC = not coded in 1988–1996 study; UTI = urinary tract infection.
       

 

Results

The peritoneal dialysis unit at UHC served 230 patients between 1988 and 1996 (4133 patient–months of care) and 170 patients between 1996 and 2000 (2738 patient–months). In the first period, 354 episodes of peritonitis were seen; in the second period, the number of episodes was 191. In 1988–1996, 11.3% of 230 patients developed enteric peritonitis due to visceral injury, as compared with 9.4% in 1996–2000 (5,6) (Table I). Figure 1 summarizes the microbiologic characteristics and risk (expressed per patient–year) of all peritonitis episodes. An overall reduction in peritonitis was seen, owing to a specific decrease in infections caused by gram-positive organisms and by Pseudomonas and Xanthomonas. Also, fewer cultures showed “no growth.” In contrast, the risk of infection with gram-negative or enteric organisms remained stable.
Table III summarizes the number and risk of episodes of peritonitis classified by etiologic source of infection. In 1988–1996, the categories “Recurrent infection” and “Prior antibiotics” were not separated from the category “Technique failure.” Also, the 1996–2000 study added two categories not used in 1988–1996: “Presumed enteric” and “Surgical contamination.” Despite the differences in methodology, the risk of anatomically documented visceral injury as a cause of enteric peritonitis remained unchanged between the two time periods (0.11 and 0.10).
Table II summarizes anatomically documented sources of visceral injury (5,6) in comparison with other series. The new category, “Presumed enteric” (Table III), includes 6 of 7 individuals in whom the presentation and clinical course of peritonitis were consistent with a diagnosis of diverticulitis. Aggressive antibiotic therapy resulted in resolution without the need for surgical exploration.
In 1988–1996, 13 deaths were related to enteric peritonitis. In 1996–2000, 7 deaths were related to abdominal catastrophe. Table I compares death rates in other published series.

Discussion and Conclusion

The risk of abdominal catastrophe is comparable to that of exit/tunnel infections in patients treated by peritoneal dialysis (Table III). We believe that, owing to a focus on microbiologic systems of classification, peritonitis resulting from visceral injury has received inadequate attention. The markedly increased morbidity and mortality among patients with peritonitis caused by gram-negative organisms, as compared with gram-positive organisms, is well recognized (15–17). Guidelines from the International Society for Peritoneal Dialysis (ISPD) suggest that a culture positive for multiple gram-negative organisms or anaerobic bacteria should lead to serious consideration of intra-abdominal pathology with evaluation for surgical exploration. However, a recent report of polymicrobial peritonitis demonstrated by culture identified an enteric source in less than 7% of cases. Instead, polymicrobial infection most commonly accompanied exit/tunnel infections, and heralded the need for catheter removal, but not necessarily abdominal exploration (18). Moreover, gram-positive or culture-negative dialysate specimens can accompany anatomically documented abdominal catastrophe. For this reason, we propose that classification of peritonitis by source, and not by organism, may lead to a more focused response to each episode of peritonitis.
The incidence of visceral perforation in patients treated by peritoneal dialysis is 50-fold to 100-fold greater than in patients of comparable age, race, sex, and underlying disease treated by transplantation or hemodialysis (5). The peritoneal dialysis technique may predispose to abdominal catastrophe through alterations in peritoneal morphology, ultrastructure, and immune function, and through the direct effects of high glucose dialysate as an irritant and culture medium. The prophylactic use of antibiotics to prevent peritonitis after surgical techniques that invade the peritoneum (such as vaginal hysterectomy), or in association with bacteremia (after colonoscopy or dental procedures) has found acceptance. Thus, the importance of antibiotic prophylaxis should be reassessed in defined clinical settings that have a high likelihood of progressing to abdominal catastrophe.
Specific clinical scenarios, such as treatment of severe constipation, may result in peritonitis with enteric organisms (19). But a perforated diverticulum may also seal rapidly, causing only a small and self-limited contamination of the peritoneum. We propose that high-dose intravenous and intraperitoneal antibiotic regimens, similar to those used for diverticulitis in patients not on dialysis, should be studied in the setting of visceral injury as a cause of peritonitis.
Recovery without laparotomy has been described by some authors, and may explain apparent differences in the incidence of this disorder between various centers (8). However, a large continuous contamination is unlikely to resolve with antibiotic treatment alone. Therefore, the indications for, and outcomes of, surgical interventions must be carefully categorized and analyzed. Whether surgery is indicated or not, a high index of suspicion for enteric sources will lead to timely surgical consultation, as well as early initiation of antibiotic coverage for enteric and anaerobic organisms, which may make surgery unnecessary.

Acknowledgments

Supported in part by DK-45619 from the National Institutes of Health (MW), and by the Leonard C. Rosenberg Renal Research Foundation of the Center for Dialysis Care, Cleveland. The authors thank Dr. Lavinia Negrea, Ms. Angela Priester, and Ms. Jean Gordon, and they remember the assistance of Ms. Kathleen Ferrara.

References

  1. Tzamaloukas A, Obermiller L, Murata G, et al. CAPD peritonitis due to intra-abdominal pathology [Abstract]. Perit Dial Int 1992; 12(Suppl 2):S59.
  2. Wakeen M, Zimmerman S, Bidwell D. Viscus perforation in peritoneal dialysis patients: diagnosis and outcome. Perit Dial Int 1994; 14:371–7.
  3. Wellington J, Rody K. Acute abdominal emergencies in patients on long-term ambulatory peritoneal dialysis. Can J Surg 1993; 36:522–4.
  4. Bustos E, Rotellar C, Mazzoni M, et al. Clinical aspects of bowel perforation in patients undergoing continuous ambulatory peritoneal dialysis. Semin Dial 1994; 7:355–9.
  5. Harwell C, Newman L, Cacho C, et al. Abdominal catastrophe: visceral injury as a cause of peritonitis in patients treated by peritoneal dialysis. Perit Dial Int 1997; 17:586–94.
  6. Newman L, Cacho C, Schulak J, et al. Abdominal catastrophe revisited: can the outcome of enteric peritoneal contamination be improved? Perit Dial Int 2001; [In press].
  7. Slingeneyer A, Mion C, Beraud JJ, Oules R, Branger B, Balmes M. Peritonitis, a frequently lethal complication of intermittent and continuous ambulatory peritoneal dialysis. Proc Eur Dial Transplant Assoc 1981; 18:212–21.
  8. Spence P, Mathews R, Khanna R, et al. Indications for operation when peritonitis occurs in patients on chronic ambulatory peritoneal dialysis. Surg Gynecol Obstet 1985; 161:450–2.
  9. Prowant B, Nolph K, Ryan L, Twardowski Z, Khanna R. Peritonitis in continuous ambulatory peritoneal dialysis: analysis of an 8-year experience. Nephron 1986; 43:105–9.
  10. Tranæus A, Heimbürger O, Granqvist S. Diverticular disease of the colon: a risk factor for peritonitis in continuous peritoneal dialysis. Nephrol Dial Transplant 1990; 5:141–7.
  11. Germain M, Harlow P, Mulhern J, et al. Low protein catabolic rate and serum albumin correlate with increased mortality and abdominal complications in peritoneal dialysis patients. Adv Perit Dial 1992; 8:113–15.
  12. Rotellar C, Sivarajan S, Mazzoni M, et al. Bowel perforation in CAPD patients. Perit Dial Int 1992; 12:396–8.
  13. Suh H, Wadhwa NK, Cabralda T, Sorrento J. Endogenous peritonitis and related outcome in peritoneal dialysis patients. Adv Perit Dial 1996; 12:192–5.
  14. Caravaca F, Ruiz–Calero R, Dominguez C. Risk factors for developing peritonitis caused by micro-organisms of enteral origin in peritoneal dialysis patients. Perit Dialysis Int 1998; 18:41–5.
  15. Fried L, Bernardini J, Johnston J, et al. Peritonitis influences mortality in peritoneal dialysis patients. J Am Soc Nephrol 1996; 7:2176–82.
  16. Bunke C, Brier M, Golper T. Outcomes of single organism peritonitis in peritoneal dialysis: gram negatives versus gram positives in the Network 9 Peritonitis Study. Kidney Int 1997; 52:524–9.
  17. Troidle L, Gorban–Brennan N, Kliger A, Finkelstein F. Differing outcomes of gram-positive and gram-negative peritonitis. Am J Kidney Dis 1998; 32: 623–8.
  18. Kim G, Korbet S. Polymicrobial peritonitis in continuous ambulatory peritoneal dialysis patients. Am J Kidney Dis 2000; 36:1000–8.
  19. Singharetnam W, Holley J. Acute treatment of constipation may lead to transmural migration of bacteria resulting in gram-negative, polymicrobial, or fungal peritonitis. Perit Dial Int 1996; 16:423–44.

Corresponding author:

Miriam F. Weiss, md, Division of Nephrology, University Hospitals of Cleveland, 11100 Euclid Avenue, Cleveland, Ohio 44106 U.S.A.