- Email: [email protected]
Outcome of colectomy for Clostridium difficile colitis: a plea for early surgical management
The American Journal of Surgery (2008) 196, 384 –388
Clinical Surgery-American
Outcome of colectomy for Clostridium difficile colitis: a plea for early surgical management Jason F. Hall, M.D.*, David Berger, M.D. Department of Surgery, Massachusetts General Hospital, WANG–ACC 455, Boston, MA 02115, USA KEYWORDS: Clostridium difficile; Clostridium difficile colitis; Toxic megacolon; Fulminant colitis; Subtotal colectomy
Abstract BACKGROUND: Fulminant Clostridium difficile colitis is a common nosocomial infection that occurs with increasing frequency. METHODS: A total of 3,237 consecutive cases of C difficile cytotoxin–positive stool samples from 1998 to 2006 were reviewed. Commonly referenced indicators for surgical intervention were gathered on the day of surgery. The preoperative characteristics of patients surviving subtotal colectomy were compared with those who did not survive. RESULTS: Thirty-six patients underwent colectomy. Twenty-three patients (64%) were discharged from the hospital alive. Preoperative intubation and vasopressor requirement were risk factors for in-hospital mortality (odds ratio [OR], 7.15; 95% confidence interval [95% CI], 1.28 –39.8 and OR, 6.0; CI, 1.08 –33, respectively). Patients who had a recent surgical procedure had a lower in-hospital mortality rate (OR, .11; 95% CI, .02–.52). CONCLUSIONS: Fulminant C difficile colitis is associated with a high mortality rate. Development of a vasopressor requirement or need for intubation are ominous signs and should lead to rapid surgical intervention. © 2008 Elsevier Inc. All rights reserved.
Clostridium difficile is a spore-forming, gram-positive bacteria that produces a broad spectrum of clinical presentations ranging from mild diarrhea to fulminant colitis associated with multisystem organ failure and shock. Recent use of antibiotics is the most common risk factor for the development of C difficile colitis. As the use of antibiotics has become ubiquitous, the incidence of C difficile colitis and subsequently fulminant C difficile colitis also have increased. In the overwhelming majority of cases, C difficile colitis can be treated with oral antibiotics such as metronidazole or vancomycin. In addition, administration of probiotics (Lactobacillus, Saccharomyces boulardii) and refeeding of donor stool also have been advocated.1–3 Occasionally, patients fail antibiotic regimens and progress * Corresponding author. Tel.: ⫹1-617-283-8507. E-mail address: [email protected] Manuscript received June 10, 2007; revised manuscript November 20, 2007
0002-9610/$ - see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.amjsurg.2007.11.017
to fulminant colitis. The onset of fulminant colitis often requires subtotal colectomy and ileostomy, which has a high mortality rate ranging from 35% to 57%.4,5 The surgeon’s goal is to determine the proper timing of surgery because excessive delay in surgical management likely leads to increased morbidity and mortality. Conversely, performing surgery too early in the course of this disease commits the patient to loss of a significant portion of their colon and a temporary or permanent ileostomy. The objective of this study was to analyze retrospectively the outcome of surgical treatment of C difficile colitis in our hospital and to determine the preoperative factors that should trigger surgical intervention and reduce in-hospital mortality rates.
Methods All adult patients who had a positive C difficile cytotoxin assay were identified for the 8-year period from January
J.F. Hall and D. Berger Table 1
Colectomy for C difficile colitis
385
Comparison of preoperative characteristics of patients undergoing subtotal colectomy Survivors
Age APACHE II Heart rate White blood cell count Albumin level Lactate level Creatinine level pH Metronidazole treatment
67 21.7 115 33 2.0 2.7 2.3 7.3 151
⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
Nonsurvivors
13 y 10.9 20 bpm 19 th/cm2 .6 g/dL 1.8 mmol/L 1.4 mg/dL .11 231 h
1998 to January 2006 in a single institution. Patients suspected of having C difficile infection were tested for cytotoxin A and B by using a commercially available immunoassay (Product # 712050; Meridian Bioscience, Inc.). A chart review was performed to identify patients who underwent a colectomy secondary to fulminant C difficile colitis. Fulminant C difficile colitis was defined as any patient with a severe systemic inflammatory response (fever, hypotension, tachypnea, leukocytosis, and/or requirement for volume resuscitation) in the context of a clinically documented Clostridium-associated colitis that was severe enough that death seemed likely without urgent colectomy.4 In addition, their illness had to be severe enough to merit surgical consultation. Patients who died before surgical consultation or who did not receive surgery were excluded from analysis. There was no standard protocol for proceeding to surgery, however, hemodynamic instability, peritonitis, and failure to respond to medical management were common indications for surgical intervention. All patients who underwent a colectomy for C difficile colitis had the diagnosis confirmed by the features consistent with infectious colitis on gross and microscopic pathologic examinations. Chart review was performed on those patients requiring surgery. Data points considered included age, recent surgical procedure, APACHE II score6 24 hours before surgery, heart rate, white blood cell count, serum albumin level, serum creatinine level, serum lactate level, serum pH level, intubation status, and vasopressor requirement on the day of surgery. Patients’ use of immunosuppressive medications including glucocorticoids, purine synthesis inhibitors, cyclosporine, tacrolimus, and recent treatment with chemotherapeutic agents were noted. Patients with solid-organ transplants or human immunodeficiency virus also were noted. Significant postoperative complications such as ventilator dependence (ventilator requirement ⬎ 14 d), pneumonia, renal failure, deep vein thrombosis, and abdominal fluid collections requiring drainage also were recorded. A 2-tailed Student t test determined the statistical significance of variables between 2 groups. Odds ratios (OR) were calculated according to the method described by Bland and Altman.7
69 24.5 105 37 1.8 3.6 2.1 7.25 81
⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
13 y 9.9 17 20 th/cm2 .5 g/dL 2.4 mmol/L 1.3 mg/dL 9.9 75 h
P value .43 .67 .87 .61 .76 .76 .51 .67 .77
Results Patient characteristics During the 8-year data period there were 3,237 cases of C difficile cytotoxin–positive assays. Thirty-six patients underwent a colectomy. The clinical characteristics of patients undergoing colectomy are presented in Table 1. Twentyeight percent of patients had a recent history of medically treated C difficile colitis. A history of recent medically treated C difficile colitis was not associated with survival (OR, 1.46; 95% confidence interval [95% CI], .30 –7.0). Age at presentation was not a statistically significant predictor of survival. In our cohort, age older than 80 years was not a significant predictor of mortality. Interestingly, 12 of 23 patients surviving total abdominal colectomy were age 80 or older. Preoperative laboratory factors including APACHE II scores, white blood cell count, albumin level, lactate level, creatinine level, pH level, or immunosuppression also were not significant predictors of survival. Patients who were intubated before surgery had an increased inhospital mortality rate (OR, 7.2; 95% CI, 1.2–39). Patients on vasopressors preoperatively had an increased in-hospital mortality rate (OR, 6; CI, 1.08 –33). Seventy-four percent of patients surviving total abdominal colectomy underwent a recent surgical procedure (Table 2). Patients who had a recent surgical procedure (within 60 days) were more likely to survive than their counterparts. Categories of recent procedures in both groups are outlined in Table 3. All patients in our cohort received antibiotic therapy within 60 days of the onset of C difficile colitis. The commonly prescribed antibiotics are outlined in Table 4. There was no difference in the percentage of survivors versus nonsurvivors who received preoperative metronidazole therapy (86% vs 92%; P ⬎ .05). There was also no statistically significant difference in the duration of preoperative metronidazole treatment in the survivor and nonsurvivor groups (151 ⫾ 231 vs 81 ⫾ 75 h; P ⫽ .77).
Diagnostic procedures Computed tomography (CT) was used commonly to aid in the diagnosis of C difficile colitis. Thirty-two of 36
386 Table 2
The American Journal of Surgery, Vol 196, No 3, September 2008 Comparison of preoperative clinical characteristics of patients undergoing subtotal colectomy
Intubation No intubation Vasopressors No vasopressors Immunosuppressed Immunocompetent Recent surgery No recent surgery
Survivors (23)
Nonsurvivors (13)
P value
OR (95% CI)
43% 57% 48% 52% 43% 57% 73% 27%
84% 16% 84% 16% 23% 77% 23% 77%
.03
7.2 (1.2–39.8)
.04
6.0 (1.08–33)
.29
.4 (.08–1.8)
.005
.1 (.02–.52)
(10) (13) (11) (12) (10) (13) (17) (6)
patients (94%) undergoing total abdominal colectomy for C difficile colitis were interrogated with CT. CT predicted the findings at surgery in 94% of cases. Twenty-one patients (66%) had scans consistent with pancolitis. Nineteen percent of patients presented with only right-sided colitis and 15% presented with isolated left-sided colitis. Endoscopy seldom was used to confirm diagnosis. Three patients underwent rigid sigmoidoscopy or colonoscopy. All 3 endoscopies confirmed the presence of pseudomembranes. One procedure was associated with perforation.
Surgical management Thirty-four patients (94%) underwent subtotal colectomy and end ileostomy. A right colectomy and ileotransverse anastomosis was performed in a critically ill patient with suspected fulminant colitis. Findings at surgery were suggestive of a cecal bascule. Subsequent analysis of the specimen revealed C difficile colitis. A left colectomy with end colostomy and Hartmann’s pouch was performed in a patient with an acute abdomen and positive C difficile cytotoxin assay after a thoracoabdominal aneurysm repair. Findings at surgery revealed a perforated sigmoid colon. This patient’s subsequent microscopic and gross pathology revealed C difficile colitis. Both patients survived to hospital discharge. For patients who survived to hospital discharge, the average length of stay was 33 ⫾ 28 days. The average intensive care unit (ICU) length of stay was 12 ⫾ 14 days. Seventy-four percent of these patients had at least 1 com-
Table 3 Classification of surgical interventions in patients less than 60 days before subtotal colectomy
General surgery Orthopedic surgery Vascular surgery Cardiac surgery Lung transplant Neurosurgery Gynecology Urology
Survivors
Nonsurvivors
4 4 2 2 2 1 1 1
1 1 1
(11) (2) (11) (2) (3) (10) (3) (10)
plication. Forty-three percent of patients had at least 2 complications. The most common complication was ventilator dependence (34%). Ventilator dependence was defined as the need for ventilatory support for greater than 14 postoperative days. Acute renal failure (26%), fluid collection (17%), pneumonia (13%), and wound dehiscence (13%) were notable complications. Sixty-one percent of patients who died before hospital discharge had at least one complication. Ventilator dependence (31%) and pneumonia (31%) were the most common complications in this group. Hospital length of stay was 20 ⫾ 31 days. Most patients (92%) died in the ICU. The ICU length of stay was 17 ⫾ 25 days. There were no specific complications that were associated with mortality.
Comments With the increase in the use and misuse of antibiotic therapy, the incidence and severity of clinically evident C difficile colitis is increasing.8 C difficile colitis commonly presents with mild to moderate diarrhea and usually is treated easily with oral metronidazole or vancomycin. A small cohort of patients, however, will develop a fulminant colitis leading to significant morbidity and mortality. This colitis often is unresponsive to the best medical therapy. Fulminant colitis should be treated with a subtotal colectomy and end ileostomy in most cases.
Table 4 Classes of antibiotics used in patients less than 60 days before subtotal colectomy
Cephalosporin Fluoroquinolone Clindamycin Vancomycin Penicillin Multiple antibiotics Aminoglycosides Flagyl
Survivors (23)
Nonsurvivors (13)
12 4 3 4 0 3 1 1
7 2 4 1 2 2 0 0
Excludes treatment for C difficile colitis.
J.F. Hall and D. Berger
Colectomy for C difficile colitis
The diagnosis of C difficile colitis often is suspected after the onset of diarrhea. The diagnosis is confirmed through the use of a cytotoxin A or B immunoassay. This test has a high sensitivity and specificity, although test results may not be available for more than 24 hours.9 In recent years, the hypervirulent strain NAP1/027 has emerged.10 In some reports this strain has increased the rate of emergency colectomies in those institutions. Our hospital does not currently test for this hypervirulent strain. The presence of pseudomembranes on the colonic mucosa is diagnostic of C difficile infection. Several investigators have suggested that flexible sigmoidoscopy should be used in all hospitalized patients suspected of C difficile colitis.11 Patients in our study group tended not to receive preoperative endoscopy because most of their diagnoses were confirmed with an immunotoxin assay and imaging. CT is another useful adjunct in the diagnosis of C difficile colitis. Common CT findings include pericolonic stranding, colonic thickening, and ascites.12 In our cohort, CT predicted the surgical findings in 94% of cases. CT evidence of colitis in the presence of diarrhea and abdominal tenderness should alert the clinician to potentially progressive fulminant colitis. Because up to 12% of C difficile toxin assays can be false negatives, CT evidence of pancolitis with the appropriate clinical history may be an indication for surgery.4 Fulminant C difficile colitis tends to affect elderly populations. The average age of patients in our cohort was 69 ⫾ 12 years. Dallal et al4 noted that there were no survivors of fulminant C difficile colitis older than 80 years of age. This result has been misinterpreted by surgeons to suggest that elderly patients should not be candidates for intervention. Others have suggested that patients with severe C difficile colitis and numerous medical comorbidities should not receive resection.13 In our cohort, however, 52% of patients surviving a subtotal colectomy were older than age 80. In addition, the average APACHE II score in both survivor and nonsurvivor groups was greater than 20, indicating that many of these patients had significant organ dysfunction. Neither age older than 80 or APACHE II score predisposed these patients to mortality. These factors suggest that early intervention is the key to survival. Although many of the previously indicated clinical signs are not predictive of survival, it is clear that a ventilator or vasopressor requirement are ominous signs. When these are present in a patient with fulminant colitis an urgent surgery is required. Previous work has shown that a vasopressor requirement despite adequate volume resuscitation is an indication for colectomy in patients with C difficile colitis.11,12 The data presented in this report show that a preoperative vasopressor requirement in the 24-hour period before surgery is associated with a markedly increased risk of death after subtotal colectomy for C difficile colitis. This suggests that in this group of patients hypotension is a relatively late sign. Once vasopressors are required to maintain adequate perfusion it is likely that the patient’s condi-
387 tion is sufficiently advanced such that a favorable outcome is significantly less likely. Similarly, patients requiring mechanical ventilation preoperatively had an increased risk of mortality after subtotal colectomy. Onset of respiratory failure is another ominous sign that likely is a harbinger of multisystem organ failure. Once the patient has progressed past this point, it is also significantly less likely that an intervention will be of clinical benefit. Several sources have noted better outcomes after fulminant C difficile colitis in patients who had recent transplantation.4 Immunosuppressed patients in our cohort did not derive any survival advantage. However, we did note that a recent surgical procedure was protective after subtotal colectomy (Table 2). We suspect this effect is owing to the presence of these patients on a surgical service. Surgeons may be more aggressive in the treatment of this disease, leading to surgery before signs of organ failure, and consequent death. We believe that this underscores the necessity for early intervention and have attributed the difference to delay in surgical consultation and subsequent treatment. We understand that this is an assumption that can be better delineated only by concrete data. This study was limited by its retrospective nature and its small sample size, however, it is impossible to prospectively study this population. Although C difficile colitis is common, it is rare for it to progress to fulminant colitis. Furthermore, it would be impossible to randomize patients to either delayed or nonsurgical treatment because this would have serious ethical ramifications. The key to treatment of fulminant C difficile colitis is early surgical intervention. Our institution advocates this approach and our in-hospital mortality rate of 36% is significantly lower than those of other groups reporting mortality rates ranging from 41% to 80%.4,5,12,14,15 It is possible that some patients in this series may have survived hospitalization without surgery; however, the data presented in this article show that even in a series of comparably low mortality rates, a vasopressor or ventilatory requirement significantly increased mortality rates.
Conclusions C difficile colitis is common. Our study was a retrospective analysis of our institution’s experience with C difficile colitis and represents the second largest surgical series in the literature. Surgical treatment of this disease is associated with a high mortality rate if not addressed in a timely fashion. Commonly quoted preoperative indicators are not useful in determining the timing of surgeries. Failure to respond to appropriate medical therapy with associated systemic inflammatory response or peritonitis are indications for surgery. The development of a vasopressor requirement or need for intubation are ominous signs that also are indications for urgent surgical intervention. Early surgical intervention is the key to survival.
388
The American Journal of Surgery, Vol 196, No 3, September 2008
References 1. Fekety R, Shah AB. Diagnosis and treatment of Clostridium difficile colitis. JAMA 1993;269:71–5. 2. Bartlett JG. Clostridium difficile: history of its role as and enteric pathogen and the current state of knowledge about the organism. Clin Infect Dis 1994;18(Suppl 4):S265–72. 3. Aas J, Gessert CE, Bakken JS. Recurrent Clostridium difficile colitis: case series involving 18 patients treated with donor stool via nasogastric tube. Clin Infect Dis 2003;36:580. 4. Dallal RM, Harbrecht BG, Boujoukas AJ, et al. Fulminant Clostridium difficile: an underappreciated and increasing cause of death and complications. Ann Surg 2001;235:363–72. 5. Synnott K, Mealy K, Merry C, et al. Timing of surgery for fulminating pseudomembranous colitis. Br J Surg 1998;85:229 –31. 6. Knaus WA, Draper EA, Wagner DP, et al. APACHE II: a severity of disease classification system. Crit Care Med 1985;13:818 –29. 7. Bland JM, Altman DG. The odds ratio. BMJ 2000;320:1468. 8. Medich DS, Lee KK, Simmons RL, et al. Laparotomy for fulminant pseudomembranous colitis. Arch Surg 1992;127:847–53.
9. Marts BC, Longo WE, Vernava AM, et al. Patterns and prognosis of Clostridium difficile colitis. Dis Colon Rectum 1994;37:837– 45. 10. Loo VG, Poirier L, Miller M, et al. A predominantly clonal multiinstitutional outbreak of Clostridium difficile associated diarrhea with high associated morbidity and mortality. N Engl J Med 2005;353: 2442–9. 11. Johal SS, Hammond J, Solomon K, et al. Clostridium difficile associated diarrhea in hospitalized patients; onset in the community and role of flexible sigmoidoscopy. Gut 2004;53:673– 4. 12. Longo WE, Mazuski JE, Virgo KS, et al. Outcome after colectomy for Clostridium difficile colitis. Dis Colon Rectum 2003;47:1620. 13. Rubin M, Bodenstein LE, Kent KC. Severe Clostridium difficile colitis. Dis Colon Rectum 1995;38:350. 14. Klipfel AA, Schein M, Fahoum B, et al. Acute abdomen and Clostridium difficile colitis: still a lethal combination. Dig Surg 2000;17: 160 –3. 15. Grundfest-Borniatowski S, Quader M, Alexander F, et al. Clostridium difficile colitis in the critically ill. Dis Colon Rectum 1995; 39:619 –23.
Clinical Surgery-American
Outcome of colectomy for Clostridium difficile colitis: a plea for early surgical management Jason F. Hall, M.D.*, David Berger, M.D. Department of Surgery, Massachusetts General Hospital, WANG–ACC 455, Boston, MA 02115, USA KEYWORDS: Clostridium difficile; Clostridium difficile colitis; Toxic megacolon; Fulminant colitis; Subtotal colectomy
Abstract BACKGROUND: Fulminant Clostridium difficile colitis is a common nosocomial infection that occurs with increasing frequency. METHODS: A total of 3,237 consecutive cases of C difficile cytotoxin–positive stool samples from 1998 to 2006 were reviewed. Commonly referenced indicators for surgical intervention were gathered on the day of surgery. The preoperative characteristics of patients surviving subtotal colectomy were compared with those who did not survive. RESULTS: Thirty-six patients underwent colectomy. Twenty-three patients (64%) were discharged from the hospital alive. Preoperative intubation and vasopressor requirement were risk factors for in-hospital mortality (odds ratio [OR], 7.15; 95% confidence interval [95% CI], 1.28 –39.8 and OR, 6.0; CI, 1.08 –33, respectively). Patients who had a recent surgical procedure had a lower in-hospital mortality rate (OR, .11; 95% CI, .02–.52). CONCLUSIONS: Fulminant C difficile colitis is associated with a high mortality rate. Development of a vasopressor requirement or need for intubation are ominous signs and should lead to rapid surgical intervention. © 2008 Elsevier Inc. All rights reserved.
Clostridium difficile is a spore-forming, gram-positive bacteria that produces a broad spectrum of clinical presentations ranging from mild diarrhea to fulminant colitis associated with multisystem organ failure and shock. Recent use of antibiotics is the most common risk factor for the development of C difficile colitis. As the use of antibiotics has become ubiquitous, the incidence of C difficile colitis and subsequently fulminant C difficile colitis also have increased. In the overwhelming majority of cases, C difficile colitis can be treated with oral antibiotics such as metronidazole or vancomycin. In addition, administration of probiotics (Lactobacillus, Saccharomyces boulardii) and refeeding of donor stool also have been advocated.1–3 Occasionally, patients fail antibiotic regimens and progress * Corresponding author. Tel.: ⫹1-617-283-8507. E-mail address: [email protected] Manuscript received June 10, 2007; revised manuscript November 20, 2007
0002-9610/$ - see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.amjsurg.2007.11.017
to fulminant colitis. The onset of fulminant colitis often requires subtotal colectomy and ileostomy, which has a high mortality rate ranging from 35% to 57%.4,5 The surgeon’s goal is to determine the proper timing of surgery because excessive delay in surgical management likely leads to increased morbidity and mortality. Conversely, performing surgery too early in the course of this disease commits the patient to loss of a significant portion of their colon and a temporary or permanent ileostomy. The objective of this study was to analyze retrospectively the outcome of surgical treatment of C difficile colitis in our hospital and to determine the preoperative factors that should trigger surgical intervention and reduce in-hospital mortality rates.
Methods All adult patients who had a positive C difficile cytotoxin assay were identified for the 8-year period from January
J.F. Hall and D. Berger Table 1
Colectomy for C difficile colitis
385
Comparison of preoperative characteristics of patients undergoing subtotal colectomy Survivors
Age APACHE II Heart rate White blood cell count Albumin level Lactate level Creatinine level pH Metronidazole treatment
67 21.7 115 33 2.0 2.7 2.3 7.3 151
⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
Nonsurvivors
13 y 10.9 20 bpm 19 th/cm2 .6 g/dL 1.8 mmol/L 1.4 mg/dL .11 231 h
1998 to January 2006 in a single institution. Patients suspected of having C difficile infection were tested for cytotoxin A and B by using a commercially available immunoassay (Product # 712050; Meridian Bioscience, Inc.). A chart review was performed to identify patients who underwent a colectomy secondary to fulminant C difficile colitis. Fulminant C difficile colitis was defined as any patient with a severe systemic inflammatory response (fever, hypotension, tachypnea, leukocytosis, and/or requirement for volume resuscitation) in the context of a clinically documented Clostridium-associated colitis that was severe enough that death seemed likely without urgent colectomy.4 In addition, their illness had to be severe enough to merit surgical consultation. Patients who died before surgical consultation or who did not receive surgery were excluded from analysis. There was no standard protocol for proceeding to surgery, however, hemodynamic instability, peritonitis, and failure to respond to medical management were common indications for surgical intervention. All patients who underwent a colectomy for C difficile colitis had the diagnosis confirmed by the features consistent with infectious colitis on gross and microscopic pathologic examinations. Chart review was performed on those patients requiring surgery. Data points considered included age, recent surgical procedure, APACHE II score6 24 hours before surgery, heart rate, white blood cell count, serum albumin level, serum creatinine level, serum lactate level, serum pH level, intubation status, and vasopressor requirement on the day of surgery. Patients’ use of immunosuppressive medications including glucocorticoids, purine synthesis inhibitors, cyclosporine, tacrolimus, and recent treatment with chemotherapeutic agents were noted. Patients with solid-organ transplants or human immunodeficiency virus also were noted. Significant postoperative complications such as ventilator dependence (ventilator requirement ⬎ 14 d), pneumonia, renal failure, deep vein thrombosis, and abdominal fluid collections requiring drainage also were recorded. A 2-tailed Student t test determined the statistical significance of variables between 2 groups. Odds ratios (OR) were calculated according to the method described by Bland and Altman.7
69 24.5 105 37 1.8 3.6 2.1 7.25 81
⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
13 y 9.9 17 20 th/cm2 .5 g/dL 2.4 mmol/L 1.3 mg/dL 9.9 75 h
P value .43 .67 .87 .61 .76 .76 .51 .67 .77
Results Patient characteristics During the 8-year data period there were 3,237 cases of C difficile cytotoxin–positive assays. Thirty-six patients underwent a colectomy. The clinical characteristics of patients undergoing colectomy are presented in Table 1. Twentyeight percent of patients had a recent history of medically treated C difficile colitis. A history of recent medically treated C difficile colitis was not associated with survival (OR, 1.46; 95% confidence interval [95% CI], .30 –7.0). Age at presentation was not a statistically significant predictor of survival. In our cohort, age older than 80 years was not a significant predictor of mortality. Interestingly, 12 of 23 patients surviving total abdominal colectomy were age 80 or older. Preoperative laboratory factors including APACHE II scores, white blood cell count, albumin level, lactate level, creatinine level, pH level, or immunosuppression also were not significant predictors of survival. Patients who were intubated before surgery had an increased inhospital mortality rate (OR, 7.2; 95% CI, 1.2–39). Patients on vasopressors preoperatively had an increased in-hospital mortality rate (OR, 6; CI, 1.08 –33). Seventy-four percent of patients surviving total abdominal colectomy underwent a recent surgical procedure (Table 2). Patients who had a recent surgical procedure (within 60 days) were more likely to survive than their counterparts. Categories of recent procedures in both groups are outlined in Table 3. All patients in our cohort received antibiotic therapy within 60 days of the onset of C difficile colitis. The commonly prescribed antibiotics are outlined in Table 4. There was no difference in the percentage of survivors versus nonsurvivors who received preoperative metronidazole therapy (86% vs 92%; P ⬎ .05). There was also no statistically significant difference in the duration of preoperative metronidazole treatment in the survivor and nonsurvivor groups (151 ⫾ 231 vs 81 ⫾ 75 h; P ⫽ .77).
Diagnostic procedures Computed tomography (CT) was used commonly to aid in the diagnosis of C difficile colitis. Thirty-two of 36
386 Table 2
The American Journal of Surgery, Vol 196, No 3, September 2008 Comparison of preoperative clinical characteristics of patients undergoing subtotal colectomy
Intubation No intubation Vasopressors No vasopressors Immunosuppressed Immunocompetent Recent surgery No recent surgery
Survivors (23)
Nonsurvivors (13)
P value
OR (95% CI)
43% 57% 48% 52% 43% 57% 73% 27%
84% 16% 84% 16% 23% 77% 23% 77%
.03
7.2 (1.2–39.8)
.04
6.0 (1.08–33)
.29
.4 (.08–1.8)
.005
.1 (.02–.52)
(10) (13) (11) (12) (10) (13) (17) (6)
patients (94%) undergoing total abdominal colectomy for C difficile colitis were interrogated with CT. CT predicted the findings at surgery in 94% of cases. Twenty-one patients (66%) had scans consistent with pancolitis. Nineteen percent of patients presented with only right-sided colitis and 15% presented with isolated left-sided colitis. Endoscopy seldom was used to confirm diagnosis. Three patients underwent rigid sigmoidoscopy or colonoscopy. All 3 endoscopies confirmed the presence of pseudomembranes. One procedure was associated with perforation.
Surgical management Thirty-four patients (94%) underwent subtotal colectomy and end ileostomy. A right colectomy and ileotransverse anastomosis was performed in a critically ill patient with suspected fulminant colitis. Findings at surgery were suggestive of a cecal bascule. Subsequent analysis of the specimen revealed C difficile colitis. A left colectomy with end colostomy and Hartmann’s pouch was performed in a patient with an acute abdomen and positive C difficile cytotoxin assay after a thoracoabdominal aneurysm repair. Findings at surgery revealed a perforated sigmoid colon. This patient’s subsequent microscopic and gross pathology revealed C difficile colitis. Both patients survived to hospital discharge. For patients who survived to hospital discharge, the average length of stay was 33 ⫾ 28 days. The average intensive care unit (ICU) length of stay was 12 ⫾ 14 days. Seventy-four percent of these patients had at least 1 com-
Table 3 Classification of surgical interventions in patients less than 60 days before subtotal colectomy
General surgery Orthopedic surgery Vascular surgery Cardiac surgery Lung transplant Neurosurgery Gynecology Urology
Survivors
Nonsurvivors
4 4 2 2 2 1 1 1
1 1 1
(11) (2) (11) (2) (3) (10) (3) (10)
plication. Forty-three percent of patients had at least 2 complications. The most common complication was ventilator dependence (34%). Ventilator dependence was defined as the need for ventilatory support for greater than 14 postoperative days. Acute renal failure (26%), fluid collection (17%), pneumonia (13%), and wound dehiscence (13%) were notable complications. Sixty-one percent of patients who died before hospital discharge had at least one complication. Ventilator dependence (31%) and pneumonia (31%) were the most common complications in this group. Hospital length of stay was 20 ⫾ 31 days. Most patients (92%) died in the ICU. The ICU length of stay was 17 ⫾ 25 days. There were no specific complications that were associated with mortality.
Comments With the increase in the use and misuse of antibiotic therapy, the incidence and severity of clinically evident C difficile colitis is increasing.8 C difficile colitis commonly presents with mild to moderate diarrhea and usually is treated easily with oral metronidazole or vancomycin. A small cohort of patients, however, will develop a fulminant colitis leading to significant morbidity and mortality. This colitis often is unresponsive to the best medical therapy. Fulminant colitis should be treated with a subtotal colectomy and end ileostomy in most cases.
Table 4 Classes of antibiotics used in patients less than 60 days before subtotal colectomy
Cephalosporin Fluoroquinolone Clindamycin Vancomycin Penicillin Multiple antibiotics Aminoglycosides Flagyl
Survivors (23)
Nonsurvivors (13)
12 4 3 4 0 3 1 1
7 2 4 1 2 2 0 0
Excludes treatment for C difficile colitis.
J.F. Hall and D. Berger
Colectomy for C difficile colitis
The diagnosis of C difficile colitis often is suspected after the onset of diarrhea. The diagnosis is confirmed through the use of a cytotoxin A or B immunoassay. This test has a high sensitivity and specificity, although test results may not be available for more than 24 hours.9 In recent years, the hypervirulent strain NAP1/027 has emerged.10 In some reports this strain has increased the rate of emergency colectomies in those institutions. Our hospital does not currently test for this hypervirulent strain. The presence of pseudomembranes on the colonic mucosa is diagnostic of C difficile infection. Several investigators have suggested that flexible sigmoidoscopy should be used in all hospitalized patients suspected of C difficile colitis.11 Patients in our study group tended not to receive preoperative endoscopy because most of their diagnoses were confirmed with an immunotoxin assay and imaging. CT is another useful adjunct in the diagnosis of C difficile colitis. Common CT findings include pericolonic stranding, colonic thickening, and ascites.12 In our cohort, CT predicted the surgical findings in 94% of cases. CT evidence of colitis in the presence of diarrhea and abdominal tenderness should alert the clinician to potentially progressive fulminant colitis. Because up to 12% of C difficile toxin assays can be false negatives, CT evidence of pancolitis with the appropriate clinical history may be an indication for surgery.4 Fulminant C difficile colitis tends to affect elderly populations. The average age of patients in our cohort was 69 ⫾ 12 years. Dallal et al4 noted that there were no survivors of fulminant C difficile colitis older than 80 years of age. This result has been misinterpreted by surgeons to suggest that elderly patients should not be candidates for intervention. Others have suggested that patients with severe C difficile colitis and numerous medical comorbidities should not receive resection.13 In our cohort, however, 52% of patients surviving a subtotal colectomy were older than age 80. In addition, the average APACHE II score in both survivor and nonsurvivor groups was greater than 20, indicating that many of these patients had significant organ dysfunction. Neither age older than 80 or APACHE II score predisposed these patients to mortality. These factors suggest that early intervention is the key to survival. Although many of the previously indicated clinical signs are not predictive of survival, it is clear that a ventilator or vasopressor requirement are ominous signs. When these are present in a patient with fulminant colitis an urgent surgery is required. Previous work has shown that a vasopressor requirement despite adequate volume resuscitation is an indication for colectomy in patients with C difficile colitis.11,12 The data presented in this report show that a preoperative vasopressor requirement in the 24-hour period before surgery is associated with a markedly increased risk of death after subtotal colectomy for C difficile colitis. This suggests that in this group of patients hypotension is a relatively late sign. Once vasopressors are required to maintain adequate perfusion it is likely that the patient’s condi-
387 tion is sufficiently advanced such that a favorable outcome is significantly less likely. Similarly, patients requiring mechanical ventilation preoperatively had an increased risk of mortality after subtotal colectomy. Onset of respiratory failure is another ominous sign that likely is a harbinger of multisystem organ failure. Once the patient has progressed past this point, it is also significantly less likely that an intervention will be of clinical benefit. Several sources have noted better outcomes after fulminant C difficile colitis in patients who had recent transplantation.4 Immunosuppressed patients in our cohort did not derive any survival advantage. However, we did note that a recent surgical procedure was protective after subtotal colectomy (Table 2). We suspect this effect is owing to the presence of these patients on a surgical service. Surgeons may be more aggressive in the treatment of this disease, leading to surgery before signs of organ failure, and consequent death. We believe that this underscores the necessity for early intervention and have attributed the difference to delay in surgical consultation and subsequent treatment. We understand that this is an assumption that can be better delineated only by concrete data. This study was limited by its retrospective nature and its small sample size, however, it is impossible to prospectively study this population. Although C difficile colitis is common, it is rare for it to progress to fulminant colitis. Furthermore, it would be impossible to randomize patients to either delayed or nonsurgical treatment because this would have serious ethical ramifications. The key to treatment of fulminant C difficile colitis is early surgical intervention. Our institution advocates this approach and our in-hospital mortality rate of 36% is significantly lower than those of other groups reporting mortality rates ranging from 41% to 80%.4,5,12,14,15 It is possible that some patients in this series may have survived hospitalization without surgery; however, the data presented in this article show that even in a series of comparably low mortality rates, a vasopressor or ventilatory requirement significantly increased mortality rates.
Conclusions C difficile colitis is common. Our study was a retrospective analysis of our institution’s experience with C difficile colitis and represents the second largest surgical series in the literature. Surgical treatment of this disease is associated with a high mortality rate if not addressed in a timely fashion. Commonly quoted preoperative indicators are not useful in determining the timing of surgeries. Failure to respond to appropriate medical therapy with associated systemic inflammatory response or peritonitis are indications for surgery. The development of a vasopressor requirement or need for intubation are ominous signs that also are indications for urgent surgical intervention. Early surgical intervention is the key to survival.
388
The American Journal of Surgery, Vol 196, No 3, September 2008
References 1. Fekety R, Shah AB. Diagnosis and treatment of Clostridium difficile colitis. JAMA 1993;269:71–5. 2. Bartlett JG. Clostridium difficile: history of its role as and enteric pathogen and the current state of knowledge about the organism. Clin Infect Dis 1994;18(Suppl 4):S265–72. 3. Aas J, Gessert CE, Bakken JS. Recurrent Clostridium difficile colitis: case series involving 18 patients treated with donor stool via nasogastric tube. Clin Infect Dis 2003;36:580. 4. Dallal RM, Harbrecht BG, Boujoukas AJ, et al. Fulminant Clostridium difficile: an underappreciated and increasing cause of death and complications. Ann Surg 2001;235:363–72. 5. Synnott K, Mealy K, Merry C, et al. Timing of surgery for fulminating pseudomembranous colitis. Br J Surg 1998;85:229 –31. 6. Knaus WA, Draper EA, Wagner DP, et al. APACHE II: a severity of disease classification system. Crit Care Med 1985;13:818 –29. 7. Bland JM, Altman DG. The odds ratio. BMJ 2000;320:1468. 8. Medich DS, Lee KK, Simmons RL, et al. Laparotomy for fulminant pseudomembranous colitis. Arch Surg 1992;127:847–53.
9. Marts BC, Longo WE, Vernava AM, et al. Patterns and prognosis of Clostridium difficile colitis. Dis Colon Rectum 1994;37:837– 45. 10. Loo VG, Poirier L, Miller M, et al. A predominantly clonal multiinstitutional outbreak of Clostridium difficile associated diarrhea with high associated morbidity and mortality. N Engl J Med 2005;353: 2442–9. 11. Johal SS, Hammond J, Solomon K, et al. Clostridium difficile associated diarrhea in hospitalized patients; onset in the community and role of flexible sigmoidoscopy. Gut 2004;53:673– 4. 12. Longo WE, Mazuski JE, Virgo KS, et al. Outcome after colectomy for Clostridium difficile colitis. Dis Colon Rectum 2003;47:1620. 13. Rubin M, Bodenstein LE, Kent KC. Severe Clostridium difficile colitis. Dis Colon Rectum 1995;38:350. 14. Klipfel AA, Schein M, Fahoum B, et al. Acute abdomen and Clostridium difficile colitis: still a lethal combination. Dig Surg 2000;17: 160 –3. 15. Grundfest-Borniatowski S, Quader M, Alexander F, et al. Clostridium difficile colitis in the critically ill. Dis Colon Rectum 1995; 39:619 –23.