Early Colectomy Saves Lives in Toxic Megacolon due to Clostridium difficile Infection

Nasim Ahmed, MD, FACS; Yen-Hong Kuo, PhD


South Med J. 2020;113(7):345-349. 

In This Article


Toxic colitis is one of the fulminant courses of C. difficile infection that occurs rarely.[3] In most instances, first-line antibiotics such as vancomycin and new medications such as fidaxomicin to treat acute C. difficile treatment are not always helpful in fulminant cases[14] and require operative intervention. Our study showed that performing an early colectomy significantly reduces the 30-day overall mortality.

Sailhamer and colleagues published results of their study of fulminant C. difficile colitis and found in-hospital mortality to be just above 34%.[18] This study analyzed 12 years' worth of data from 1996 to 2007 from a tertiary-care hospital and found a 4.1% incidence rate of fulminant C. difficile colitis. Of 199 patients, 75 (37.7%) patients underwent colectomy and the remaining patients were treated conservatively. The independent factors that led to colectomy included peritonitis, abdominal distension and positive computed tomography findings, free air, free fluid, thick wall, and distended colon. The study showed an age of 70 years or older and leukocytosis WBC ≥35,000/μL or leukopenia WBC <4000/μL, bandemia >10%; a need for vasopressors and vent-support were predictors of mortality, which was consistent with other studies.[3,7,19–22] The overall mortality rate in the colectomy group was lower when compared with the noncolectomy group; however, the study did not reach statistical significance (32% vs 36%, P = 0.08). The study also showed that patients who underwent colectomy early after intensive care unit admission had a better chance of survival.

There is evidence from prior studies, including the above study, that early operative intervention may be beneficial;[20–23] however, recognizing the patients who can benefit the most from early colectomies is not always easy. Few patient characteristics, including leukocytosis >50,000/μL, serum lactate level ≥5 mmol/L, and hemodynamic collapse and impending shock were described that led to colectomy in patients with fulminant C. difficile colitis.[19]

There is no clear guidance on the timing of colectomy in fulminant colitis cases. Even when colectomy is performed after the presentation of the above findings, mortality remained high.[5] In the present study, we accessed 5 years of the NISQIP data of all patients who underwent colectomy for toxic megacolon caused by C. difficile infection, based on the inclusion criteria. We compared the groups of patients who underwent colectomy before and after presenting with septic shock. On univariate analysis, the two groups had significance differences in baseline characteristics; therefore, propensity score matching and pair-matched analyses were performed to better balance the groups and remove some of the selection bias. After propensity score matching, no significant differences were found in patient demography, race, comorbidities, use of preoperative vent-support and blood transfusion, type of operation, ASA-class, and W-class. The average patient age was 65 years old, and patients were predominantly female. The majority of the patients underwent total abdominal colectomy, as others have suggested.[24] The patients who underwent a colectomy early compared with late colectomy had a lower mortality rate (13 [21%] vs 28 [45%], P = 0.009). The absolute risk difference was 0.24, with 95% CI 0.07–0.42, and the odds of survival in the early group were 3.1, with 95% CI 1.4–13.6. The patients in the late group died relatively early when compared with the early group (median [interquartile range] 5 [2.5–12] vs 12 [3.5–18], P = 0.39). There also was a significantly smaller number of total hospital stays in the early group compared with late group (median [95% CI] 20 [14–34] vs 25 [21–37], P = 0.011). There was no significant difference regarding postoperative complications except in the late group. A higher proportion of patients in the late group continued to have septic shock after colectomy when compared with the early group (38.2% vs 8.25%, P = 0.002).

This study was performed using the NISQIP database. This data repository houses millions of US patients' surgical data. This was a retrospective study; therefore, it carries the same limitation as any other retrospective study. The propensity score matching and pair-matched analyses between the two groups were used to remove selection bias and balance the groups; however, propensity score matching does not account for unknown and unmeasured variables.