WSES guidelines for management of Clostridium difficile infection in surgical patients

Clostridium difficile spores survive the acidic environment of the stomach and germinate in the intestine [19]. They act as an environmental reservoir for C. difficile and can facilitate spread among patients, as well as contributing to the high recurrence rates observed in CDI. The primary toxins produced by this bacterium are toxins A and B [20]. Some strains of C. difficile also produce binary toxin. Toxins A and B act as glucosyltransferases, promoting the activation of Rho GTPases leading to disorganization of the cytoskeleton of the colonocyte, and eventual cell death [21]. Since CDI is a toxin mediated infection, non toxigenic C. difficile strains are non-pathogenic. Over the years the respective roles and importance toxins A and B has been debated. Toxin A was thought to be the major virulence factor for many years, [22–24]. It is now established that both toxins A and B are important for inducing colonocyte death and colitis. In addition to toxins A and B, some strains produce a third toxin known as binary toxin [25–29]. Binary toxin has an ADP-ribosyltransferase function, which also leads to actin depolymerization [30, 31]. It has been demonstrated in C. difficile strains associated with nosocomial outbreaks of CDI with increased clinical severity [32, 33].

Typing is useful to differentiate C. difficile strains and to obtain epidemiological information. Different typing methods for C. difficile are actually available: restriction endonuclease analysis (REA), pulsed-field gel electrophoresis (PFGE), multilocus sequence typing (MLST), repetitive-element PCR typing, toxinotyping, multilocus variable-number tandem-repeat analysis (MLVA) and PCR- ribotyping [34].

C. difficile strains with increased virulence traits (hyper virulent), have been described in the last 10 years. In particular, PCR-ribotype 027, also known as North American pulsed-field gel electrophoresis type 1 (NAP1) or restriction endonuclease analysis group BI, has been associated with increased disease severity, recurrence and significant mortality [35].

Asymptomatic colonization may occur in 6 to 50 % of long-term care facility residents depending on whether CDI is endemic [36, 37]. In a 15-month prospective study of 4143 patients performed in six Canadian hospitals in Quebec and Ontario [38], 184 (4.4 %) had asymptomatic colonization at the time of unit admission, and 123 (3.0 %) had health care–associated C. difficile colonization.

Risk factors for CDI may be divided into three general categories: host factors (immune status, co-morbidities), exposure to CD spores (hospitalizations, community sources, long-term care facilities) and factors that disrupt normal colonic microbiome (antibiotics, other medications, surgery) [39].

Risk factors identified to date include, age more than 65 years, comorbidity or underlying conditions, inflammatory bowel diseases, immunodeficiency (including human immunodeficiency virus infection, hematologic malignancies and chemotherapy), malnutrition, and low serum albumin level [3, 40]. Diabetes mellitus is increasingly recognized as a risk factor for hospital and community-acquired CDI [41]. More recently, gene polymorphisms (e.g. IL-8) may be associated with increased risk for CDI but further studies are needed [42].

The effect of prior appendectomy on the development of C. difficile colitis has been debated in literature [43].

A recent review by Seretis et al. [44] of five studies conducted retrospectively was published in 2014. Although the results were conflicting regarding the impact of prior appendectomy on the occurrence or relapse of CDI, it appeared that an in situ appendix did not impact on the development of CDI.

In the retrospective analysis by Clanton et al. [45] on 55 patients who underwent colectomy for CDI between 2001 and 2011, a prior appendectomy was noted in 24 of 55 specimens (44, 99 % CI: 0.280–0.606). This was compared to an observed lifetime rate of appendectomy of 17.6 %. The rate of appendectomy in the cohort of patients who later underwent colectomy for CDI was significantly higher than would be expected in the general population (44 % vs 18 %, P < 0.01).

In a second retrospective study [46], of 388 patients with an intact appendix, 20 (5.2 %) developed fulminant infection and required colectomy, whereas of the 119 patients with a previous appendectomy, 13 (10.9 %) required colectomy. An increased severity of disease, indicated by increased rate of colectomy, occurred for the group with a history of appendectomy (P = 0.03).

A sub-group analysis of a large population based study published in 2013 [47] showed that appendectomy was not associated with adverse outcomes in CDI. Patients with appendectomy before CDI had no differences in risk factors, treatment, or outcomes including treatment failure, development of severe or severe-complicated CDI, and recurrence rates as compared with patients without appendectomy.

Larger prospective studies are needed to assess the impact of prior appendectomy on development and severity of CDI.

Factors that increase risk of exposure to C. difficile spores, such as increased duration of hospital stay may increase the risk of CDI. A length of stay > 2 weeks has been shown to be a risk factor for CDI [48]. Hospitals with well implemented infection prevention and control measures may reduce the risk of patients of developing CDI [49].

The indigenous gut microbiota is the complex community of microorganisms that populates the gastrointestinal tract. This micro-ecosystem plays a crucial role in protecting the intestines by providing resistance to colonization and infection by pathogenic organisms [50]. Gut microbiota also has immeasurable effects on homeostasis in the host [51]. Under normal conditions, the human gut microbiota may impede pathogen colonisation through general mechanisms such as direct inhibition through bacteriocins, nutrient depletion (consuming growth-limiting nutrients) or stimulation of host immune defences [38] though the exact mechanism by which the microbiota protects against CDI is unknown [52]. Disruption of the normal balance of colonic microbiota as a consequence of antibiotic use or other stressors, is, however, likely to be important [53].

Community-acquired CDI (CA-CDI) has been demonstrated in populations previously thought to be at low-risk, including younger patients not previously exposed to antibiotics [99]. Suggested risk factors include increasing outpatient antibiotic prescriptions, greater use of acid-suppression medications, an increase in the proportion of asymptomatic carriers in the community and novel risk factors like food and water contamination [100]. A sub-group analysis of a population-based epidemiological study of CDI in Olmsted County, Minnesota from 1991–2005 was published in 2012 [101]. Of 157 CA-CDI cases, the median age was 50 years and 75.3 % were female. Among CA-CDI cases, 40 % required hospitalization, 20 % had severe and 4.4 % had severe-complicated infection, 20 % had treatment failure and 28 % had recurrent CDI.

Recently a systematic review and meta-analysis investigated the association between commonly prescribed medications and comorbidities with CA-CDI [41]. Twelve publications (n = 56,776 patients) met inclusion criteria. Antimicrobial (odds ratio, 6.18; 95 % CI 3.80–10.04) and corticosteroid (1.81; 1.15–2.84) exposure were associated with increased risk of CA-CDI. Among the comorbidities, inflammatory bowel disease (odds ratio, 3.72; 95 % CI, 1.52–9.12), renal failure (2.64; 1.23–5.68), hematologic cancer (1.75; 1.02–5.68), and diabetes mellitus (1.15; 1.05–1.27) were associated with CA-CDI. By location, antimicrobial exposure was associated with a higher risk of CA-CDI in the United States, whereas proton-pump inhibitor exposure was associated with a higher risk in Europe. By life stages, the risk of CA-CDI associated with antimicrobial exposure greatly increased in adults older than 65 years.

In a meta-analysis by Garey et al. [102] found that continued use of non-C. difficile antibiotics after diagnosis of CDI (OR: 4.23; 95 % CI: 2.10–8.55; P < 0.001), concomitant receipt of antacid medications (OR: 2.15; 95 % CI: 1.13–4.08; P = 0.019), and older age (OR: 1.62; 95 % CI: 1.11–2.36; P = 0.0012) were significantly associated with an increased risk of recurrent CDI. Other factors identified in individual studies include age, hospital exposure, comorbid conditions, severe underlying illness, poor quality of life scores, initial disease severity and previous recurrent CDI [103, 104].

A recent systematic review and meta-analysis [105] was published to evaluate current evidence on the risk factors for recurrent CDI. A total of 33 studies (n = 18,530) met the inclusion criteria. The most frequent independent risk factors associated with recurrent CDI were age ≥65 years (risk ratio [RR], 1.63; 95 % confidence interval [CI], 1.24–2.14; P = .0005), additional antibiotics during follow-up (RR, 1.76; 95 % CI, 1.52–2.05; P < .001), use of proton-pump inhibitors (PPIs) (RR, 1.58; 95 % CI, 1.13–2.21; P = .008), and renal insufficiency (RR, 1.59; 95 % CI, 1.14–2.23; P = .007). The risk was also greater in patients previously on fluoroquinolones (RR, 1.42; 95 % CI, 1.28–1.57; P < .001).

The spectrum of symptomatic CDI ranges from mild diarrhea to severe disease or fulminant colitis and as many as 30 % of patients may develop recurrent CDI [106, 107].

Though diarrhea is the hallmark symptom of CDI it may not be present initially, possibly due to colonic dysmotility either from previous underlying conditions or possibly from the disease process itself [108].

This is especially important in surgical patients who may have a concomitant ileus. Therefore, in surgical patients it is important to have a high index of suspicion for the development of CDI.

Diarrhea may be accompanied by mild abdominal pain and cramps and if prolonged may result in altered electrolyte balance and dehydration. When this occurs in patients with severe comorbidity, particularly after surgery, non-severe CDI may increase morbidity significantly [109].

Severe CDI is associated with increased abdominal cramping and pain and constitutional features such as fever, leukocytosis, and hypoalbuminemia. The absence of diarrhoea in these patients may signal a progression to fulminant infection [110]. Though a wide variety of severity predictors for severe CDI has been described [111–115] international consensus for the definition of severe CDI is lacking [6, 7, 116].

One systematic review identifying risk factors for adverse outcomes of CDI was published by Abou Chakra et al. in 2012 [114]. Except for leukocytosis, albumin and age, there was much heterogeneity in the data and most studies were limited by small sample sizes.

To investigate the prognostic value of fever, leukocytosis, and renal failure, Bauer et al. [113] in 2012 analyzed the database of two randomized controlled trials, which contained information for 1105 patients with CDI. They found that both leucocytosis and renal failure were useful predictors of a complicated course of CDI. Miller et al. [115] in 2013 subsequently published an analysis of the same two clinical therapeutic trials to validate a categorization system to stratify CDI patients into severe or mild-moderate groups. A combination of five simple and commonly available clinical and laboratory variables (ATLAS) measured at the time of CDI diagnosis were able to accurately predict treatment response to CDI therapy. The ATLAS criteria included: age, treatment with systemic antibiotics, leucocyte count, albumin and serum creatinine [115].

The progression to fulminant C. difficile colitis is relatively infrequent [109] (1–3 % of all CDI) although mortality in this group of patients remains high due to the development of toxic megacolon with colonic perforation, peritonitis and septic shock and subsequent organ dysfunction. Systemic symptoms may result from toxin-induced inflammatory mediators released locally in the colon [117–119]. Studies have demonstrated a significant rise in the number of cases of fulminant colitis associated with multiple organ failure and increased mortality in recent years associated with the hypervirulent 027 strain of C. difficile [120, 121]. Early diagnosis and treatment is therefore important in reducing the mortality associated with fulminant colitis. Patients who present with organ failure including increased serum lactate or vasopressor requirements, should be assessed immediately with regard to early operative intervention [121].

Recurrence of symptoms after initial therapy for C. difficile, develops in 10–30 % of cases, and this often presents a clinical challenge. Patients may have several episodes of recurrence that may occur over a period of years [122–127]. Recurrence and reinfection are therefore difficult to distinguish by symptoms alone, but may be distinguished if the strain of C. difficile is typed.

RCDI may be either a consequence of germinating resident spores remaining in the colon after antibiotic treatment has stopped, or re-infection from an environmental source.

Even though consensus regarding factors associated with CDI recurrence is not universal learning algorithms have been developed to predict CDI recurrence with good sensitivity [128].

Ultimately distinction between recurrence and reinfection can only be achieved if the strain of C. difficile is ‘typed’ using molectular epidemiology [129].

Patients who develop CDI have increased hospital length-of-stay, higher medical care costs, more hospital re-admissions, and higher mortality [130–132].

These consequences are also found for surgical patients with CDI.

In the Zerey et al. analysis [8] epidemiologic data suggested that the infection was most prevalent after emergency operations and among patients having intestinal tract resections. Infection with C. difficile was an independent predictor of increased length of stay, which increased by 16.0 days (95 % CI 15.6, 16.4 days; p < 0.0001) in the presence of infection. Total charges increased by $77,483 (95 % CI $75,174, $79,793; p < 0.0001), and there was a 3.4-fold increase in the mortality rate (95 % CI 3.02, 3.77; p < 0.0001) compared with patients who did not acquire C. difficile.

In the Abdelsattar et al. study [11] three procedure groups had higher odds of postoperative CDI: lower-extremity amputations (adjusted odds ratio [aOR], 3.5; P = .03), gastric or esophageal operations (aOR, 2.1; P = .04), and bowel resection or repair (aOR, 2; P = .04). Postoperative CDI was independently associated with increased length of stay (mean, 13.7 days vs 4.5 days), emergency department presentations (18.9 vs 9.1 %) and readmissions (38.9 vs 7.2 %, all P < .001).

Data from Nationwide Inpatient Sample database in patients who underwent vascular surgery [79], showed that in 2011 patients who had experienced CDI had median length of stay 15 days (IQR 9, 25 days) compared with 8.3 days for matched patients without CDI, in-hospital mortality 9.1 % (compared to 5.0 %), and $13,471 extra cost per hospitalization. The estimated cost associated with CDI in vascular surgery in the United States was about $98 million in 2011. Data from the National Inpatient Sample examined just patients with lumbar surgery and found CDI increased length of stay by 8 days, hospital costs by 2-fold and increased inpatient mortality by 36-fold [133].

Higher mortality was also observed for liver transplant recipients (from 2000 to 2010) at Detroit hospital [134].

The ACS-NSQIP database from 2005 to 2010 was used by Lee et al. to study emergently performed open colectomies for a primary diagnosis of C difficile colitis in US [135]. The overall mortality was 33 % (111/335). Age 80 years or older, preoperative dialysis dependence, chronic obstructive pulmonary disease), and wound class III were associated to patients mortality. Thrombocytopenia (platelet count < 150 × 10³/mm³), coagulopathy (International Normalized Ratio > 2.0), and renal insufficiency (blood urea nitrogen > 40 mg/dL) were associated with a higher mortality as well.

Recently a study was performed to quantify additional hospital stay attributable to CDI in four European countries, by analyzing nationwide hospital-episode data [5]. Patients in England had the longest additional hospital stay attributable to CDI at 16.09 days, followed by Germany at 15.47 days, Spain at 13.56 days, and The Netherlands at 12.58 days, derived using regression analysis. Propensity score matching indicated a higher attributable length of stay of 32.42 days in England, 15.31 days in Spain, and 18.64 days in The Netherlands. Outputs from this study consistently demonstrate that in European countries, for patients whose hospitalization is complicated by CDI, the infection causes a statistically significant increase in hospital length of stay.

12) Unnecessary antimicrobial agent(s) and proton pump inhibitors should be discontinued if CDI is suspected (Recommendation 1 C).

13) Empirical therapy for CDI should be avoided unless there is a strong suspicion for CDI. If a patient has a strong suspicion for CDI, empirical therapy for CDI should be considered while awaiting test results (Recommendation 1 B).

In cases of suspected severe CDI, antimicrobial agent(s) should be discontinued, if possible [158].

A meta-analysis addressing factors associated with prolonged symptoms and severe disease due to Clostridium difficile showed that continued use of antimicrobials for infections other than CDI is significantly associated with an increased risk of CDI recurrence [159].

When antimicrobial therapy is indicated for symptomatic cases with a positive C. difficile toxin result, options include metronidazole, oral or intraluminal vancomycin and fidaxomicin [160–166].

14) Metronidazole is recommended for the treatment of mild-moderate disease (Recommendation 1 A).

Given at a dose of 500 mg orally 3 times a day for 10 days, metronidazole has been shown to be an inexpensive and effective treatment of non-severe CDI [167]. Metronidazole can also be administrated intravenously with or without intraluminal vancomycin in patients unable to take oral medication e.g. those with post-surgical ileus.

A Cochrane analysis published in 2011 [167] reviewed 15 studies on the antibiotic treatment for CDI in adults. In three randomized controlled trials comparing symptomatic cure between metronidazole and vancomycin, no statistically significant difference was found [167]. Symptomatic cure was achieved in 79 % of patients who received vancomycin compared with 71 % of patients who received metronidazole (three studies; 335 patients; RR 0.91; 95 % CI 0.81–1.03, p 0.14).

15) Oral vancomycin is recommended for treatment of patients with severe disease, or for patients with mild-moderate disease who do not respond to metronidazole. (Recommendation 1 A).

Vancomycin orally 125 mg four times daily for 10 days is considered superior to metronidazole in severe C. difficile disease [168–170]. This may reflect the superior pharmacokinetic properties of vancomycin which is concentrated in the gut lumen. Doses of up to 500 mg have been used in some patients with severe CDI [7] although there is little evidence for this in the literature.

16) In patients in whom oral antibiotics cannot reach the colon, vancomycin may be administered by enema and metronidazole can be given intravenously (Recommendation 1 B).

Intravenous vancomycin has no effect on CDI since the antibiotic is not excreted into the colon. Vancomycin enema may be an effective therapy for patients who cannot tolerate the oral preparation or patients with ileus who have delayed passage of oral antibiotics from the stomach to the colon. Trans-stoma vancomycin may also be effective in surgical patients with Hartmann resection, ileostomy, or colon diversion. A single-hospital, retrospective chart review on 47 consecutive patients with C. difficile colitis treated with intra-colonic vancomycin (ICV) was published by Kim PK et al. in 2013 [171]. Thirty-three of 47 patients (70 %) with severe C. difficile colitis responded to adjunct ICV with complete resolution without surgery. Multivariable analysis suggested that failures to intra-colonic vancomycin enemas occurred in patients who were older and frail with albumin < 2.5 g/dl and early surgery should be considered for those patients. Early surgery should also be offered to those patients who are failing maximal medical therapy that include ICV enemas.

17) Fidaxomicin may be used to treat CDI, especially in the patients at higher risk for recurrence (e.g. elderly patients with severe underlying disease or those requiring receiving concomitant antibiotics) (Recommendation 1 A).

Fidaxomicin orally 200 mg twice daily for 10 days may be an alternative to vancomycin in some patients with CDI [172, 173].

Fidaxomicin was non-inferior to vancomycin for initial cure of CDI in two prospective trials [164, 165]. In a first double-blind, randomized, non-inferiority trial [164] 629 adults with acute symptoms of C. difficile infection and a positive result on a stool toxin test were enrolled and randomly assigned to receive fidaxomicin (200 mg twice daily) or vancomycin (125 mg four times daily) orally for 10 days. The rates of clinical cure with fidaxomicin were non-inferior to those with vancomycin in both the modified intention-to-treat analysis (88.2 % with fidaxomicin and 85.8 % with vancomycin) and the per-protocol analysis (92.1 % and 89.8 %, respectively). Significantly fewer patients in the fidaxomicin group than in the vancomycin group had a recurrence of the infection, in both the modified intention-to-treat analysis and the per-protocol analysis. In a second multi-centre, double-blind, randomized, non-inferiority trial [165] 535 patients, 16 years or older with acute, toxin-positive C difficile infection were randomly allocated (1:1) to receive oral fidaxomicin (200 mg every 12 h) or oral vancomycin (125 mg every 6 h) for 10 days. Non-inferiority was shown for both the modified intention-to-treat analysis (15.4 % vs. 25.3 %, P = 0.005) and the per-protocol analysis (13.3 % vs. 24.0 %, P = 0.004). Patients receiving concomitant antibiotics for other infections had a higher cure rate with fidaxomicin (46 [90 · 2 %] of 51) than with vancomycin (33 [73 · 3 %] of 45; p = 0 · 031). Fidaxomicin may be useful for treating patients who are considered at high risk for recurrence (elderly patients with multiple comorbidities who are receiving concomitant antibiotics). However, it is important to note that there are no data available on the efficacy of Fidaxomicin in severe life-threatening disease.

The use of other antibiotics such as tigecycline [174, 175] fusidic acid, teicoplanin, rifamixin [167] and nitazoxanide [176], has been described in the literature, but they are not currently recommended for general use.

Patients with fulminant colitis (FC) who progress to systemic toxicity require surgical intervention.

To determine clinical predictors for the development of fulminant colitis in patients with CDI a 10-year retrospective review of FC patients who underwent colectomy was performed and compared with randomly selected age- and sex-matched non-fulminant CDI patients at a single institution study by Girotra in 2012 [177]. Predictive clinical and laboratory features included: old age (>70 years), prior CDI, profound leukocytosis (>18,000/mm³), hemodynamic instability, use of anti-peristaltic medications, and a clinical triad of increasing abdominal pain, distention and diarrhea.

18) Patients with severe CDI who progress to systemic toxicity should undergo early surgical consultation and evaluated for potential surgical intervention (Recommendation 1 C).

Patients with severe CDI who progress to systemic toxicity are likely to have serious comorbidities. Delaying surgery in this group leads to increased likelihood of adverse outcomes [178], although some reports show that a short period of medical optimization can improve outcomes before colectomy [179].

There are no reliable clinical and/or laboratory findings that can predict those patients who will respond to medical therapy and those who will subsequently need surgery [180].

Data comparing mortality rates between surgical and medical treatment for fulminant C. difficile colitis were published in a recent systematic review by Stewart et al. [181]. Five hundred and ten patients with FC were identified in six studies. Emergency colectomy for patients with FC provided a survival advantage compared with continuing antibiotics. When all six studies numbering 510 patients were analysed, the pooled adjusted odds ratio of mortality comparing surgery with medical therapy, and weighted by the contribution of each study, was 0.70 (0.49–0.99) leading the authors to conclude that emergency colectomy has a therapeutic role in treating complicated C. difficile colitis.

Patients presenting with organ failure (acute renal failure, mental status changes, or cardiopulmonary compromise) also need prompt intervention.

The timing of surgical intervention is the key for survival of patients with FC [182–185].

Seder et al. [186] described 6,841 patients with CDI and showed a decreased mortality associated with surgery performed before the need for vasopressor requirement, especially in the patients <65 years old. Hall et al. [184] reviewed 3,237 consecutive cases of CDI and showed an increased mortality rate when surgical exploration was performed after intubation or the development of respiratory failure and the use of vasopressors.

Recently a risk scoring system (RSS) for daily clinical practice was designed by van der Wilden et al. [187]. Age greater than 70 years was assigned 2 points, white blood cell count equal to or greater than 20,000 x 10⁹/L or equal to or less than 2,000 x 10⁹/L was assigned 1 point, cardiorespiratory failure was assigned 7 points, and diffuse abdominal tenderness on physical examination was assigned 6 points. A value of 6 points was determined to be the threshold for reliably dividing low-risk (<6) from high-risk (≥6) patients. Only patients with cardiorespiratory failure or diffuse abdominal tenderness were high risk.

Ferrada et al. [188] reviewed the existing literature on the treatment of CDI and published practice management guidelines (PMG) for the Eastern Association for the Surgery of Trauma (EAST). The authors strongly recommended, that adult patients with CDI undergo early surgery before developing shock and the need for vasopressors. Although timing remains controversial Ferrada et al. found that it was between 3 days and 5 days after diagnosis in patients who are worsening or not clinically improving [188].

Many factors have been described as predictors of mortality in patients who undergo emergency intervention.

Sailhamer et al. [189] reviewed the records of 4796 inpatients diagnosed with C difficile colitis. In 199 patients (4.1 %) with fulminant C difficile colitis the in-hospital mortality rate was 34.7 %. Independent predictors of mortality included age 70 years or older, severe leukocytosis or leukopenia (white blood cell count, >or = 35 000 × 10⁹/L or <4000 × 10⁹/L) or bandemia (neutrophil bands, >or = 10 %), and cardiorespiratory failure (intubation or vasopressors). Survival rates were higher in patients who were cared for by surgical vs nonsurgical departments.

The ACS-NSQIP database from 2005 to 2010 was used by Lee et al. to study emergency open colectomies performed for C difficile colitis in the USA [190]. The overall mortality was 33 % (111/335). Age 80 years or older, preoperative dialysis dependence, chronic obstructive pulmonary disease, and wound class III were associated high patient mortality. Thrombocytopenia (platelet count < 150 × 10³/mm³), coagulopathy (International Normalized Ratio > 2.0), and renal insufficiency (blood urea nitrogen > 40 mg/dL) were also associated with a higher mortality.

A systematic review and meta-analysis of outcomes following emergency surgery for C. difficile colitis was published by Banghu et al. [191]. Thirty-one studies were included, which presented data for 1433 patients undergoing emergency surgery for C. difficile colitis. It concluded that the strongest predictors for postoperative death were those relating to preoperative physiological status: preoperative intubation, acute renal failure, multiple organ failure and shock requiring vasopressors.

19) Resection of the entire colon should be considered to treat patients with fulminant colitis (FC) (Recommendation 1 B).

20) Diverting loop ileostomy with colonic lavage may be a useful alternative to resection of entire colon (Recommendation 2 C).

21) Patients with FC should be treated with high dose oral or by enema vancomycin (500 mg, 6 hourly) in combination with intravenous metronidazole (500 mg, 8 hourly). (Recommendation 1 C).

In the Bhangu et al. meta-analysis [191] the most commonly performed operation for treatment of FC was total colectomy with end ileostomy (89 %, 1247/1401). When total colectomy with end ileostomy was not performed, reoperation to resect further bowel was needed in 15.9 % (20/ 126). In the recent meta-analysis by Ferrada et al. [188], 17 studies comparing colectomy versus other procedures or no surgery as treatment for CDI were analyzed. The authors recommended that total colectomy (vs. partial colectomy or other surgery) is the procedure of choice for patients with C. difficile colitis.

To evaluate the role of emergency colectomy in patients with FC, and to identify subgroups of patients that may benefit Lemontagne et al. [192] published a retrospective observational cohort study of 165 cases of FC that required ICU admission or prolongation of ICU stay in 2 tertiary care hospitals of Quebec, Canada. Eighty-seven (53 %) cases died within 30 days of ICU admission, of which almost half (38 of 87, 44 %) died within 48 h of ICU admission. The independent predictors of 30-day mortality were leukocytosis > or = 50 × 10(9)/L, lactate > or = 5 mmol/L, age > or = 75 years, immunosuppression and shock requiring vasopressors. Patients who underwent an emergency colectomy were less likely to die than those treated medically. Colectomy was more beneficial in patients aged 65 years or more, in immunocompetent patients and in patients with a leukocytosis > or = 20 × 10⁹/L or lactate between 2.2 and 4.9 mmol/L.

Diverting loop ileostomy with antegrade colonic lavage may be a colon preserving alternative to total colectomy [193, 194]. To evaluate whether a minimally invasive, colon-preserving approach may be an alternative to sub-total colectomy in the treatment of FC, a historical control group study was performed at the University of Pittsburgh Medical Center or and the Veterans’ Administration Healthcare System, Pittsburgh between June 2009 and January 2011 [193]. All patients with FC were managed by a loop ileostomy, intraoperative colonic lavage with warmed polyethylene glycol 3350/electrolyte solution via the ileostomy and postoperative antegrade instillation of vancomycin flushes via the ileostomy. Forty-two patients were treated during this time period. There was no significant difference in age, sex, pharmacologic immunosuppression, and Acute Physiology and Chronic Health Evaluation-II scores between the studied cohort and historical controls. The operation was accomplished laparoscopically in 35 patients (83 %). This treatment strategy resulted in reduced mortality compared to their historical population. Preservation of the colon was achieved in 39 of 42 patients (93 %). Of note, in this study vancomycin antegrade enemas were continued via the ileostomy every 6 h for 10 days after ileostomy formation and this likely augmented the effect of the defuctioning surgery.

22) Supportive measures, including intravenous fluid resuscitation and electrolyte replacement, should be provided to all patients with severe C. difficile infection (Recommendation 1 C).

Diarrhea results in significant volume depletion and electrolyte abnormalities, and fluid and electrolyte imbalance should be promptly corrected [119, 120].

23) Early detection of shock and aggressive management of underlying organ dysfunction are essential for optimum outcomes in patients with fulminant colitis (Recommendation 1 C).

Early detection and prompt aggressive treatment of the underlying organ dysfunction is an essential component of improving outcome of critical ill patients [120].

Severe CDI may present with a fulminant course and may be associated with great morbidity and high mortality. Physiologic support including close invasive monitoring in an intensive care unit setting and aggressive resuscitation are often necessary in fulminant colitis.

Recurrence is diagnosed when CDI recurs <8 weeks after the onset of a previous episode, provided the symptoms from the previous episode resolved after completion of initial treatment and other causes have been excluded. Symptomatic recurrent C difficile infection (RCDI) occurs in approximately 20 % of patients and is challenging to treat [195]. Patients with recurrence of CDI should therefore be treated by clinicians who have experience in treating the infection.

24) Agents that may be used to treat the first recurrence of CDI include metronidazole, for non-severe RCDI, and vancomycin for severe RCDI. (Recommendation 1 B).

25) Fidaxomicin may be used as an alternative agent (Recommendation 1 B).

A systematic review on the treatment of RCDI was recently published [196]. Metronidazole and vancomycin have a good evidence base for use in RCDI but heterogeneity in treatment duration and treatment doses between the studies precluded robust conclusions. Fidaxomicin may also have a role in the treatment of first recurrence. Fidaxomicine was superior to vancomycin in terms of recurrences, with significantly less recurrence at 28 days. This was confirmed in some subgroup analysis [197].

26) In subsequent recurrence of CDI (2nd or later) oral vancomycin or fidaxomicin is recommended (Recommendation 1 B).

Vancomycin and fidaxomicin are equally effective in resolving CDI symptoms but fidaxomicin has been shown to be associated with a lower likelihood of CDI recurrence after a first recurrence [164, 165, 197]. However, there are no prospective randomized controlled trials investigating the efficacy of fidaxomicin in patients with multiple recurrences of CDI. Vancomycin is often administered using a prolonged tapered and/or pulsed regimen which may be more effective than a standard 10 to 14 day course although no RCTs have been reported [198].

27) Probiotics may be considered as an adjunctive treatment to antibiotics for immunocompetent patients with RCDI (Recommendation 2 B).

Little evidence exists to support the use of probiotics in the first episode of CDI [116]. Two randomized controlled trials showed some effectiveness for Saccharomyces boulardii CNCM I-745 in recurrent CDI. The first demonstrated a lower relapse rate compared with a placebo control group (35 vs 65 % in the placebo group) [199] and the second found that the combination of S. boulardii (1 g/d) with high dose vancomycin (2 g/d) was more effective than high dose vancomycin and placebo (17 vs 50 % recurrence rate) [200]. Other studies with Lactobacillus strains (L. rhamnosus GG or L. plantarum 299v) were stopped prematurely due to enrollment problems [201]. Probiotics should not be administered to patients at risk of bacteraemia or fungaemia [116].

There is limited evidence to support the use of probiotics for the primary prevention of CDI from developing. A meta-analysis of 11 studies was in published 2012 [202]. Two studies showed significantly lower rates of CDI among the probiotic recipients. A meta-analysis of three studies that used the probiotic combination Lactobacillus acidophilus CL1285 and Lactobacillus casei LBC80R and a combined analysis of those studies with four studies that used Saccharomyces boulardii, showed lower CDI rates in recipients of probiotics compared with recipients of placebo (risk ratio = 0.39; 95 % confidence interval 0.19–0.79. However, given the potential risk of bloodstream infection with these organisms further studies are warrented before their use can be recommended routinely.

28) Intestinal or faecal microbiota transplantation (IMT or FMT) may be an effective option for the treatment of RCDI (Recommendation 1 B).

Intestinal or faecal microbiota transplantation (IMT or FMT) has been considered as an alternative therapy to treat RCDI [203–208]. It involves infusing intestinal microorganisms (in a suspension of healthy donor stool) into the intestine of patients to restore the intestinal microbiota.

The rationale of FMT is that disruption of the normal balance of colonic flora allows C. difficile strains to grow and produce CDI. By reintroducing normal flora via donor faeces, the imbalance may be corrected, and normal bowel function re-established [203].

FMT has not been widely adopted as a therapeutic tool probably due to concerns regarding safety and acceptability [204].

A systematic literature review of IMT treatment for RCDI and pseudomembranous colitis was published in 2011 by Gough et al. [205]. In 317 patients treated across 27 case series and reports, IMT was highly effective, showing disease resolution in 92 % of cases. In those studies, 35 % of patients received IMT via enema, with a response rate of 95; 23 % patients received IMT via naso-jejunal tube by gastroscope, with a response rate of 76; and 19 % via colonoscopy, with a response rate of 89 %. Effectiveness varied by route of instillation, relationship to stool donor, volume of IMT given, and treatment before infusion.

Recently a systematic review was published by Cammarota et al. [206]. Twenty full-text case series, 15 case reports, and 1 randomized controlled study were included for the final analysis. Almost all patients treated with donors’ fecal infusion experienced recurrent episodes of CD-associated diarrhea despite standard antibiotic treatment. Of a total of 536 patients treated, 467 (87 %) experienced resolution of diarrhea. Diarrhea resolution rates varied according to the site of infusion: 81 % in the stomach; 86 % in the duodenum/jejunum; 93 % in the cecum/ascending colon; and 84 % in the distal colon. No severe adverse events were reported with the procedure.

In a recently published randomized clinical trial by van Nood et al. [208] patients with RCDI were randomised to three groups; 1) vancomycin regime only; 2) vancomycin with duodenal infused FMT and 3) vancomycin and bowel lavage. In the FMT treated group an 81% reduction in diarrhoea was observed. The FMT group were observed to have normalization of their intestinal bacterial composition which was similar to that of the donor. Although, this trial has shown exciting results, these need to be interpreted with caution as the trial included only small number of patients, was not blinded, and was aborted early due to profound differences in the groups. It has also been criticised for potentially having several potential biases.

FMT may be administered via enemas or as a slurry given via a nasogastric tube. In the fall of 2014, Youngster et al. [209] reported their experience on utilizing frozen FMT capsules in 20 patients who had RCDI. Fourteen patients (70 %) had resolution of diarrhea after the first treatment, and an additional 4 patients responded after a second treatment, for a clinical resolution rate of 90 %.

29) FMT may be effective in immunocompromised patients and patients who have had solid organ transplants (Recommendation 2 B).

Patients who are immunocompromised are at increased risk of CDI. During the last two years the first data on FMT in immunocompromised patients began to appear in the medical literature [210].

A multicenter retrospective series on the use of FMT in immunocompromised (IC) patients with CDI that was recurrent, refractory, or severe was published in 2014 [211]. Reasons for IC included: HIV/AIDS (3), solid organ transplant (19), oncologic condition (7), immunosuppressive therapy for inflammatory bowel disease (IBD; 36), and other medical conditions/medications (15).

This series demonstrated the effective use of FMT for CDI in IC patients with few serious adverse events or related adverse events.

30) IVIG should only be used as adjunct therapy in patients with multiple recurrent or fulminant CDI until results from large, randomized controlled trials are available (Recommendation 2 C).

IVIG treatment has been proposed based on the evidence that the level of immune response to C. difficile colonization is a major determinant of magnitude and duration of clinical manifestations. Passive immunization with IVIG has been reported to be successful in several small series. A review by Abourgergi [212] of fifteen small, mostly retrospective and non-randomized studies documented success with IVIG in the treatment of protracted, recurrent, or severe CDI. The authors concluded IVIG should only be used as adjunct therapy until results from large, randomized controlled trials are available.

31) Infusion with monoclonal antibodies may be of use to prevent recurrences of CDI, particularly in patients with CDI due to the 027 epidemic strain (Recommendation 2 C).

In a phase II clinical trial [213], the use of monoclonal antibodies to toxins A and B as an adjunct to antibiotics was shown to decrease recurrence rates in patients with CDI compared with placebo (7 vs. 25 % respectively; 95 % confidence interval, 7 to 29; P < 0.001). The recurrence rates among patients with the epidemic BI/NAP1/027 strain were 8 % for the antibody group compared with 32 % for placebo (P = 0.06); among patients with more than one previous episode of CDI, recurrence rates were 7 and 38 %, respectively (P = 0.006). The authors concluded that the addition of monoclonal antibodies against C. difficile toxins to antibiotic agents significantly reduced the recurrence of C. difficile infection. The findings of this study require confirmation before firm recommendations can be made.

32) Tube feeding patients should be clinically assessed due to their risk for developing CDI (Recommendation 2 C).

It is widely accepted that enteral nutrition (EN) maintains gut mucosal integrity which leads to decreased intestinal permeability, decreased infections, and an improved immunological status. EN during episodes of diarrhea may be well tolerated and may improve enterocyte healing and maintenance of enzyme activity [214, 215]. EN, however, has also been associated with increased risk of CDI [216]. Bliss, et al. evaluated 76 tube-fed and non tube-fed hospital patients for the development of CDI [217]. Patients were controlled for age, severity of illness and duration of hospitalization. Patients who were tube-fed were statistically more likely to develop C difficile associated diarrhea (20 versus 8 % p = 0.03). One of the reasons may be prolonged use of elemental diets. It is known that critically ill patients tolerate feeding well if the feed is given in elemental form and delivered beyond the stomach into the jejunum because it is totally absorbed within the upper small intestine [218]. Elemental diets are completely absorbed within the small intestine and therefore deprive the colonic microbiota of their source of nutrition, such as dietary fiber, fructose oligosaccharides, and resistant starch [219]. The resultant suppression of colonic fermentation may therefore lead to the disruption of the normal gut flora and the creation of a “permissive” environment for C. difficile colonization and subsequent infection. In feeding tube patients the conversion of elemental diet feeding to a diet containing adequate indigestible carbohydrate after the first week of critical illness may, in theory, be useful.

Recently, Puri et al. [220] reported that daily concomitant treatment with 4 g cholestyramine in patients receiving long-term intravenous ceftriaxone (2 to 4 g ceftriaxone daily, for an average of >10 weeks) was associated with CDI in only three out of 46 patients (6.5 %) compared with 23.1 % of those receiving ceftriaxone alone [221]. Cholestyramine (or colestyramine) is a hydrophilic, water insoluble, non-digestible basic anion-exchange resin which can bind luminal TcdA and TcdB.

Studies have also investigated the possible value of exogenous Phosphatidylcholene (PC) administration for reinforcement of the mucus layer [222, 223]. Mucus or “exogenous” mucus in the form of PC may have a synergistic role with secretory IgA as a barrier against C. difficile toxin A though additional studies are needed to demonstrate its clinical benefit before recommendations can be made [222, 223].

33) The use of anti-peristaltic agents for the treatment of CDI should be discouraged. If anti-peristaltic, if used in isolation agents, are used to control persistent symptoms in patients with CDI they must always be accompanied by medical therapy (Recommendation 2 C).

A review of the literature regarding anti-motility treatment of CDI found 55 patients with CDI who were exposed to anti-motility agents [224].

Nine patients (16 %) died, and 27 patients (49 %) had unknown outcomes. Seventeen patients (31 %) with CDI developed colonic dilation; 5 of these patients with severe CDI died. However, all patients who experienced complications or died were given anti-motility agents alone initially, without an appropriate antibiotic and 23 patients who received metronidazole or vancomycin co-administered with the anti-motility agent experienced no complications. Further study of the role of anti-motility agents in providing symptomatic relief and reducing environmental contamination with infectious stool may be warranted though, until there is clear evidence of benefit, their use in patients with CDI should be avoided [116].

34) Proper antimicrobial stewardship in selecting an appropriate antibiotic and optimizing its dose and duration to cure an infection may prevent the emergence of C. difficile (Recommendation 1 B).

Despite vigorous infection control measures until recently, CDI was causing an increasing problem in healthcare facilities worldwide. As CDI is thought to follow disruption to the normal bacterial flora of the colon occurring as a consequence of antibiotic use [225], it is logical that antibiotic stewardship programs may be useful in preventing CDI [226]. Good antimicrobial stewardship involves ensuring appropriate antibiotic choice and optimizing antibiotic dose and duration to cure an infection while minimizing toxicity and conditions conducive to CDI. Recently, a systematic review [227] of interventions to improve antibiotic prescribing practices for hospital inpatients suggested that reducing excessive antibiotic prescribing can prevent hospital-acquired infections and that interventions to increase effective prescribing improve clinical outcome. It would appear that cephalosporin and quinolone antibiotics may be particularly high risk, in this context [116, 228].

35) Patients with suspected or proven CDI should be placed in contact (enteric) precautions (Recommendation 1 B).

Prompt identification of patients with symptomatic CDI is essential so that appropriate isolation precautions can be put into effect.

This is particularly important in reducing environmental contamination as spores can survive for months in the environment [229], despite regular use of environmental cleaning agents.

Contact (enteric) precautions patients with CDI should be maintained until the resolution of diarrhea, which is demonstrated by passage of formed stool for at least 48 h. Patients with known or suspected CDI should ideally be placed in a private room [116, 230] with en-suite hand washing and toilet facilities. If a private room is not available known CDI patients may be cohort nursed in the same area [231] though the theoretical risk of transfection with different strains exists.

This is supported by a retrospective cohort of 2859 patients by Chang et al. [232]. Patients who were roommates or neighbors of a patient with CDI were at risk of nosocomial acquisition of CDI (RR, 3.94; 95 % CI, 1.27–12.24).

36) Hand hygiene with soap and water is a cornerstone of the prevention of C. difficile . Hand hygiene, contact precautions and good cleaning and disinfection of the environment and patient care equipment, should be used by all health-care workers contacting any patient with known or suspected CDI (Recommendation 1 B).

Hand hygiene with soap and water and the use of contact precautions along with good cleaning and disinfection of the environment and patient equipment, should be used by all health-care workers contacting any patient with known or suspected CDI. Hand hygiene is a cornerstone of prevention of nosocomial infections, including C. difficile. Alcohol-based hand sanitizers are highly effective against non–spore-forming organisms, but they may not kill C. difficile spores or remove C. difficile from the hands [233, 234].

The most effective way to remove them from hands is through hand washing with soap and water.

For environmental cleaning, hypochlorite disinfection such as sodium hypochlorite solutions are suggested for regular use in patient areas where C. difficile transmission is ongoing [231].

Though disposable glove use during care of a patient with CDI may be effective in preventing the transmission of C. difficile [230], these must be removed at the point of use and hands thoroughly decontaminated afterwards through soap and water hand washing.