C. difficile toxin A elicits intestinal fluid secretion and neutrophil infiltration by both mast cell-dependent and -independent pathways, and substance P participates in both pathways[38].

Extensive mitochondrial damage occurs within 15 min in cells exposed to toxin A. Diminished ATP concentrations and increased oxygen radicals contribute to cytotoxicity from this bacterial toxin[39].

The toxins damage the tight junctions of the intestinal epithelium. Tight junctions are crucial determinants of barrier function in intestinal epithelia, and are regulated by Rho guanosine triphosphatase. Rho kinase (ROCK) is a downstream effector of Rho. ROCK inhibition in calcium switch assays has shown that ROCK is necessary for the assembly of tight and adherens junctions. ROCK also is critical for assembly of apical junctional proteins and F-actin cytoskeleton organization during junctional formation[40].

C. difficile toxins A and B are glucosyltrans-ferases, which catalyze the inactivation of Rho proteins. C. difficile toxins act via translocation into target cells, and do their damage through autocatalytic processes by inactivating low-molecular-mass GTP-binding proteins of the Rho GTPase family involved in cellular signaling. This leads to cytotoxicity, including depolymerization of the target cell’s actin cytoskeleton. Thus, these toxins glycosylate members of the Rho GTPase family, and this GTPase inactivation leads to depolymerization of the cell’s actin cytoskeleton and, ultimately, cell death[41]. In addition, the C. difficile toxins further damage the intestine’s target cells by initiating massive cellular immune responses; i.e. neutrophilic infiltration with up-regulation and release of cytokines, such as interleukin (IL)-8, IL-6, IL-1β, leukotrienes B4 and interferon-γ.

Part of the mammalian immune response falls to the innate immune system called defensins and, specifically, human α-defensins produced by leukocytes, mucosal epithelial cells, and skin. Defensins, one of evolution’s major groups of antibiotic peptides, have broad-spectrum antibiotic activity against Gram-positive and Gram-negative bacteria, fungi, and viruses[42–44]. Defensins are characterized by a conserved 6-cysteine array. Each cysteine has intra-molecular disulfide bonds that are essential to protection against proteolysis[45]. Defensins also are known to contribute to wound healing, chemotaxis, and cytokine function[4647]. Defensins are part of two major groups of antimicrobial peptides: defensins and cathelicidins. These groups of human defensins consist in part of alpha, beta and omega defensins, human neutrophil protein (HNP)-1, HNP-3, cathelicidin LL-37 and enteric human defensin (HD)-5. These peptides play a role in the innate immune response, by deactivating various microbial pathogens, as well as specific bacterial exotoxins.

The antibiotic activity of both HNPs and HD5 is well documented in host defenses against enteric pathogens[4849]. HD5 and HD6 are produced and stored in Paneth cell secretory granules[50], along with a variety of additional Paneth cell products demonstrated to have antimicrobial and immune activity[51–55]. The impact of defensins on C. difficile disease has been described by Giesemann et al[56] and others[57–60].

Giesemann et al[56] have studied the effects of α-defensin HNP-1, HNP-3, and enteric HD-5 on the activity of C. difficile toxins A and B. They found that the treatment of cells with human α-defensins caused a loss of cytotoxicity of toxin B, but not of toxin A. In this study, only α-defensins, but not β-defensin-1 or cathelicidin LL-37, inhibited toxin B-catalyzed in vitro glucosylation of Rho GTPases in a competitive manner. This indicates that human α-defensins interact with high affinity for C. difficile toxin B. Defensins thereby provide a defense mechanism against clostridial glucosylating cytotoxins. At high concentrations, defensins (HNP-1 ≥ 2 &mgr;mol/L) also cause high-molecular-mass aggregates of C. difficile toxins, thus further decreasing their toxic effects on target cells.

C. difficile has been found in approximately 3% of normal adults and up to 40% of hospitalized patients[7]. However, as Salzman emphasizes: “only about one third of patients harboring C. difficile develop colitis, whereas the rest remain asymptomatic[61]”. Giesemann et al[56] have shown that α-defensins inhibit C. difficile toxin B, which offers insight into the possibility of different inflammatory responses in patients who develop CDAC versus others who do not. Salzman feels that “α-defensins show an additional antitoxin activity, in which HD5 is more effective; i.e. the stimulation of toxin aggregation.” Giesemann has shown that HD5, used at concentrations that normally can be found in the small intestine, is effective at causing aggregation of toxin B, thus effectively preventing the toxin’s ability to enter cells and interact with its target. These findings suggest an additional mechanism of antitoxin activity by α-defensin HD5.

This ability of HD5 to cause toxin B aggregation may provide an explanation for both the asymptomatic carriage of this pathogen and the frequency of patient relapse following antibiotic treatment, especially if the small intestine is a reservoir for C. difficile carriage in the gut. Salzman postulated that C. difficile is able to maintain colonization of the small intestine, but unable to cause colitis, because the high concentration of HD5 at this site neutralizes the secreted exotoxin.

In summary, Salzman feels that, in the small intestine, high concentrations of HD5 result in toxin B aggregation and therefore, the prevention of intoxication. While, in the large intestine, inadequate amounts of α-defensin are present to aggregate or inhibit toxin B, resulting in epithelial intoxication, inflammation, and neutrophilic infiltration.

Usually, C. difficile that transits through the large bowel will be prevented from finding a niche by the normal colonic microbiota. Yet, if the microbial ecology of the colon is disrupted, perhaps through antibiotic treatment, C. difficile can colonize the large intestine. Salzman postulates that, under these conditions, HD5 concentration is reduced by diffusion and dilution; thus, C. difficile exotoxins become free to interact with colonic enterocytes, resulting in intoxication, inflammatory responses, and infectious colitis.

Many patients are colonized with C. difficile, but have no symptoms. Perhaps C. difficile is harbored in the small intestine, where its toxic effects are well neutralized. Lawrence has claimed that about 20% of hospitalized adults are C. difficile carriers; and, in LTCFs, the carriage rate may approach 50%[62]. Although asymptomatic, these individuals shed pathogenic organisms and serve as a reservoir for environmental contamination. About 3% of healthy adults and 20%-40% of hospitalized patients are colonized with C. difficile, which in healthy persons is metabolically inactive in the spore form. Many patients have C. difficile as an asymptomatic organism in their intestine on hospital admission, and it only becomes a problem after they are treated with antibiotics, if, in fact, it ever induces symptoms. Exposure to antibiotics that disrupt the colonic microbial flora appears to be the most important risk factor for CDAD.

Asymptomatic colonized patients can act as a reservoir for the transmission of CDAD. Data, however, are limited regarding whether the treatment of these asymptomatic carriers leads to a decrease in the nosocomial transmission of C. difficile. Thirty asymptomatic C. difficile carriers were randomly assigned to one of three treatment groups: oral vancomycin 125 mg four times daily; metronidazole 500 mg orally twice daily; or placebo. Johnson et al[63] have found that nine of 10 patients receiving vancomycin became culture-negative during and immediately after treatment, compared to three of 10 receiving metronidazole and two of 10 receiving placebo. However, this decolonization was transient, as most patients became re-colonized within weeks. Thus, metronidazole does not appear to be effective for the treatment of asymptomatic carriers. In the setting of a hospital outbreak in which temporary elimination of the organism is felt necessary to reduce horizontal transmission, vancomycin may be a useful tool[63].

Riggs et al[64] have reported on molecular typing of C. difficile performed on asymptomatic carriers using pulsed-field gel electrophoresis. They found that 35 (51%) of 68 asymptomatic patients were carriers of toxigenic C. difficile, and 13 (37%) of these patients carried epidemic strains. They have also reported that 87% of isolates found in skin samples and 58% of isolates found in environmental samples were identical to concurrent isolates found in stool samples. Spores on the skin of asymptomatic patients were transferred easily to the investigators’ hands, again accounting for spread to persons in contact. This might be an explanation for the McFarland et al[65] observation that nosocomial CDAD frequently is transmitted between hospitalized patients, and that the organism often is present on the hands of hospital personnel caring for such patients. Kyne et al[66] have studied prospectively C. difficile infections in hospitalized patients who were receiving antibiotics, and identified no evidence of immune protection against repeat colonization by C. difficile. However, after colonization, there is an association between a systemic anamnestic response to toxin A, as demonstrated by increased serum levels of IgG antibody against toxin A, and asymptomatic carriage of C. difficile.

Although the antibiotics most frequently implicated in predisposition to C. difficile infection are fluoroquinolones, clindamycin, cephalosporins and penicillins, virtually all antibiotics, including metronidazole and vancomycin, can predispose to C. difficile. De Andrés et al[76] reported a case of C. difficile colitis associated with valacyclovir treatment.

The risk of CDAD in hospitalized patients receiving antibiotics may be compounded by co-existing disorders that require treatment with PPI therapy, which inhibits one’s defenses against ingested bacteria by virtually eliminating gastric acid[77]. Dial et al[78] estimated an adjusted risk ratio for C. difficile-associated disease with the current use of PPIs as 2.9 (95% CI: 2.4-3.4); and with H2-receptor antagonists, the rate ratio was 2.0 (95% CI: 1.6-2.7). The authors also uncovered an elevated rate of CDAD in patients on non-steroidal anti-inflammatory drugs (rate ratio, 1.3; 95% CI: 1.2-1.5). Thus, the consumption of drugs other than antibiotics may put one at increased risk for community-acquired C. difficile.

PPI therapy is also associated with an increased risk of recurrent C. difficile colitis. Patients receiving PPIs have been found to be 4.17 times as likely to have recurrence as their counterparts not receiving them[79]. This relationship between PPI therapy and C. difficile was elucidated by Jump who found that the survival of vegetative C. difficile in gastric contents obtained from patients receiving PPIs was also increased at a pH of > 5[80].

The incidence of severe CDAD is increasing in peripartum women. A PubMed search identified 24 recorded cases of peripartum C. difficile infection. Most patients (91%) had received prophylactic antibiotics during delivery or for treatment of bacterial infections. Two cases without known risk factors were found, by polymerase chain reaction analysis, to be infected with an epidemic and hypervirulent C. difficile strain. These cases demonstrate the need for clinicians to consider C. difficile infection in pregnant and peripartum patients with diarrhea, even if they do not have the traditional risk factors for C. difficile infection, such as antibiotic use or concurrent hospitalizations[81].

The Agency for Healthcare Research and Quality (AHRQ) is the lead US Federal agency charged with improving the quality, safety, efficiency, and effectiveness of health care. AHRQ data make clear that one of the challenges in accurately diagnosing CDAD is that it is not unusual for patients who acquire C. difficile to have multiple co-morbidities. Thus, multiple co-morbidities put patients at risk for C. difficile infection. AHRQ found that hospitalized patients with CDAD had over 10 diagnoses, versus six diagnoses among patients without CDAD[82]. According to recent AHRQ data, four out of the top 20 most common principle diagnoses observed with CDAD are infections (sepsis, pneumonia, urinary tract infection, and skin infection), where antibiotic use would be difficult to avoid[82].

Sixteen patients, representing an incidence rate of 0.16%, developed a C. difficile infection after total joint arthroplasty (TJA) at one institution. Those at risk for developing CDAD after TJA were patients with deteriorated physical status and those who had received more than one antibiotic postoperatively[83].

In addition, C. difficile is now considered to be a significant cause of diarrhea in heart transplant recipients, and the post-transplantation period is now considered one of greater risk[84]. With C. difficile infection, CDAC prior to 2000 was a rare complication in this patient group; but 38 of the 43 reported cases of CDAC in these patients occurred after 2000. Therefore, C. difficile is now also one of the most common causes of diarrhea in patients who have undergone solid organ transplantation[85]. Another group of patients at increased risk are post orthotopic liver transplant patients. Testing for C. difficile toxins among orthotopic liver transplant patients with nosocomial diarrhea revealed that 63% of samples are toxin-positive[86].

The development of life-threatening toxic megacolon secondary to CDAC now must be considered in solid organ recipients. Toxic megacolon was reported in five patients by Stelzmueller et al[85].

The risk of C. difficile infection was 14.9 cases per 1000 surgical procedures among patients who received preoperative prophylaxis (PAP) during the period 2003-2005, which is a significant increase compared with 0.7 cases per 1000 surgical procedures during the period 1999-2002 (P < 0.001). Independent risk factors associated with C. difficile infection in patients given PAP alone, were older age, the administration of cefoxitin (rather than cefazolin) alone or in combination with another drug, and the year of surgery. Thus, in the context of a large epidemic of C. difficile infection associated with the emergence of a novel strain of organism, 1.5% of patients who had received PAP as their sole antibiotic treatment developed C. difficile infection. In situations in which the only purpose of PAP is to prevent infrequent and relatively benign infections, the risks of PAP may outweigh its benefits, especially in elderly patients[87].

Unfortunately, the incidence of C. difficile infection is increasing in US surgical patients even without PAP, and infection with C. difficile is most prevalent after emergency operations and among patients who have undergone intestinal tract resections[88].

IBD patients are at greater risk than the general population for acquiring C. difficile infection[89]. Issa et al[90] performed a retrospective, observational study in IBD patients to evaluate the impact of C. difficile. They found that the rate of C. difficile infection had increased from 1.8% of IBD patients in 2004 to 4.6% in 2005 (P < 0.01). The proportion of IBD patients within the total number of C. difficile infections at their institution increased from 7% in 2004 to 16% in 2005 (P < 0.01). In 2005, IBD colonic involvement was found in the vast majority (91%) of C. difficile-infected patients, a clear majority (76%) had contracted infection as an outpatient, and antibiotic exposure was identified in 61% of IBD patients with C. difficile infection. Over the period 2004-2005, more than half of the infected IBD patients required hospitalization, and 20% required colectomy. Univariate and multivariate analyses identified maintenance immunomodulator use and colonic involvement as independent risk factors for C. difficile infection in IBD. The authors also reported a nationwide doubling in the rate of C. difficile infection among hospitalized UC patients between 1998 and 2004. The pathologic/endoscopic features of pseudomembranous colitis CDAC varies as a spectrum, with some patients exhibiting only mild inflammatory changes confined to the superficial epithelium, and typical pseudomembranes and crypt abscesses may not be present. The more severe cases demonstrate marked mucin secretion, and more intense inflammation. Intense necrosis of the full thickness of the mucosa, with a confluent pseudo-membrane, can become more prominent as disease severity increases.

The association between IBD and C. difficile may be due to a variety of factors, including antibiotic use for treatment of other gastrointestinal pathogens and frequent hospitalizations for the management of IBD flares. Many of these patients are taking immunosuppressive medications that may confer additional risk of C. difficile infection. C. difficile, and specifically its toxins, have been implicated as a risk factor for the exacerbation of the inflammatory process in up to 5% of patients with ulcerative colitis or Crohn’s disease. A severe clinical course may result from C. difficile infection superimposed on IBD, including the precipitation of toxic colitis and toxic megacolon.

CDAC in patients with IBD carries a higher mortality than in patients with C. difficile without underlying IBD. On multivariate analysis, patients in the C. difficile-IBD group had a four times greater mortality than patients admitted to hospital for IBD alone (AOR = 4.7, 95% CI: 2.9 to 7.9) or C. difficile alone (AOR = 2.2, 95% CI: 1.4 to 3.4), and stayed in the hospital for 3 d longer (95% CI: 2.3 to 3.7 d). Significantly higher mortality, endoscopy and surgery rates were found in patients with ulcerative colitis compared with Crohn’s disease (P < 0.05) who had associated C. difficile[91]. The median times from admission to a positive C. difficile test result for non-IBD was much longer than in Crohn’s disease and ulcerative colitis patients (4.0, 0.8, and 0.5 d, respectively). C. difficile infections in IBD are confirmed predominantly within 48 h of admission, suggesting most were acquired before hospitalization. CDAD rates approximately doubled in Crohn’s disease (9.5 to 22.3/1000 admissions) and tripled in ulcerative colitis (18.4 to 57.6/1000). Length of stay was similar among the groups. For all years combined, the adjusted odds ratios for CDAD in all IBD, Crohn’s disease, and ulcerative colitis admissions were 2.9 (95% CI: 2.1-4.1), 2.1 (1.3-3.4), and 4.0 (2.4-6.6), respectively[92].

Patients with severe C. difficile infection, especially IBD patients, require prompt diagnosis and management, since failure to diagnose the infection can lead to inappropriate treatment with glucocorticoids or immunosuppressive therapy. Furthermore, C. difficile may be difficult to distinguish from an IBD relapse, given the similar symptoms of diarrhea, abdominal pain, and low-grade fever. Thus, a high index of suspicion is required when evaluating IBD patients with apparent flares, especially those who recently have received antibiotics and/or have been hospitalized.

Thus, speedy diagnosis largely requires the use of laboratory tools, since endoscopy may not be helpful early, because IBD patients may not develop pseudomembranes. Given the underlying colonic pathology, patients with IBD who develop C. difficile colitis frequently require colectomy (20 percent in one series)[90].

Lower endoscopy is a useful tool for the diagnosis of C. difficile. This is especially when: (1) there is a high level of clinical suspicion for C. difficile, despite a negative laboratory assay; (2) prompt C. difficile diagnosis is needed before laboratory results can be obtained; (3) C. difficile infection fails to respond to antibiotic therapy; or (4) when there is an atypical disease presentation, and C. difficile is suspected, as with ileus, acute abdomen, leukocytosis or diarrhea.

Endoscopy is not indicated in patients with classic clinical findings and a positive stool toxin assay. Conversely, endoscopy may be contra-indicated, especially in the setting of fulminant colitis, due to the risk of perforation.

Endoscopic findings: Pseudomembranes are pathognomonic for CDAC, but are not found in all areas of the colon, even in severe cases; thus, findings may be patchy. Pseudomembranes may be absent in the rectosigmoid area, but may be visualized more proximally with colonoscopy. This is true in patients with co-existing IBD. Pseudomembranes are raised yellow or off-white plaques, up to 2 cm in diameter, which are randomly scattered over the colorectal mucosa with normal intervening mucosa, and that cannot be removed by lavage. The pseudomembranes form when C. difficile toxin-induced cytoskeleton disruption causes shallow ulcerations on the intestinal mucosal surface. It is postulated that ulcer formation allows for the release of serum proteins, mucus, and inflammatory cells, which appear grossly on the colorectal mucosal surface as pseudomembranes. Light and scanning electron microscopy after exposure to either of the C. difficile toxins reveal patchy damage and exfoliation of superficial epithelial cells, while crypt epithelium remains intact. Fluorescent microscopy of phalloidin-stained sections shows that both toxins cause the disruption and condensation of cellular F-actin[96].

Other colonic mucosal findings include bowel-wall edema, erythema, friability, and inflammation, with or without pseudomembranes. This manifests on the abdominal CT scan as thickening of the colonic wall.

Colonoscopic findings among 16 patients with histologically-proven antibiotic-associated PMC or CDAC were described by Seppälä et al[97]. Pseudomembranes were found in only five of 16 (31%) patients by sigmoidoscopy, but were found in 11 of 13 patients (85%) in whom colonoscopy also was performed. These findings suggest the importance of colonoscopy in the early diagnosis of CDAC, because the typical endoscopic changes of pseudomembranes are limited to the colon above the rectosigmoid area in most patients. Consequently, colonoscopy should be performed, instead of sigmoidoscopy, at least in clinically suspected CDAC cases[98].

C. difficile colitis is usually associated with a mild/moderate course, but may progress to fulminant colitis. Fulminant colitis develops in 3%-8% of patients. The manifestations of fulminant colitis typically include severe lower quadrant or diffuse abdominal pain, diarrhea, abdominal distention, fever, hypovolemia, lactic acidosis, and marked leukocytosis (up to 40 000 white blood cells/microL or higher). Diarrhea may be less prominent in patients with prolonged ileus, due to pooling of secretions in the dilated, atonic colon. Other potential complications of fulminant colitis include toxic megacolon and bowel perforation[73].

Toxic megacolon is a clinical diagnosis based upon the finding of colonic dilatation (> 7 cm in its greatest diameter) accompanied by severe systemic toxicity. Abdominal plain films also may demonstrate small-bowel dilatation, air-fluid levels (mimicking an intestinal obstruction or ischemia), and ‘thumb printing’ (scalloping of the bowel wall) due to submucosal edema. Toxic megacolon may be complicated by bowel perforation.

This latter complication presents with abdominal rigidity, involuntary guarding, diminished bowel sounds, rebound tenderness, and severe localized tenderness in the left or right lower quadrants. Abdominal radiographs may demonstrate free intra-abdominal air. Thus, patients with toxic megacolon must be followed with daily upright abdominal X-rays to ascertain if perforation has occurred. Patients with toxic megacolon should be evaluated for surgical resection. Once fulminant colitis is diagnosed, subtotal colectomy with ileostomy usually is required. In these patients who develop a marked leukocytosis or bandemia, surgery is advisable, because the leukocytosis frequently precedes hypotension. The requirement for vasopressor therapy carries a poor prognosis, according to Shen et al[99].

Lamontagne et al[100] has documented that emergency colectomy reduces mortality in patients with fulminant CDAD. The independent predictors of 30-d mortality in their study were leukocytosis ≥ 50 × 10⁹/L (AOR, 18.6; 95% CI: 3.7-94.7); serum lactate ≥ 5 mmol/L (AOR, 12.4; 95% CI: 2.4-63.7); age ≥ 75 years (AOR, 6.5; 95% CI: 1.7-24.3); immunosuppression (AOR, 7.9; 95% CI: 2.3-27.2); and shock requiring vasopressor therapy (AOR, 3.4; 95% CI: 1.3-8.7). After adjusting for these confounders, patients who had an emergency colectomy were less likely to die than those treated medically. Colectomy also seemed more beneficial in patients 65 years or older; in those who were immune-competent; and those with leukocytosis ≥ 20 × 10⁹/L or a serum lactate level between 2.2 and 4.9 mmol/L.

Small-bowel involvement with C. difficile enteritis is unusual[101]. Potential risk factors for small-bowel involvement with C. difficile enteritis include prior gastrointestinal surgery (including colonic resection) and advanced age[102]. Such patients may present with ileitis and high ileostomy output and may be at increased risk for fulminant disease.

Small-bowel involvement with C. difficile infection enteritis has been described increasingly since 2000. Usually, this occurs in patients with a history of a prior colectomy or total procto-colectomy for severe and extensive IBD. The ileal mucosa appears to be at increased risk for inflammatory disease in the specific subset of patients who have undergone a prior colectomy[67]. Serious post-colectomy concerns, like severe ileostomy dysfunction with high ileostomy volumes and marked diarrhea, have been known to occur after pan-procto-colectomy and restorative ileo-anal pouch formation. They are almost always due to a non-C. difficile enteritis. This non-CDAD post colectomy enteritis can be life threatening; fortunately it is steroid/ immunosuppressive responsive, according to Gooding et al[103]. This picture can be mimicked by C. difficile infection.

Lundeen et al[104] reported that high ileostomy volumes may result from C. difficile enteritis in patients who have undergone colectomy for ulcerative colitis. All of the ileostomy output was positive for C. difficile toxins. These patients responded to metronidazole and/or vancomycin, in contrast to subjects with the former, non-CDAD entity.

Refractory or treatment-resistant pouchitis also may occur with C. difficile infection[105]. C. difficile infection involving ileal pouch-anal anastomosis is common, and occurs with or without the previous receipt of antibiotics[99]. Diagnosing recurrent C. difficile infection can be difficult in this group of patients, especially in the 20% without diarrhea.

All health care facilities must develop rapid communica-tion between the laboratory and the treating physician. At the Mayo Medical School, the time between electronic medical record reporting of a positive result for a test for C. difficile toxin in stool and the ordering of antimicrobial therapy was compared during consecutive periods when results were not telephoned (n = 274) and when results were telephoned (n = 90) to the clinical service[106]. The mean times to the ordering of antimicrobial therapy were 11.9 and 3.6 h, respectively (P < 0.001). The clinical implications of this 8-h delay may be important, especially in patients with severe disease. Early recognition of CDAD caused by NAP1/027, followed by the initiation of rapid treatment, can help to prevent complications and further spread of the bacterium[107].

Current laboratory testing lacks a single assay that is sensitive, specific, and rapid. Peterson et al[108] used clinical criteria that required at least three loose stools in one day, as part of the reference standard for a positive test result supporting CDAD (Table 2). They found that real-time PCR and anaerobic culture assays were significantly more sensitive than the enzyme immunoassay (P < 0.01 to P < 0.05). Real-time PCR has an assay turnaround time of < 4 h, and is both more sensitive than, and as rapid as enzyme immunoassay. They feel that it is a feasible laboratory option to replace enzyme immunoassay for toxigenic C. difficile detection in clinical practice, as well as for use during the development of new therapeutic agents.

Tests for the presence of C. difficile and its toxins are imperfect, and false positives and false negatives are not uncommon. McFarland[30] found that false-negative results occur in 29%-56% of cases. False-negative results may occur when specimens are not promptly tested or not kept refrigerated until testing is performed. Also, there is a relatively high false-negative rate, due to the fact that 100-1000 pg of toxin must be present for an EIA test to be positive. Utilizing up to three serial EIA tests may increase the diagnostic yield by as much as 10 percent, if the initial test is negative. If CDAD is suspected, despite negative initial testing, submission of multiple specimens and verifying that the laboratory is testing for both the A and B toxins is mandatory (Table 3).

Enzyme immunoassays are labor-intensive tests, requiring several hours of technician time and an assay reader. The batching of specimens increases cost efficiency, but may delay the reporting of results, especially if tests are not done every day. Rapid enzyme immunoassay is more costly for each test performed but, for laboratories that process only occasional samples, it appears to provide prompt, reliable, and cost-effective results.

Enzyme immunoassay rapid cards have been evaluated, in terms of their ability to detect C. difficile toxins A and B. For one such card, the EIAPrem, the positive predictive value (PPV) was 75/85 samples (88.2%; CI: 79% to 94%) and the negative predictive value (NPV) was 360/361 samples (99.7%; CI: 98% to 99%). For a review of all card performances, see references[109–112].

Killgore et al[113] compared the results of analyses done with seven C. difficile typing techniques: multi-locus variable-number tandem-repeat analysis (MLVA); amplified fragment length polymorphism; surface layer protein A gene sequence typing; PCR-ribotyping; restriction endonuclease analysis (REA); multi-locus sequence typing; and pulsed-field gel electrophoresis (PFGE). All techniques appeared to be capable of detecting outbreak strains; but only REA and MLVA exhibited sufficient discrimination to distinguish strains from different outbreaks.

Comparison of four enzyme immunoassays (Bartels Prima System C. difficile Toxin A EIA, Cambridge Biotech Cytoclone A+B EIA, Meridian Diagnostics Premier C. difficile Toxin A EIA, and TechLab C. difficile Tox-A Test EIA) found that, although enzyme immunoassays were less sensitive than cytotoxin assay, they provide same-day results and may be useful in laboratories without tissue culture facilities[114].

ELISA Toxin A+B is a reliable method with 100% specificity and sensitivity in the rapid diagnosis of

C. difficile. Its results can be utilized until culture results are obtained. The specificity of the Toxin A latex test is 100%; however, its use alone as a primary rapid diagnostic test is not recommended, because of its low (30.7%) sensitivity. This was shown when all of the culture positive samples underwent testing by ELISA Toxin A+B method and were found to be 100% positive, but only four of these positive culture samples (30.7%) yielded positive results with the Toxin A latex test[115].

Overall, the new-generation assays still are less sensitive than the cytotoxin assay; however, their advantages are that they provide same-day results; they can be used as a screening test; and they may be useful in laboratories without tissue-culture facilities. Results from a study by Vanpoucke et al[111] could not recommend one single assay over the other for the diagnosis of CDAD.

Therefore, the cytoxin assay test (CYTA) is highly sensitive and specific, but it is difficult to perform, and results are not available for 24-48 h[15]. What further complicates efforts to determine if toxin was present on admission is that C. difficile toxin is very unstable. The toxin degrades at room temperature and may be undetectable within 2 h after collection of a stool specimen. Given the cost and complexity of culture and cytotoxicity assays, most laboratories rely on tests for toxin A detection only. Moreover, enzyme immunoassays generally are available at lower cost and provide more rapid results, usually within 4 h. Their sensitivity generally ranges from 60% to 90%, and specificity from 75% to 100%. Testing of a single diarrheal stool generally is sufficient to make the diagnosis of CDAD; but unfortunately, doing so misses a substantial proportion of cases. Therefore, testing only should be performed on three loose stool specimens.

The cytotoxin assay test, though the ‘gold standard’ for assaying C. difficile toxins A and B, is labor-intensive, requires tissue-cultured cells and an inverted microscope, and needs overnight incubation before results can be read.

Twenty percent of C. difficile infection patients relapse, despite adequate therapy. Risk factors for relapse are presented in Table 6. Diagnosing recurrent C. difficile infection can be difficult, especially in the 20% without diarrhea. The usual treatment for recurrent C. difficile infection is a repeat course of metronidazole, unless the patient has severe disease. Tapered and pulsed dosing schedules of vancomycin have been investigated for the treatment of C. difficile infection that recurs after an initial course of vancomycin (Table 7). An example of an oral vancomycin taper schedule is as follows: 125 mg qid× 10-14 d; 125 mg bid× 7 d; 125 mg daily × 7 d; 125 mg once every 2 d × 8 d; and 125 mg once every 3 d × 15 d[126]. The treatment of recurrent C. difficile infection with various vancomycin daily doses (2 g/d, 1 g/d, and 500 mg/d) and administration schedules (daily vancomycin followed by tapered or pulsed dose vancomycin therapy) was reported by McFarland et al[123]. They found that tapered and pulsed dosing schedules of vancomycin result in significantly better C. difficile infection cure rates than traditional vancomycin dosing.

Wenisch et al[127] conducted a prospective, randomized study to compare the efficacy of the oral drugs fusidic acid, metronidazole, vancomycin, and teicoplanin in the treatment of CDAD. Treatment resulted in clinical cure greater than 90% with all the agents: 94% vancomycin, 96% teicoplanin, 93% fusidic acid, and 94% metronidazole. However, recurrent clinical symptoms occurred in 16% of patients treated with vancomycin or metronidazole, 7% of those treated with teicoplanin, and 28% of those treated with fusidic acid. There was asymptomatic carriage of C. difficile toxin in 13% of patients treated with vancomycin, 16% with metronidazole, 4% with teicoplanin and 24% with fusidic acid. No adverse effects related to therapy were observed with vancomycin or teicoplanin. Considering the costs of treatment, their findings suggest that metronidazole is the drug of choice for CDAD, and that glycopeptides should be reserved for patients who cannot tolerate metronidazole or who do not respond to treatment with this drug.

Studies on probiotics for C. difficile infection have been inconclusive and conflicting, with respect to treatment benefit. Nonetheless, the use of probiotics is becoming more widespread.

Pillai and Nelson conducted a meta-analysis to assess the potential therapeutic effects of probiotics for C. difficile infection[128]. Randomized, prospective studies (1966-2007) using probiotics alone or in conjunction with conventional antibiotics for the treatment of documented C. difficile colitis were eligible for inclusion. Ultimately, four studies met the inclusion criteria and were included in the review. The four studies examined the use of probiotics in conjunction with conventional antibiotics (vancomycin or metronidazole) for the treatment of recurrence or the initial episode of C. difficile colitis in adults. All of the studies were small and had methodological issues. A statistically-significant benefit of probiotics combined with antibiotics was detected in only one study. The authors concluded that, overall, there is insufficient evidence to recommend probiotic therapy as an adjunct to antibiotic therapy for C. difficile colitis. There also is no evidence to support the use of probiotics alone in the treatment of C. difficile colitis.

In 1994, McFarland et al[129] reported that patients receiving Saccharomyces boulardii were significantly less likely than patients receiving placebo to experience a recurrence of C. difficile diarrhea (RR 0.59; 95% CI: 0.35 to 0.98). Consequently, in a later meta-analysis, he compared the efficacy of probiotics for the prevention of AAD and the treatment of CDAD. Across 25 randomized controlled trials (RCTs), probiotics significantly reduced the relative risk of AAD (RR 0.43, 95% CI: 0.31 to 0.58, P < 0.001)[130]. Across six randomized trials, probiotics had significant efficacy for CDAD (RR 0.59, 95% CI: 0.41 to 0.85, P = 0.005).

This time, McFarland et al[129] concluded that a variety of different types of probiotic show promise as effective therapies for these two diseases. Again using meta-analysis, three types of probiotics (S. boulardii, Lactobacillus rhamnosus GG, and probiotic mixtures) were found to significantly reduce the development of AAD. Only S. boulardii was effective for CDAD.

Treatment of recurrent C. difficile infection with high-dose vancomycin plus S. boulardii is the only treatment combination that has been evaluated in a prospective, randomized, controlled trial and found to generate a significant trend toward reduced recurrent C. difficile infection[131]. Lactobacillus spp. have been evaluated for use in recurrent C. difficile infection, but data on regimens containing these organisms are poorly derived and conflicting. Fungemia with its administration has been reported in immunocompromised hosts. Therefore, its use is not appropriate in this group[132].

Relapse of C. difficile occurs in 10%-25% of patients treated with metronidazole or vancomycin. Furthermore, multiple relapses may occur in the same patient. An alternative approach to patients with recurrent CDAD involves the administration of the entire fecal flora from a healthy individual, which is referred to as fecal bacteriotherapy. Borody et al[133] reviewed 84 fecal transplantation therapies for severe cases of relapsing, or recurrent C. difficile infection (via various routes of administration). They found that 80% resulted in a good clinical response, resolution, or cure[133]. A review of eight reports on the infusion of feces or fecal bacteria revealed an optimistic cure rate, without recurrence in most patients[134]. In a study involving 18 patients treated with healthy donor stools via a nasogastric tube, 15 patients were recurrence-free at 90 d (two died of unrelated causes and one experienced recurrence)[135]. The patients described in these reports[133136–144] included those with symptomatic relapse after receiving multiple courses of antibiotics; e.g. vancomycin, and/or metronidazole, and/or rifampicin together with cholestyramine. Case series have suggested a clinical benefit of fecal bacteriotherapy in patients with severe or recurrent CDAD who have failed to respond to standard approaches. Although the data are limited to case series, fecal bacteriotherapy has been used successfully to treat relapsing C. difficile infection. The precise mechanisms for the benefits of fecal bacteriotherapy are unclear. The reappearance of Bacteroides species after treatment suggests that Bacteroides species may be involved in the restoration of the presumably antibiotic-damaged flora in the colon.

Successful treatment with two or more fecal enemas has been described in other reports, according to Borody, Leis, & Gerald Pang (http://www.up-to-date.com2008), involving a total of 23 patients with PMC who were refractory to antibiotic therapy, or who had experienced multiple relapses. In one study of 16 patients with severe, refractory disease treated over an 18-year period, 13 responded dramatically with decreases in diarrhea, temperature and leukocytosis. In a report describing nine patients, the single administration of a fecal enema (5-10 gm homogenized stool in pasteurized cow’s milk) was effective in seven. In another case report, according to Borody, Leis, & Gerald Pang (http://www.up-to-date.com), the one-time administration of bacteriotherapy was effective when 500 mL of fecal infusion in saline was delivered throughout the colon via a colonoscope. The authors hypothesized that the greater area of re-colonization by fecal bacteria created a greater capacity to inhibit spore formation proximal to the splenic flexure. The use of the colonoscope to deliver fecal bacteria has an added theoretical advantage of permitting delivery of the active flora components to the distal small bowel, where C. difficile can reside. In addition, the colonoscope may permit the proximal delivery of flora in patients with a dilated colon, although colonoscopy must be performed extremely cautiously in this setting, because of the risk of perforation. One of the current authors (JSB) has utilized this modality with similar results.

Aas et al[135] reported on 18 subjects who received donor stool by nasogastric tube for recurrent C. difficile infection over a 9-year period at a single institution. During the period between the initial diagnosis of C. difficile colitis and the stool treatments, the 18 subjects received a total of 64 courses of antimicrobials (range, 2-7 courses; median, three courses). During the first 90 d after receipt of treatment with stool, two patients died of unrelated illnesses. Only one of the 16 survivors experienced a single recurrence of C. difficile colitis over the 90-d follow-up. No adverse effects associated with stool treatment were observed. Patients with recurrent C. difficile colitis may benefit from the introduction of stool from healthy donors via a nasogastric tube.

Lund-Tønnesen et al[140] reported on 18 patients with CDAD who were given homologous feces from one healthy donor. In 17 patients, feces were instilled via a colonoscope, and in one patient via a gastric stoma. Fifteen patients were clinically cured, and no relapses were observed; however, it is important to note that three patients with severe colitis did not respond to the treatment.

In recalcitrant, recurrent C. difficile infection, one should attempt initially to use probiotics that have been shown to be effective in published studies. Subsequently, in patients who remain seriously ill from recurrent C. difficile infection, fecal bacteriotherapy may be used when other approaches have been unsuccessful[133145]. The above-mentioned study by Lund-Tønnesen et al[140], in which three patients with severe colitis did not respond to the treatment, while only the remaining less-severely ill patients were clinically cured, and no relapses were observed may indicate that this may serve as rescue therapy for patients with recurrent C. difficile. But its role in patients with severe C. difficile infection remains unproven.

Suggested protocol for fecal bacteriotherapy: Barody’s protocol is as follows. (a) Donor stool and blood are screened for pathogens and viruses before infusion. CBC, serological testing for hepatitis A, B, and C; HIV-1 and HIV-2 and syphilis, stool culture for enteric bacterial pathogens, and light microscopy examination of stool sample for parasites and ova are performed. (b) The donor is clinically well, with the passage of normal, daily stools, and has had no intake of antibiotics for the last 6 mo. (c) The donor should not be a close relative, living in the same household such as a husband or child, theoretically to avoid use of flora from a silent carrier of the same pathogen. (d) The recipient is evaluated for HIV and hepatitis markers to avoid future questions about transmission. (e) Oral vancomycin (500 mg twice daily for 7 d) is administered, and then followed by a single oral lavage with 3-4 L of polyethylene glycol with electrolyte purgative (such as GoLYTELY). (f) Although the lavage is skipped in patients too ill to tolerate it, vancomycin pretreatment is used, whenever possible. (g) 200-300 gm of donor stool suspended in 200-300 mL of sterile normal saline (homogenized briefly in a kitchen blender to a liquid consistency) is administered via an enema within 10 min of preparation, and this is repeated daily for 5 d. (h) Initial infusion may be filtered and infused via colonoscopy, preferably into the terminal ileum to address known ileal presence of C. difficile. (i) At least five consecutive days of rectal enemas are administered, using donor stools. (j) The enema should be retained for at least 6 h (loperamide pretreatment may help), followed by a high-fiber meal and overall diet. (k) Although some patients are unable to retain the enema initially for prolonged periods, it appears that coating of the mucosa by the infusate is adequate. (l) Adverse effects have been transient and mild, and have consisted primarily of abdominal gurgling, gas and borborygymi-expected post-enema symptoms. Recurrence has not been observed with follow-up of 1-3 years in most patients, even though a number of patients subsequently have required antibiotics for unrelated infections.

In summary, patients who develop a second episode of C. difficile infection after successful treatment of the first episode may be at increased risk for developing complications. Although different drugs and regimens have been used, vancomycin may be the best option; and the combination of high-dose vancomycin plus S. boulardii is the only treatment combination that has been evaluated in a prospective, randomized, controlled trial to demonstrate a significant trend toward reduced recurrent C. difficile infection. Fecal bacteriotherapy seems promising and is undergoing further testing at this time

Taylor et al[156] examined the safety and pharmacokinetics of a novel neutralizing human monoclonal antibody against C. difficile toxin A (CDA1) in 30 healthy adults whose median age was 27.5 years. While there were no serious adverse events related to its use, 21 of 30 reported non-serious adverse events were possibly related to CDA1. These included transient blood pressure changes requiring no treatment, nasal congestion, headache, abdominal cramps, nausea, and self-limited diarrhea. The authors concluded that, at least in healthy subjects, the administration of CDA1 as a single intravenous infusion is safe and well tolerated.

The importance of toxin production in the patho-physiology of C. difficile diarrhea has prompted consideration of anion-binding resins as a possible alternative to antimicrobial therapy. An advantage of resin therapy is that the bowel flora is not altered, as occurs with antibiotics (e.g. vancomycin or metronidazole), which may allow for more rapid reconstitution of normal colonic flora. Anion-exchange resins bind vancomycin as well as toxins; thus, the resin must be taken at least 2 h or 3 h apart from the vancomycin. Suggested regimens are colestipol (5 g every 12 h) or cholestyramine (4 g three or four times daily) for 1-2 wk, usually in conjunction with vancomycin.

Tolevamer, a novel toxin-binding polymer, has been developed to ameliorate C. difficile-associated disease without adversely affecting normal flora. Tolevamer has been tested for its ability to neutralize clostridial toxins produced by the epidemic BI/027 strains, thereby preventing toxin-mediated tissue culture cell rounding. The titers of toxin-containing C. difficile culture supernatants were determined using confluent cell monolayers, and then the supernatants were used in assays containing dilutions of tolevamer to determine the lowest concentration of drug that prevented ≥ 90% cytotoxicity. Tolevamer neutralized toxins in the supernatants of all C. difficile strains tested. Specific antibodies against the large clostridial toxins TcdA and TcdB also neutralized the cytopathic effect, suggesting that tolevamer specifically neutralizes these toxins, and that the binary toxin (whose genes are carried by the BI/027 strains) is not a significant source of cytopathology against tissue culture cells in vitro[157].

However, tolevamer is not FDA approved or commercially available, to date. Castanospermine has been identified as an inhibitor of the Rho/Ras-glucosylating Clostridium sordellii lethal toxin and C. difficile toxin B. Microinjection of castanospermine into embryonic bovine lung cells prevents the cytotoxic effects of toxins. The inhibitor binds in a conformation that brings its four hydroxyl groups and its N atom almost exactly into the positions of the four hydroxyls and the ring oxygen of the glucosyl moiety of UDP-glucose, respectively[158]. It is in its early stage of development.

Testing the feasibility of active vaccination against C. difficile and its toxins in high-risk individuals currently is ongoing[159]. C. difficile toxoid vaccine has induced immune responses to toxins A and B in patients with CDAD, and has been associated with resolution of recurrent diarrhea. This parenteral C. difficile vaccine, which contains toxoid A and toxoid B, has been reported to be safe and immunogenic in healthy volunteers. Three patients with multiple episodes of recurrent CDAD were vaccinated. Two of the three exhibited an increase in serum IgG antitoxin A antibodies (three- and four-fold increases), and in serum IgG antitoxin B antibodies (52 and 20-fold). Both individuals also developed cytotoxin-neutralizing activity against toxins A and B. Prior to vaccination, the subjects had required nearly continuous treatment with oral vancomycin for 7, 9, and 22 mo, respectively, to treat recurrent episodes of CDAD. After vaccination, all three subjects discontinued treatment with oral vancomycin without any further recurrence. Thus, C. difficile toxoid vaccine induced immune responses to toxins A and B in patients with CDAD, and was associated with resolution of recurrent diarrhea.

Vaccination with a partially-purified preparation of inactivated toxins A and B is also undergoing current study. Several studies have shown that the humoral immune response of the host to C. difficile toxins A and B influences the clinical course of CDAD, as well as the risk of relapse[160–162].

Another vaccine, containing toxoids A and B, has been shown to induce adequate antibody responses in healthy volunteers[163]. The efficacy of this vaccine subsequently was evaluated in an open-label study involving three patients with recurrent C. difficile colitis[159]. Following four intramuscular inoculations over an 8-wk period, all three patients discontinued antibiotic treatment without further recurrence over 6 mo of follow-up.

A retrospective review was performed on 264 C. difficile toxin-positive patients (November 2003-January 2005), which documented 14 patients with severe, refractory, recurrent C. difficile diarrhea who were treated with intravenous immunoglobulin (Flebogamma, 150-400 mg/kg)[164]. Patients received a median of three (range, 1-5 g/L) courses of vancomycin or metronidazole before receiving intravenous immunoglobulin. All had hypoalbuminemia (median, 22 g/L; range, 18-33 g/L) and raised C-reactive protein (median, 47 mg/L; range, 25-255 g/L) at the time of infusion. The median white cell count was 15.3 × 10⁹/L (range, 4-24 g/L). Eight patients had evidence of pancolitis on abdominal imaging, suggesting severe C. difficile diarrhea. All patients tolerated intravenous immunoglobulin without side effects. Nine (64%) responded with bowel habits normalizing in a median of 10 (range, 2-26) d; one patient received two doses. One patient had a partial response from two doses, but died 2 mo later after a recurrence. Thus, intravenous immunoglobulin may be effective for severe, refractory, or recurrent C. difficile diarrhea after failed conventional treatment.

In patients with C. difficile colitis, a progressive, systemic inflammatory state may develop that is unresponsive to medical therapy; some cases ultimately will progress to colectomy or death. C. difficile colitis is a significant and increasingly common cause of death. Surgical treatment of C. difficile colitis has a high death rate once the fulminant expression of the disease is present[165]. These authors reviewed 2334 hospitalized patients with C. difficile colitis from January 1989 to December 2000. In the setting of CDAD before the predominance of the hypervirulent strain, 64 patients died or underwent colectomy for pathology-proven C. difficile colitis. Unfortunately, those patients who underwent colectomy for C. difficile colitis had an overall death rate of 57%. Significant predictors of death after colectomy were preoperative vasopressor requirements and older age.

Fulminant C. difficile colitis is associated with a high mortality rate. As in the former study, Hall et al[166] found that the development of a vasopressor requirement and the need for intubation are ominous signs which should lead to rapid surgical intervention. From 1998 to 2006, they studied a total of 3237 consecutive patients with C. difficile cytotoxin-positive stool samples. 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. They found that 36 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 (OR: 7.15; 95% CI: 1.28-39.8 and OR: 6.0; 95% CI: 1.08-33, respectively). Patients who had a recent surgical procedure experienced a lower in-hospital mortality rate (OR: 0.11; 95% CI: 0.02-0.52).

In the setting of CDAD due to the hypervirulent strain, some patients have progressed from severe disease to death in less than 48 h. Emergency colectomy has prevented mortality in some patients with fulminant CDAD. The decision to perform an emergency colectomy remains largely empirical[100]. In a retrospective observational cohort study of 165 cases of CDAD, among those patients who required ICU admission or prolongation of ICU stay between January 2003 and June 2005 at two tertiary care hospitals in Quebec, 53% died within 30 d of ICU admission, and almost half (44%) within 48 h of ICU admission. The independent predictors of 30-d mortality were: leukocytosis ≥ 50 ×10⁹/L (AOR: 18.6; 95% CI: 3.7-94.7), lactate ≥ 5 mmol/L (AOR: 12.4; 95% CI: 2.4-63.7), age ≥ 75 years (AOR: 6.5; 95% CI: 1.7-24.3), immunosuppression (AOR: 7.9; 95% CI: 2.3-27.2) and shock requiring vasopressors (AOR: 3.4; 95% CI: 1.3-8.7). After adjustment for these confounders, patients who had an emergency colectomy were less likely to die (AOR: 0.22; 95% CI: 0.07-0.67, P = 0.008) than those treated medically. Surgical intervention is indicated in the setting of peritoneal signs, severe ileus, or toxic megacolon; but colectomy also seems more beneficial in patients aged 65 years or older, in the immune competent, and in those with a leukocytosis ≥ 20 × 10⁹/L or serum lactate between 2.2 and 4.9 mmol/L.

The standard of care for patients undergoing emergency surgical intervention for CDAD is a total colectomy (with preservation of the rectum) and ileostomy, since primary anastomosis is not feasible acutely due to the pancolitis associated with severe disease. However, after colonic inflammation has subsided, closure of the ileostomy and ileorectal anastomosis can be performed (Table 8).

We must address environmental reservoirs to help to limit transmission. C. difficile has been cultured not only from patient bathrooms and bedpans, but from stethoscopes, blood pressure cuffs, and hospital furniture. The initial step is identifying possible C. difficile patients, especially in long-term care facilities (LTCFs). Quinn et al[168] determined that only 111 facilities (42.2%) had a protocol to identify residents with C. difficile infection, and most (77.5%) did not test for C. difficile unless a resident had severe diarrhea. Only 58.5% of the facilities placed residents with C. difficile infection in private rooms, and 60.9% cohorted residents infected with C. difficile with other residents with C. difficile colonization or infection. Only 66 facilities (25.1%) had a program to control the use of antimicrobial agents.

Findings suggest that asymptomatic carriers of epidemic and non-epidemic C. difficile strains have the potential to contribute significantly to disease transmission in long-term care facilities. Thirty-five (51%) of 68 asymptomatic patients were carriers of toxigenic C. difficile, and 13 (37%) of these patients carried epidemic strains. Compared with non-carriers, asymptomatic carriers had higher percentages of skin (61% vs 19%; P = 0.001) and environmental contamination (59% vs 24%; P = 0.004). Eighty-seven percent of isolates found in skin samples and 58% of isolates found in environmental samples were identical to concurrent isolates found in stool samples. Spores on the skin of asymptomatic patients were transferred easily to investigators’ hands. Previous C. difficile-associated disease (P < 0.001) and previous antibiotic use (P = 0.017) were associated with asymptomatic carriage, and the combination of these two variables was predictive of asymptomatic carriage (sensitivity, 77%; specificity, 58%; PPV, 66%; NPV, 70%)[64].

In a prospective study of 27 patients with C. difficile-associated disease, it was found that C. difficile frequently contaminated multiple skin sites, including groin, chest, abdomen, forearms, and hands, and was easily acquired on investigators’ hands. Skin contamination often persisted on patients’ chest and abdomen after resolution of diarrhea. Thus, skin contact of the patient by a health-care worker is a means of C. difficile transmission[169].

It is important to emphasize that asymptomatic fecal excretion of C. difficile is transient in most patients, and treatment with metronidazole is not effective. Although treatment with vancomycin is temporarily effective in asymptomatic carriers, it is also associated with a significantly higher rate of C. difficile carriage 2 mo after treatment and, therefore, is not recommended[63].

An increase in hospital-acquired C. difficile infection rate was found at the University of Pittsburg Medical Center. A comprehensive C. difficile infection control ‘bundle’ was implemented by hospital personnel to control the outbreak of C. difficile infection. This C. difficile infection control bundle consisted of education, increased and early case-finding, expanded infection-control measures, the development of a C. difficile infection management team, and antimicrobial management. Process measures, antimicrobial usage, and hospital-acquired C. difficile infection rates were analyzed, and C. difficile infection isolates were typed. The rates of compliance with hand hygiene and isolation were 75% and 68%, respectively.

The C. difficile infection management team evaluated a mean 31 patients per month (11% were evaluated for moderate or severe disease). The use of antimicrobial therapy associated with increased C. difficile infection risk decreased by 41% during the period 2003-2005. The aggregate rate of C. difficile infection during the period 2001-2006 decreased to 4.8 infections per 1000 HDs; and, by 2006, it had decreased to 3.0 infections per 1000 HDs, a rate reduction of 71%. During the period 2000-2001, the proportion of severe C. difficile infection cases peaked at 9.4% (37 of 393 C. difficile infections were severe); this rate decreased to 3.1% in 2002 and further decreased to 1.0% in 2006, a 78% overall reduction. In 2005, 13% of C. difficile isolates were type BI (20% were hospital acquired), which represented a significant reduction from 2001. These authors concluded that the outbreak of C. difficile infection with the BI strain in hospital was controlled after implementing this infection control ‘bundle’. Thus, early identification, coupled with appropriate control measures, reduces the rate of C. difficile

infection and the frequency of adverse events. However, it requires a multipronged approach.

Methods of contact precautions and control: (1) Place patients with C. difficile in private rooms. (2) If private rooms are not available, place these patients in rooms with other patients who have C. difficile-associated disease. (3) Perform hand hygiene procedures preferably using soap and water-not alcohol. To reduce the transmission of C. difficile spores, environmental disinfection with 10% sodium hypochlorite and hand-washing with soap and water can be effective at removing the spores from hands and surfaces.

Strict antiseptic procedures should be followed by health care workers in contact with the patient, and these procedures should include the use of disposable gloves, and a mask and gown. Because alcohol is ineffective at killing C. difficile spores, health care workers must frequently wash their hands with soap and water, rather than with alcohol-based waterless hand sanitizers, especially when caring for CDAD patients. Patient-care equipment (e.g. blood-pressure cuffs, stethoscopes and thermometers) should either be used only for the infected patient or cleaned well before they are used with another patient.

Enhanced environmental cleaning following a regular schedule with dilute bleach should be used to eliminate C. difficile spores from all patient contact surface areas. These spores may remain on infected surface areas for months or even years.

In addition, note the ability of the vegetative form of C. difficile to survive on moist surfaces. On dry surfaces, vegetative C. difficile cells die rapidly, whereas they remained viable for up to 6 h on moist surfaces in room air. This illustrates the importance of washing and drying room surfaces when cleaning contaminated rooms.

A very important method of controlling outbreaks of C. difficile-associated disease should be restricting the use of antimicrobial agents that have been implicated as risk factors for the disease, as recommended by Gerding et al[116]. Davey et al[170] documented that interventions to improve antibiotic prescribing practices to hospital inpatients can be successful, and that they can reduce antimicrobial resistance and the rates of hospital-acquired infections.

Effective surveillance of antibiotic-resistant bacteria and CDAD must be intensified in every healthcare setting, but especially in long-term care and rehabilitation facilitie. All these facilities must have easy laboratory access for prompt and active surveillance culturing and C. difficile cyto-toxin testing, for both A and B, at the earliest indication of any infection or CDAD. In addition, Furuno et al[171] advise that those patients at higher risk for carriage of antibiotic-resistant bacteria should be identified early for active surveillance targeting-culturing for methicillin-resistant S. aureus (MRSA) and vancomycin-resistant enterococci (VRE); e.g. patients who report having had antibiotics or prior hospital admissions within the past year. The authors found that this was very cost-effective, saving a projected $19 000-$26 000 relative to non-directed hospital wide screening for resistant organisms (MRSA and VRE), during the 8-mo study period at their tertiary care facility. They also found that there often is a significant delay between the onset of CDAD symptoms and the full implementation of CDC contact precautions.

Recommended for all acute care hospitals using strength of recommendations and quality of evidence from Table 9.

A. Components of a C. difficile infection prevention program: (1) Use contact precautions for infected patients, with a single-patient room preferred (A-II for hand hygiene, A-I for gloves, B-III for gowns, and B-III for single-patient room). (2) Ensure cleaning and disinfection of equipment and the environment (B-III for equipment and B-II for the environment). (3) Implement a laboratory-based alert system to provide immediate notification to infection prevention and control personnel and clinical personnel about patients with newly diagnosed C. difficile infection (B-III). (4) Conduct C. difficile infection surveillance and analyze and report C. difficile infection data (B-III). (5) Educate healthcare personnel, housekeeping personnel, and hospital administration about C. difficile infection (B-III). (6) Educate patients and their families about C. difficile infection, as appropriate (B-III). (7) Measure compliance with CDC or World Health Organization hand-hygiene and contact precaution recommendations (B-III).

Perform a C. difficile infection risk assessment. These special approaches are recommended for use in locations and/or populations within the hospital for which outcome data and/or risk assessment suggest lack of effective control despite implementation of basic practices.

A. Approaches to minimize C. difficile transmission by healthcare personnel: (1) Intensify the assessment of compliance with process measures (B-III). (2) Perform hand hygiene with soap and water as the preferred method before exiting the room of a patient with C. difficile infection (B-III). (3) Place patients with diarrhea under contact precautions while C. difficile test results are pending (B-III). (4) Prolong the duration of contact precautions after the patient becomes asymptomatic until hospital discharge (B-III).

B. Approaches to minimize C. difficile infection transmission from the environment: (1) Assess the adequacy of room cleaning (B-III). (2) Use sodium hypochlorite (bleach)-containing cleaning agents for environmental cleaning. Implement a system to coordinate with the housekeeping department if it is determined that sodium hypochlorite is needed for environmental disinfection (B-II).

C. Approaches to reduce the risk of C. difficile infection acquisition: Initiate an antimicrobial stewardship program (A-II).

(1) Do not test patients without signs or symptoms of C. difficile infection for C. difficile (B-II). (2) Do not repeat C. difficile testing at the end of successful therapy for a patient recently treated for C. difficile infection (B-III).

In the US, The Deficit Reduction Act of 2005 (P.L. 109-171) requires the Centers for Medicare & Medicaid Services (CMS), the US federal agency which administers Medicare, Medicaid, and the State Children’s Health Insurance Program to deny the assignment of a case to a higher DRG (payment to a health care facility) based on the occurrence of one of a selected number of hospital-acquired conditions, if that condition was acquired during the hospitalization. This rule is named CMS-1390-P: Medicare Program; Proposed Changes to the Hospital Inpatient Prospective Payment Systems and Fiscal Year 2009 Rate - Provisions on Preventable Hospital-Acquired Conditions Including Infections. The US Congress requires CMS to select conditions that are high cost, high volume, or both; assigned to a higher paying DRG when present as a secondary diagnosis; and reasonably preventable through the application of evidence-based guidelines.

In its original ruling, CMS had proposed adding nine hospital-acquired conditions, including C. difficile to the list of hospital acquired infections for which it proposed to deny payment. This would have taken effect on 1 October 2008. In July 2008, however, in response to an April 2008 letter sent by all three of the US major gastroenterology organizations, CMS, in a final rule setting policies and payment rates for the hospital setting, decided not to add C. difficile to the list of ‘Hospital-Acquired Conditions for Which It Will Deny Payment’.

The gastroenterology organizations in its April 2008 letter to CMS focused on the last criterion necessary for ‘non-reimbursement’-reasonably preventable through the application of evidence-based guidelines.

US gastroenterology organizations, in the letter to CMS, made much about the alleged fact that alcohol-based products are effective against the majority of microorganisms other than (i.e. not for) C. difficile, with the statement that “Alcohol-based products, in compliance with CDC guidelines, have played a significant role in potentially complicating efforts to avoid the spread of CDAD.”

“Indeed”, stated the letter, quoting Shen et al[184] “the proportion of all hand hygiene episodes performed with soap and water dropped from 90% to 15%, three years after the introduction of alcohol hand gels in one U.S. teaching hospital”. The letter continued by stating “that the trade-off of higher overall compliance against more focused use of soap and water is one that CMS must consider given that ‘CDC guidelines have played a significant role in potentially complicating efforts to avoid the spread of CDAD’. “In conclusion”, ended the letter to CMS, “the ACG, AGA and ASGE urge CMS not to add C. difficile to its list of hospital-acquired conditions for which additional payment as a complicating condition would not be available. We strongly believe that the disease is not reasonably preventable. Adding it to the list would create a very expensive and unworkable situation for CMS, hospitals, physicians and patients.”

However, as we have seen in our above review of C. difficile infection, the fact is that the hand washing issue is controversial and basically not objectively ascertained with good RCTs. Nevertheless, one can agree that hand washing at least may be superior secondary to the mechanical shedding of C. difficile spores with vigorous hand washing.

One can agree with the arguments against adding C. difficile infection to the list of non-reimbursable hospital services because of its variable incubation period. Complicating the accurate diagnosis of CDAD is that, while symptoms typically occur within 48 h of infection, patients infected in the hospital with C. difficile usually become infected within 3 wk of admission. However, the onset of symptoms can be delayed by 2-3 mo[30]. Also, although most cases of CDAD occur on days 4-9 of antibiotic therapy[15], the subsequent diagnosis of CDAD is not always possible upon admission to the hospital, due to a variable incubation period. Therefore, one can agree that it is not reasonable to hold an inpatient hospital liable for a condition acquired in a different setting, one which is not even always detectable upon the patient’s admission into their setting.

The CMS accepted these arguments against including C. difficile infection in those nosocomial infections that are not reimbursable. However, that still does not alleviate the responsibility of each healthcare facility to make aggressive attempts to counteract this problem. One can look to successful efforts made by others.

The University of Pittsburgh Medical Center developed a program, as mentioned in our review, consisting of education, increased early case finding, expanded infection-control measures, development of a C. difficile infection management team, and microbial management. The aggregate rate of C. difficile infection decreased from 7.2 infections per 1000 (9.4 during a peak) hospital discharges to 4.8 infections per 1000 hospital discharges, and later, was 3.0 infections per 1000 hospital discharges. The rates of compliance with hand hygiene and isolation were 75% and 68%, respectively[185].