Clostridioides difficile infection

• Clostridioides difficile infection (CDI) is colitis caused by a bacterium that forms hardy spores and releases two powerful toxins. It is the most common healthcare-associated infection in the United States, with nearly 250,000 hospital-acquired cases each year, and its toxin-driven diarrhoea can range from a few loose stools to a life-threatening, obstructed, dilated colon.
• The story of CDI is almost always the story of antibiotics: a course of clindamycin, a cephalosporin, a fluoroquinolone or a carbapenem disrupts the normal gut microbiota, and C. difficile, already present or newly acquired, multiplies and fills the ecological space.
• Spores resist drying, heat and alcohol gel. Contact precautions and hand washing with soap and water are not optional extras; they are the practical defence against the spread of the organism.
• Most mild and moderately severe disease is cured with one of two antibiotics: fidaxomicin is now preferred, and oral vancomycin is the standard alternative. Metronidazole is used only when neither is available.
• About two in ten patients treated with metronidazole or vancomycin recur, and more than half of those who recur once recur again. Treatment therefore does not end when the diarrhoea stops: the inciting antibiotic is stopped, acid suppression is reviewed, and recurrence prevention is planned.

By the end of this page you should be able to:

1. Explain why broad-spectrum antibiotics, proton-pump inhibitors and hospitalisation allow C. difficile to cause disease, and name the antibiotics at highest risk.
2. Describe the actions of toxins A and B in the colon and how they produce diarrhoea, colitis and pseudomembranes.
3. Distinguish asymptomatic carriage from mild diarrhoea, pseudomembranous colitis and fulminant colitis, including the picture of ileus and toxic megacolon.
4. Apply current testing rules: which patients to test, how the two-step algorithms work, and when not to test.
5. Classify severity from the white cell count and serum creatinine, and choose management accordingly.
6. Prescribe first-line and fulminant treatment with drug, dose, route and duration, and describe the two surgical options.
7. Counsel a patient after a first episode about recurrence risk and the drugs and biotherapeutics used to prevent further recurrences.
8. State safe infection control practice, including hand hygiene, isolation and environmental cleaning, and describe CDI care when tests or drugs are limited.

C. difficile is a normal-looking bacterium that behaves like a seed. Its vegetative form produces the toxins that make us ill, but its spores survive on bed rails, toilet seats and door handles for months. Inside the colon it is usually kept in check by the dense community of bacteria that shares the bowel. Illness begins when that community is thinned, most often by an antibiotic, and C. difficile is left to germinate, multiply and release its toxins. The damage is therefore toxin-mediated rather than invasive: the organism does not need to enter the bloodstream to make someone very sick.

C. difficile has become the defining hospital-acquired intestinal infection of the modern antibiotic era. There are an estimated 450,000 infections each year in the United States, with 14,000 associated deaths. ^[1] A US population surveillance programme that followed 10 cities from 2011 to 2017 counted about 306,500 healthcare-associated cases in 2011 falling to 235,700 in 2017, while community-associated cases rose over the same period from 170,000 to 226,400. ^[1] In other words, the infection has not gone away; its centre of gravity has shifted partly out of the hospital. Nearly 250,000 hospital-acquired cases still occur annually, and among patients admitted to hospital with CDI, 5 to 15 percent need intensive care and about a quarter are readmitted within 30 days, mostly because of recurrence. ^[3] In the intensive care unit, CDI develops in 2 to 5 percent of patients, and mortality in that setting runs from 20 to 40 percent. ^[3]

Globally, the picture is similar in broad outline, though incidence rates, ribotype distributions and antimicrobial resistance patterns vary by region. One statistic that unites settings is the burden on older patients: age of 65 years or more is consistently the demographic risk factor, and frailty, renal impairment and recent antibiotic exposure stack on top of it.

The greatest single modifiable risk factor is antibiotic exposure. Antibiotics disrupt the gut microbiota and open an ecological niche that C. difficile, whether newly acquired or already carried, expands into. A meta-analysis of hospital-acquired CDI in nearly 16,000 patients ranked the antibiotic classes by risk: third-generation cephalosporins were highest, followed by clindamycin, fourth-generation cephalosporins, carbapenems, fluoroquinolones and penicillin combinations such as piperacillin-tazobactam. Community data point even more sharply to clindamycin and later-generation cephalosporins. Across settings the high-risk list in clinical practice is clindamycin, cephalosporins, fluoroquinolones, carbapenems and broad-spectrum penicillin combinations, with risk rising as cumulative dose, number of antibiotics and duration of treatment increase. Even a single course of perioperative prophylaxis adds a little risk, which is why de-escalation and early cessation of antibiotics matter.

Other medications and conditions raise risk, though in different ways:

• Proton-pump inhibitors and other acid suppression. Earlier observational studies suggested an association, and the US drug regulator issued a warning for omeprazole in 2012. Current ICU-focused evidence, including a meta-analysis of nearly 4,000 intensive care patients, does not show an association between proton-pump inhibitors and CDI. Long-term use outside the intensive care unit may still matter a little, but proton-pump inhibitors should not be withheld when there is a clear indication purely because of CDI anxiety.
• Age over 65, hospitalisation and long-term care. Two of the three major population-level risk factors, alongside antibiotic exposure.
• Inflammatory bowel disease multiplies risk roughly fivefold in the community and is a cause of severe disease. ^[3]
• Immunosuppression: corticosteroids, chemotherapy, transplantation, haematological malignancy and biologic therapy, especially TNF blockade.
• Renal impairment, diabetes, malignancy, tube feeding and recent gastrointestinal surgery.
• Prior CDI itself, which is the strongest predictor of recurrence.

Community-associated cases now form a large and rising share of all CDI. These patients are younger, have fewer healthcare exposures and are more often women; many have had a recent outpatient antibiotic, especially for a respiratory or urinary infection.

Three players matter: the microbiota, the spore and the two toxins. Understanding their interaction lets you reason through the disease instead of memorising it.

The microbiota. A healthy colon holds trillions of bacteria that compete for nutrients, maintain the mucus barrier and produce short-chain fatty acids from fibre. This community suppresses C. difficile by primitive ecology: there is no free food and the environment is unfriendly. Broad-spectrum antibiotics, especially those excreted in bile and reaching the colon, remove that competition. Empiric blame has historically fallen on clindamycin, but any antibiotic can be the trigger in an individual patient.

The spore. Whether it arrives from the hospital environment or is already carried in the colon, C. difficile responds to disruption by germinating. Its spores resist alcohol gel, survive on dry surfaces and pass through stomach acid. Germination is promoted by primary bile acids such as cholate, which become more available when the bile-acid-metabolising microbiota is depleted. Secondary bile acids, which suppress germination, fall. So the same antibiotics that thin the bacteria also change the bile acid balance in favour of the pathogen.

Toxin A and toxin B. Only toxigenic strains produce disease, and the relevant virulence factors are two large clostridial cytotoxins, toxin A and toxin B. Both are single-chain proteins internalised by colonocytes; both inactivate Rho-family GTPases by glucosylation, dismantling the actin cytoskeleton, disrupting tight junctions, triggering fluid secretion and provoking a vigorous neutrophilic inflammatory response that thickens the mucosa. Toxin A was historically called the enterotoxin and toxin B the cytotoxin, and the historic 027 (NAP1/BI) epidemic strain produces both in greater quantity and an additional binary toxin. The clinical importance of the binary toxin is unsettled. What matters at the bedside is that disease is toxin-driven: there is no mucosal invasion in the ordinary sense. A strain that carries the toxin gene but is not expressing toxin causes no illness, which is exactly why test choice matters.

Pseudomembranes. In severe focal colitis the damaged mucosa exudes fibrin, mucus and neutrophils, which layer over denuded patches as yellow-white plaques. Not every patient with severe disease has pseudomembranes at endoscopy; their absence does not exclude CDI, and their presence, although characteristic, is not specific to it.

The spectrum is the spectrum of colonic toxin exposure: the patient can be an asymptomatic carrier, mildly unwell, or critically ill with a leaking, dilated colon. None of the presentations is specific.

Asymptomatic colonisation is present in up to 10 percent of hospitalised patients and a small proportion of healthy adults. ^[3] These patients need no treatment. Treating them harms the more useful microbiota and breeds resistance.

Mild to moderate CDI presents as new diarrhoea, usually watery, with lower abdominal cramps, a low-grade fever, anorexia and a mildly tender abdomen. The diarrhoea is the clue, not the fever: fewer than 15,000 white cells per microlitre and a creatinine no higher than 1.5 mg/dL usually sit in this group. ^[2] Bowel movements are typically 3 or more loose or watery stools in 24 hours for 2 consecutive days, often with mucus and occasionally occult blood. Frank haematochezia is unusual and should raise other thoughts.

Pseudomembranous colitis is the severe colitic end of the same process: more frequent watery or mucoid stools, fever, abdominal distension and marked tenderness. Endoscopy may show raised creamy or yellow-white plaques on an oedematous, erythematous mucosa; these may be patchy or confluent.

Fulminant colitis is the dangerous extreme: diarrhoea may actually stop. The abdomen becomes distended and tender, fever and tachycardia rise, white cells climb above 15,000 per microlitre, serum creatinine rises, lactate rises, and albumin falls. ^[2] Ileus traps gas and stool proximally, so the patient may vomit and have no bowel sounds. Toxic megacolon is the end stage, with a colon dilated worryingly on the plain abdominal film, sometimes with thumbprinting and target signs, perforation risk and shock. This is the patient who dies: mortality in fulminant disease remains 30 to 50 percent even with modern surgery. ^[3]

Many causes of diarrhoea must be sorted out before CDI is blamed. Osmotic antibiotic-associated diarrhoea, where antibiotic residues alter colonic absorption without true inflammation, is common and self-limited; pseudomembranous colitis is almost always CDI when it is present. Alternative infectious causes to keep in mind include Staphylococcus aureus, Klebsiella oxytoca, Clostridium perfringens, Salmonella and other foodborne organisms. Non-infective colitides include inflammatory bowel disease, ischaemic colitis, microscopic colitis, celiac disease and eosinophilic gastroenteritis. An acute abdomen due to CDI can mimic small-bowel obstruction, Ogilvie's syndrome, volvulus, appendicitis or ischaemia. Suspicion is sharpened by recent antibiotics, hospitalisation, long-term care, advanced age, recent chemotherapy or a known inflammatory bowel disease flare, but the final answer comes from testing appropriately selected patients.

The diagnosis is made in someone with clinically significant diarrhoea, because a stool test cannot tell infection from colonisation in a person without symptoms. Testing should be limited to patients with new-onset, unexplained diarrhoea of 3 or more unformed stools in 24 hours.

• Nucleic acid amplification tests (NAAT, PCR) detect the toxin gene. They are highly sensitive and specific when applied to the right patient, and rapid. They cannot tell whether toxin is actually being produced, so in a colonised patient they overdiagnose.
• Glutamate dehydrogenase (GDH) enzyme immunoassay detects an antigen common to both toxigenic and non-toxigenic C. difficile. It is very sensitive but not specific, so it is used as a screen.
• Toxin A/B enzyme immunoassay detects free toxin. It is highly specific but insensitive, because it needs abundant toxin to turn positive. A negative result does not exclude infection.
• Toxigenic culture and cell-cytotoxicity neutralisation assays are sensitive and specific reference standards, but they take days and are used mainly in outbreak investigation and research.

A single NAAT is too sensitive and a single toxin test is too insensitive, so modern practice uses a two-step design to separate infection from colonisation:

1. GDH screening first. If GDH is negative, CDI is very unlikely and no further test is needed.
2. If GDH is positive, test the same specimen for toxin A/B. A positive toxin test confirms CDI. A negative toxin test should be followed by NAAT for the toxin gene, because the patient may have toxin-producing infection below the toxin test's detection limit.
3. Where a NAAT is the first step, the same logic is inverted: a positive NAAT is confirmed by a toxin EIA on the same specimen, or by clinical assessment.

Which algorithm a laboratory uses depends on local capacity, but the principle is constant: combine a highly sensitive screen with a highly specific confirmation. Only liquid stool should be sent; formed stool from a person without diarrhoea should not be tested at all. The table below summarises what each result means.

Once a patient has responded, there is no role for repeat stool testing to confirm eradication. Toxin genes and glutamate dehydrogenase can remain detectable for weeks after the organisms have been cleared, and a positive result in a symptom-free patient leads to unnecessary re-treatment and to the mislabeling of colonisation as recurrence. A second episode should be diagnosed clinically and on a fresh test sample only when symptoms return. ^[1]

Full blood count, renal function, electrolytes, albumin and lactate are needed to grade severity and guide fluid resuscitation. Abdominal CT in severe disease may show colonic thickening, oedema, and signs of colitis; it is best reserved for patients who are severely ill, have complications, or need another diagnosis excluded. Abdominal radiography is used to look for ileus and toxic megacolon. Plain films in severe disease may show thumbprinting and target signs. Endoscopy is not routine. It is used when ileus is present, when the diagnosis is uncertain, when an alternative is suspected, or when faecal microbiota transplantation is being delivered. When performed, pseudomembranous plaques are seen in 50 to 85 percent of severe cases. ^[3]

Severity should be classified before treatment, because the drug and dose change with it. There is no international consensus on a single system, and IDSA/SHEA, ESCMID and ASID differ at the edges; the table below shows the consensus core.

ESCMID also adds a temperature above 38.5 degrees Celsius and marked large-bowel changes on imaging (distension, pericolonic fat stranding, colonic wall thickening) to the severe category, and bowel perforation or rapidly deteriorating renal function to the fulminant one. Patients in intensive care are often sicker than these numbers suggest, so clinical judgement matters. A patient with 12,000 white cells and a creatinine of 1.4 can still be septic and dehydrated. ^[2] Staging is dynamic and should be reassessed daily during the first 3 days. ^[2]

1. Stop the inciting antibiotic when it is safe to do so, and review every other medication for a narrower alternative: the antibiotic is the driver of the episode and of recurrence.
2. Review proton-pump inhibitors and other acid suppressants; stop them if there is no strong indication.
3. Avoid anti-peristaltic and opioid agents, which mask the clinical picture and can precipitate ileus and toxic megacolon.
4. Resuscitate the patient: fluids, electrolytes (potassium often falls), protein and nutrition, thromboprophylaxis and mobilisation.
5. Escalate early if white cells, creatinine, lactate or abdominal distension worsen, and involve surgery early in fulminant disease.

The 2021 IDSA/SHEA focused update makes fidaxomicin the preferred drug. It is a non-absorbed, narrow-spectrum macrocyclic antibiotic that reduces recurrence compared with vancomycin while achieving equivalent initial cure, preserves the microbiota better and is associated with less new colonisation by vancomycin-resistant enterococci and Candida. The decision is conditional on availability and cost; vancomycin remains an acceptable alternative.

• Fidaxomicin 200 mg by mouth twice daily for 10 days. ^[7]
• Vancomycin 125 mg by mouth four times daily for 10 days, if fidaxomicin is not available or not affordable. ^[8]
• Metronidazole 500 mg by mouth or intravenously three times daily for 10 to 14 days (oral 10 days in the ESCMID table) only when vancomycin and fidaxomicin are unavailable, for an initial non-severe episode in a low-risk patient. ^[2]

Metronidazole is no longer first-line: it has poor, unpredictable concentrations in the colon and worse cure rates in comparative studies. The argument for using it at all is access and cost. Where it is used, clinicians need to watch for the patient who is worsening despite it on day 2 or 3, because under-treatment of de facto severe disease carries excess mortality.

The principle is the same as non-severe disease, with the same drugs, but metronidazole must not be used alone:

• Fidaxomicin 200 mg by mouth twice daily for 10 days. ^[7]
• Vancomycin 125 mg by mouth four times daily for 10 days. ^[8]

Fidaxomicin is endorsed for severe non-fulminant disease by IDSA/SHEA and ESCMID. There is no good evidence for its use in fulminant, collapsing disease, where the pharmacokinetics of a non-absorbed drug become unreliable.

This is an intensive-care-level emergency. The drug combination is deliberately aggressive and partially systemically delivered:

• Vancomycin 500 mg by mouth or via nasogastric tube every 6 hours for at least 10 days, extended according to the clinical course. ^[7][3]
• Metronidazole 500 mg intravenously every 8 hours, added on the rationale that it reaches colonic tissue via the bloodstream even when gut motility is lost. A 2015 retrospective study showed lower mortality with the combination than with vancomycin alone in critically ill patients; a 2019 study did not, but it remains standard practice until better evidence exists. ^[3]
• Vancomycin 500 mg in 100 mL normal saline per rectum every 6 hours as a retention enema in anyone with ileus, to deliver drug distal to the point where the oral dose can no longer reach. ^[2][3]
• Fluid resuscitation, electrolyte management, vasopressors and organ support as needed.

High-dose oral vancomycin is justified by the fear that ileus prevents the standard dose from reaching the worst-affected bowel. European guidelines still prefer the lower dose in patients without ileus, which the European society cites for patients without ileus or toxic megacolon. ^[3]

Surgical referral should happen early and in parallel with maximal medicine, not after everything has failed and the patient is unable to survive an operation. Indications include failure of medical therapy, peritonitis, toxic megacolon, bowel perforation, septic shock requiring vasopressors, end-organ failure and rising lactate. Post-operative mortality in fulminant CDI is 30 to 50 percent, ^[3] which is why timing matters: the operation has to happen in the window after medical failure but before irreversible shock. Serum lactate in the range 2.2 to 4.9 mmol/L identifies patients most likely to benefit, ^[3] while a lactate of 5 mmol/L or higher signals futility.

Two operations are used:

• Total abdominal colectomy with end ileostomy, removing the whole colon; formerly the default.
• Diverting loop ileostomy with intraoperative colonic lavage and antegrade vancomycin flushes, which is colon-sparing and allows easier restoration of bowel continuity later. Results in the largest comparative series are equivalent in mortality to colectomy (26 versus 31 percent), ^[3] and it cannot be used if there is compartment syndrome, perforation or colonic necrosis.

Attempts at rescue with intravenous immunoglobulin, tigecycline and faecal microbiota transplantation exist for patients who fail maximal therapy but are not surgical candidates. Tigecycline can add 100 mg intravenously once then 50 mg twice daily, but the evidence is mixed. Intravenous immunoglobulin 150 to 400 mg/kg works by neutralising toxins and has been used off-label since 1991; it is not in guidelines. Cholestyramine and tolevamer are not recommended. ^[3]

Recurrence, the reappearance of symptoms after completing treatment for a prior episode (trials usually count it through 8 weeks after treatment), is the clinical problem that defines modern CDI care. Treatment with metronidazole or vancomycin is associated with recurrence in 20 to 30 percent of patients, who then have a 50 to 60 percent chance of a further recurrence. A first recurrence is usually a relapse from surviving spores rather than reinfection, because neither drug reliably kills spores.

Treating the first recurrence is similar to the initial episode:

• Discontinue unnecessary non-C. difficile antibiotics.
• If fidaxomicin or vancomycin was used first time, both are valid again, because recurrence is not usually from developed resistance.
• Fidaxomicin is preferred; vancomycin or metronidazole may be used if fidaxomicin is unavailable, with bezlotoxumab considered in high-risk patients.
• The vancomycin tapered and pulsed regimen is an acceptable alternative: 125 mg four times daily for 10 to 14 days, twice daily for 7 days, once daily for 7 days, then every 2 to 3 days for 2 to 8 weeks. This schedule lets peaks of drug kill newly germinated vegetative cells while drug-free intervals allow the recovering microbiota to re-establish, leaving fewer niches for spores. ^[2]

For a second or further recurrence, fidaxomicin (standard or extended-pulsed) or a vancomycin taper is used, and faecal microbiota transplant becomes appropriate. The extended-pulsed fidaxomicin regimen, 200 mg twice daily for 5 days followed by 200 mg every other day for 20 days, achieved the lowest recurrence rates yet seen in an antibiotic trial, particularly in hospitalised patients over 60.

Bezlotoxumab, a monoclonal antibody against toxin B given once intravenously at 10 mg/kg alongside standard antibiotics, reduced recurrence in the MODIFY trials and is conditionally recommended for patients with a prior recurrence within 6 months, or for patients on a first episode with other risk factors (age 65 or more, immunocompromise, severe disease on presentation). It is cautioned in a history of congestive heart failure, where deaths were higher (19.5 versus 12.5 percent). It was not well studied alongside fidaxomicin. Critically, Merck discontinued bezlotoxumab (Zinplava), and the European Commission withdrew the marketing authorisation at the company's request in October 2026, ^[6] so it is no longer available for routine clinical use and is now only a historical example of toxin-targeted prophylaxis. ^[2]

Faecal microbiota transplantation treats the problem that antibiotics leave behind: a depleted and dysbiotic gut. It works by restoring a complex microbial community, and restores cure in 65 to 95 percent of multiply recurrent cases; individual studies report up to 92 percent. It is traditionally delivered by colonoscopy, although enema, nasoenteric tube and capsule routes all work. It should not be used lightly: donor stool must be screened for transmissible pathogens, rare invasive infections and deaths have occurred with unscreened donations, and recent fatalities linked to extended-spectrum beta-lactamase-producing organisms remind us that donor quality is the whole story. It is reserved for patients with at least two recurrences or a first or second episode that is refractory and not improving, after appropriate antibiotic courses. ^[2][3]

Approved live biotherapeutics now sit alongside transplant as regulated, quality-controlled options, both licensed for prevention of recurrence after antibiotic treatment:

• REBYOTA (fecal microbiota, live-jslm) is a 150 mL suspension from qualified donors, given once as a rectal enema, 24 to 72 hours after the last antibiotic dose. It is approved for adults 18 years or older following antibiotic treatment for recurrent CDI. ^[4]
• VOWST (fecal microbiota spores, live-brpk) is a capsule preparation of purified Firmicutes spores, given as 4 capsules by mouth once daily for 3 consecutive days, beginning 2 to 4 days after antibiotics, with 296 mL of magnesium citrate laxative the day before and an 8-hour fast; take on an empty stomach. It reduced recurrence from 39.8 to 12.4 percent at 8 weeks in its pivotal trial. ^[5]

Both are prevention products, not treatment for active infection, and should be given with the same attention to antibiotic washout that the label specifies. Ser-109, the research name for the VOWST organism, showed a similar magnitude of benefit.

Antibiotic prophylaxis during future antibiotic courses is a narrower tool. For selected patients with a history of multiply recurrent CDI who must take systemic antibiotics for another infection, oral vancomycin prophylaxis 125 mg once or twice daily during treatment and for 5 to 7 days afterwards can be considered; routine prophylaxis with vancomycin or fidaxomicin for CDI-prone patients on antibiotics is not recommended for most, because microbiome protection matters. Rifaximin 400 mg three times daily for 20 days after a standard course can follow multiple recurrences, as can suppressive oral vancomycin 125 mg daily for patients who need frequent antibiotic courses or are not transplant candidates.

• Recurrence, with reinfection-like risk rising after each episode.
• Dehydration and electrolyte disturbance, especially hypokalaemia.
• Acute kidney injury from hypovolaemia on top of baseline renal disease.
• Hypoalbuminaemia, which both marks and worsens severe disease.
• Ileus and toxic megacolon, with perforation, peritonitis and sepsis.
• Bowel perforation and abdominal compartment syndrome.
• Lower gastrointestinal haemorrhage from mucosal necrosis.
• Malnutrition from prolonged diarrhoea and fear of eating.
• Death, which is especially common in the intensive care population. ^[3]
• Stricture and disordered evacuation are rare late sequelae.
• Reactive arthritis and bacteraemia are uncommon.
• Over-treatment and false positivity from poor testing also causes harm: unnecessary vancomycin, colonisation mistaken for disease, and the incorrect label of CDI in a patient with another diarrhoeal illness.

Most uncomplicated CDI responds quickly to appropriate therapy and resolves during the standard course of treatment. Recurrence is the dominant long-term outcome, at 20 to 30 percent after a first episode and higher thereafter. Age over 65, renal impairment, continued broad-spectrum antibiotics, proton-pump inhibitor use in some studies, and a hypervirulent ribotype worsen prognosis. Prompt discontinuation of the inciting antibiotic and appropriate antibiotic choice, including fidaxomicin where affordable, directly reduce the probability of recurrence. Fulminant disease carries the worst prognosis, particularly with late surgery.

Prevention works at four levels: antibiotic stewardship, source control with barriers and hand hygiene, environmental cleaning, and early, correct diagnosis.

• Antibiotic stewardship is the lever with the largest evidence base. A quality programme that reduced antibiotic use by 30 days of therapy per 1,000 patient-days reduced hospital-acquired CDI by 19.5 percent in one study. ^[1] Interventions that work: avoid unnecessary antibiotics, use the narrowest spectrum and shortest duration that fits the infection, de-escalate promptly on culture results, avoid clindamycin and broad cephalosporins when an equally effective alternative exists, and limit perioperative prophylaxis to one dose where protocol permits.
• Isolation. Place the patient in a single room or cohort with other CDI patients. Use gloves and gowns for any contact with the patient or their environment, and dedicated, disposable equipment. The patient remains under precautions until 48 hours after diarrhoea resolves.
• Hand hygiene. This is the one place where soap and water beats alcohol gel. Alcohol-based hand rubs do not kill C. difficile spores. Hands should be washed with soap and water with proper friction before leaving the room and between patients. ^[3]
• Environmental cleaning. Spores survive on surfaces for months. Terminal cleaning of the room with a sporicidal agent, typically diluted sodium hypochlorite (bleach used as a sporicidal solution) or another product certified sporicidal, is required after every CDI case and in outbreak situations. Routine detergent alone leaves spores viable. Cleaning of high-touch surfaces and toilet seats during the illness matters as much as the terminal clean.
• Personal equipment and laundry. Use single-use or dedicated thermometers, stethoscopes and commodes, and handle linen as contaminated.
• Screening and early recognition of outbreaks by surveillance of CDI incidence and ribotype, and rapid review of antibiotic policies when rates climb.

A safe low-resource environment need not be high-tech: single room where possible, gloves and aprons, soap-and-water hand washing for everyone, bleach disinfection of toilets, and strict discipline about not sharing bedpans. Healthcare workers in resource-limited settings often face spore cross-contamination from shared toilets and inadequate cleaning; these simple measures are the difference.

Inflammatory bowel disease. Test any patient with inflammatory bowel disease who presents with an acute flare and diarrhoea, because CDI can mimic a flare and both can coexist. If CDI is diagnosed, treat with vancomycin 125 mg by mouth four times daily for 10 to 14 days, preferring fidaxomicin if it is available and affordable. Faecal microbiota transplantation should be considered for recurrent CDI in this group.

Older adults. CDI incidence and mortality both rise sharply after 65, and the presence of comorbidity is the real driver. Early fluids, nutrition and cautious medication review are important.

Pregnancy. Vancomycin is the first choice in pregnancy, as systemic absorption is negligible. Metronidazole is generally avoided in the first trimester. Fidaxomicin data in pregnancy are limited, so vancomycin is preferred. Expert obstetric and infectious diseases advice should be sought.

Children. CDI in children usually follows antibiotic exposure or hospitalisation, and community acquisition is increasingly recognised. Treatment is age and weight-based; concurrent testing rules and the two-step approach still apply.

Immunocompromised patients. They present later and more severely, are more prone to recurrence and may benefit from faecal microbiota transplantation. Bezlotoxumab was historically considered, but is no longer available. Probiotics are not recommended in these patients because of invasive fungal and bacterial infection risk.

Where endoscopy, computed tomography and second-line drugs are limited, the honest approach is to treat more on clinical grounds and less on laboratory detail.

• Use the bedside criteria. Diarrhoea with 3 or more unformed stools in 24 hours, recent antibiotic exposure, age over 65, hospital or long-term care and a tender lower abdomen make CDI the working diagnosis. A raised white cell count and creatinine, where available, confirm severity.
• Treat early. If vancomycin is unavailable, use metronidazole 500 mg three times daily for 10 to 14 days and arrange access. Oral vancomycin, where it can be obtained, can be prepared from the intravenous formulation when the brand product is unavailable; many settings do this.
• Fulminant disease still needs high-dose vancomycin plus intravenous metronidazole and rectal vancomycin. Rectal vancomycin can be given as 500 mg in 100 mL normal saline by enema or by instillation down a rectal catheter, retained for an hour. Surgical options are limited to what is available: colectomy remains the fallback if nothing else can be done. Even partial colectomy or diverting loop ileostomy, where surgical expertise and equipment exist, carries lower morbidity.
• Diagnostic reality. Where two-step testing is unavailable, a single well-validated toxin test or NAAT is acceptable, but it must be used only on patients with diarrhoea meeting criteria, and a negative test in a patient with classical features should be treated empirically if severe.
• Prevention as above, at low cost: soap and water, single-use gloves, dedicated commodes, disinfected surfaces and discipline around sharing of pills and syringes. Antibiotic stewardship audits, even simple point-prevalence surveys of prescribing, quickly pay for themselves.
• Nutrition and fluids. In settings where oral rehydration is under-used, meticulous fluid and electrolyte replacement does more good than a CT scan.
• Referral. Patients with fulminant disease or recurrence need the nearest centre with surgery and, where possible, faecal microbiota transplantation capability. Donor stool screening programmes are being built in many regions and local protocols should be followed.

• CDI is a toxin-mediated colitis caused by toxigenic C. difficile, not an invasive infection. Toxins A and B glucosylate Rho GTPases, damage tight junctions and produce diarrhoea, colitis and pseudomembranes.
• Spores survive alcohol gel, drying and heat; soap and water hand washing removes both spores and vegetative forms.
• Broad-spectrum antibiotics are the principal risk factor: third-generation cephalosporins, clindamycin, fourth-generation cephalosporins, carbapenems, fluoroquinolones and penicillin combinations, with dose and duration multiplying risk.
• Age over 65, hospitalisation, long-term care, inflammatory bowel disease and immunosuppression are the major non-antibiotic risk factors; community-onset CDI now accounts for a large and rising share of cases.
• Only test patients with 3 or more unformed stools in 24 hours, and use a two-step algorithm (GDH screen plus toxin EIA, with NAAT when results disagree). ^[2] Do not test formed stool or asymptomatic patients, and do not test for cure.
• Non-severe disease: fidaxomicin 200 mg twice daily for 10 days, ^[7] or vancomycin 125 mg four times daily for 10 days if unavailable; ^[8] metronidazole only if neither is available.
• Severe disease: same drugs at standard doses; metronidazole must not be used alone.
• Fulminant disease: vancomycin 500 mg every 6 hours by mouth or tube plus metronidazole 500 mg intravenously every 8 hours, ^[7][3] plus vancomycin 500 mg per rectum every 6 hours if ileus; ^[2] early surgical consultation.
• Surgery: total abdominal colectomy with end ileostomy, or diverting loop ileostomy with intraoperative lavage and antegrade vancomycin flushes.
• Recurrence runs 20 to 30 percent after a first episode and higher later; ^[2] tapered or pulsed vancomycin, fidaxomicin, antibiotic stewardship and, for multiple recurrences, faecal microbiota transplantation or approved live biotherapeutics such as REBYOTA and VOWST all help. Bezlotoxumab worked but is no longer available.
• Testing for cure and treating colonisation cause more harm than good: clinical judgement guided by the tests is the safer habit.

• Acute diarrhoea and gastroenteritis
• Inflammatory bowel disease (ulcerative colitis and Crohn disease)
• Gut microbiota and dysbiosis
• Antibiotic stewardship: choosing and stopping antibiotics
• Faecal microbiota transplantation
• Ischaemic colitis and other causes of acute colitis
• Electrolyte disturbances, especially hypokalaemia
• Sepsis and multi-organ support
• Hospital infection control bundles
• Pharmacokinetics of poorly absorbed oral antibiotics