Antimicrobial Stewardship and Resistance — Written Clinical Reasoning

SAQ 1 — Integrated Management of Severe Clostridioides difficile Infection (20 marks, 30 minutes)

Prompt: Outline your integrated management plan for this patient with severe recurrent Clostridioides difficile infection, including: (a) the severity stratification you are applying and its implications for therapy, with the exact first-line drug, dose, route and duration; (b) the additional pharmacological therapy you would consider given her recurrence risk; (c) the supportive and infection-control measures; (d) the surgical and microbiological escalation pathway if she deteriorates; (e) the prevention of further recurrence; and (f) the high-yield exam traps in CDI management. [1]

Model Answer

(a) Severity stratification and first-line therapy (4 marks): [1]

This patient has severe and fulminant Clostridioides difficile infection. Severity is defined by the IDSA/SHEA criteria: a white cell count of 15 x 10^9 per litre or above (she is 26) and/or a creatinine 1.5 times the baseline or above (her creatinine has risen from 78 to 165, more than double) define severe disease; fulminant disease adds hypotension, shock, ileus or toxic megacolon. She meets the fulminant criteria because of her hypotension (blood pressure 92/58) and clinical instability, on top of severe biochemical markers [2]. The most appropriate first-line therapy for fulminant CDI is oral vancomycin 500 mg four times daily PLUS intravenous metronidazole 500 mg every 8 hours, continued until the patient is improving and the diarrhoea is settling (typically 10 to 14 days). The reason for the higher vancomycin dose (500 mg rather than 125 mg) in fulminant disease is to achieve high intraluminal colonic concentrations in the face of an inflamed, oedematous mucosa and reduced gut motility; the addition of IV metronidazole provides systemic cover because oral vancomycin is not absorbed and may not reach the colon in adequate concentration if there is an ileus. Metronidazole is no longer first-line for any non-fulminant adult CDI (replaced by fidaxomicin and oral vancomycin for non-severe and severe disease), but it retains a role as an IV adjunct in fulminant disease.

(b) Additional pharmacological therapy for recurrence prevention (3 marks): [1]

This patient is at high risk of recurrence because she is over 65, has severe CDI, has had a prior CDI episode within the past year, and is immunocompromised by her acute illness. The MODIFY I and MODIFY II trials demonstrated that a single 10 mg/kg intravenous infusion of bezlotoxumab (a human monoclonal antibody against toxin B), given WITH standard antibiotic therapy, reduced 12-week recurrence from 26.6 to 16.5 per cent in the pooled analysis, with the greatest benefit in high-risk patients [4]. I would administer bezlotoxumab on day 1 of therapy. I would NOT use faecal microbiota transplantation in the acute fulminant episode — it is reserved for second or subsequent recurrence after clinical stabilisation, not for primary treatment of fulminant disease. I would not switch to fidaxomicin at this acute stage because fulminant disease with shock warrants the high-dose vancomycin plus metronidazole combination; fidaxomicin is preferred for non-fulminant severe CDI and for recurrence, but the evidence in fulminant shock is weaker. Once she stabilises, I would consider a transition to fidaxomicin or a vancomycin taper to complete the course.

(c) Supportive and infection-control measures (3 marks): [1]

Supportive measures include intravenous fluid resuscitation (crystalloid, guided by perfusion and urine output) for her hypovolaemia and acute kidney injury, electrolyte correction (potassium and magnesium losses through diarrhoea), and avoidance of antimotility agents (loperamide is contraindicated in acute severe CDI because it precipitates toxic megacolon). I would stop the inciting ceftriaxone immediately. I would not give probiotics routinely — the evidence is modest and the patient is acutely ill. Infection control requires single-room isolation with contact precautions (gown and gloves for all contact), dedicated equipment (stethoscope, blood pressure cuff, thermometer), hand hygiene with soap and water (not alcohol-based hand rub, which is less effective against C. difficile spores) before and after every contact, and environmental cleaning with a sporicidal agent (hypochlorite-based bleach). I would notify the infection-control team for surveillance and contact tracing if there are other cases on the ward. [1]

(d) Surgical and microbiological escalation (3 marks): [1]

I would request an urgent surgical review because fulminant CDI can progress to toxic megacolon (caecal diameter above 6 cm with systemic toxicity), colonic perforation, or refractory septic shock — all of which are indications for subtotal colectomy, which is life-saving in refractory fulminant CDI. The clinical signs of deterioration that mandate surgical review are: worsening abdominal distension and tenderness, haemodynamic deterioration despite maximal medical therapy, rising lactate, falling platelet count (suggesting DIC), and radiological evidence of toxic megacolon or perforation on abdominal CT or X-ray. I would order a CT abdomen with contrast if there is any doubt about toxic megacolon, perforation or an alternative diagnosis. Microbiologically, I would send the isolate for ribotyping and binary toxin testing to identify the NAP1/BI/027 hypervirulent strain, which carries higher morbidity, mortality and recurrence [2], although the result does not change the acute management. I would involve the infectious diseases team from the outset.

(e) Prevention of further recurrence (3 marks): [1]

Recurrence is the central long-term challenge: after one recurrence the risk of a further recurrence is 40 to 65 per cent. The prevention strategy has several layers. First, minimise further antibiotic exposure — every additional antibiotic course predisposes to recurrence, and stewardship is itself preventive. Second, if she recurs, treat with fidaxomicin (which has a recurrence advantage over vancomycin from the Louie trial, 15 versus 25 per cent [6]) or a tapered-and-pulsed oral vancomycin course. Third, for a second or subsequent recurrence, arrange faecal microbiota transplantation, which achieved 94 per cent cure after one or two infusions in the van Nood randomised trial [5]. Fourth, bezlotoxumab given in the acute episode reduces recurrence for the next 12 weeks. Fifth, there is emerging evidence for the extending-pulsed fidaxomicin regimen (every-other-day dosing after an initial course) for multiply-recurrent CDI. I would give her written information about CDI, counsel her to seek early review for any recurrence of diarrhoea, and communicate the plan to her general practitioner.

(f) High-yield exam traps (4 marks): [1]

The traps are: (1) metronidazole as first-line — metronidazole is no longer first-line for any adult CDI; use fidaxomicin or oral vancomycin for non-fulminant disease, and reserve IV metronidazole as an adjunct in fulminant disease; (2) the test of cure — do NOT repeat the toxin assay after clinical resolution; toxin persists for weeks and treating an asymptomatic patient with a positive result is futile and counterproductive; clinical resolution is the endpoint; (3) antimotility agents in acute severe disease — loperamide is contraindicated in severe CDI because it precipitates toxic megacolon; (4) alcohol-based hand rub for C. difficile — soap and water is required because alcohol does not kill spores; (5) under-dosing oral vancomycin in fulminant disease — the 500 mg four times daily dose is required in fulminant disease, not the 125 mg dose used for non-severe and non-fulminant severe CDI; (6) missing toxic megacolon — any clinical deterioration in a patient with severe CDI mandates a surgical review and abdominal imaging, because subtotal colectomy is life-saving in refractory fulminant disease; (7) not addressing recurrence prevention — bezlotoxumab, fidaxomicin, and faecal microbiota transplantation all have roles in patients at high recurrence risk, and a patient with severe recurrent CDI deserves a documented recurrence-prevention plan. [1]

SAQ 2 — Carbapenem-Resistant Enterobacteriaceae Bacteraemia in a Repatriated Patient (10 marks)

Prompt: A 62-year-old man is repatriated to Australia after a 3-week hospital admission in Egypt for a motor vehicle accident complicated by a ventilator-associated pneumonia. Three days after his return, he develops fever and rigors. Blood cultures grow Klebsiella pneumoniae resistant to all cephalosporins, piperacillin-tazobactam, all carbapenems (meropenem, imipenem, ertapenem), aztreonam, and all fluoroquinolones, but susceptible to colistin and tigecycline. The laboratory reports a carbapenemase and PCR is pending for the type. Outline: (a) the immediate empiric therapy while awaiting carbapenemase typing; (b) the carbapenemase-specific definitive therapy once the type is known, with the key drug choices for KPC, NDM and OXA-48; (c) the infection-control response; (d) the nephrotoxicity-risk discussion; and (e) the source-control and microbiology review. [1]

Model Answer

(a) Immediate empiric therapy (2 marks): [1]

This patient has carbapenem-resistant Enterobacteriaceae (CRE) bacteraemia, with a mortality of 40 to 50 per cent in bacteraemia, in a patient with a classic epidemiological risk (recent hospitalisation in a high-prevalence region — the Middle East and North Africa). While the carbapenemase type is awaited, the empiric regimen is combination therapy anchored on intravenous colistin (loading dose 4 to 5 MU, then approximately 9 MU daily in two divided doses, renal-adjusted) plus high-dose meropenem (2 g every 8 hours extended infusion), with the addition of tigecycline (100 mg loading then 50 mg every 12 hours) if the source is soft-tissue or intra-abdominal. Combination therapy is used because CRE bacteraemia has high mortality, no single agent is reliably bactericidal against all carbapenemases empirically, and combination therapy (two or more active agents) appears to outperform monotherapy in observational data and meta-analyses. I would involve the infectious diseases team immediately and request urgent carbapenemase typing by PCR (for blaKPC, blaNDM, blaOXA-48, blaVIM, blaIMP) and the CarbaNP phenotypic test. [1]

(b) Carbapenemase-specific definitive therapy (3 marks): [1]

The definitive therapy is driven by the carbapenemase type, because the inhibitors (avibactam, vaborbactam) have differential activity: [1]

• KPC (Klebsiella pneumoniae carbapenemase) — definitive therapy is ceftazidime-avibactam 2.5 g intravenously every 8 hours (avibactam inhibits class A carbapenemases including KPC) or meropenem-vaborbactam 4 g every 8 hours (vaborbactam is active against KPC). These agents have transformed KPC-CRE management and avoid the nephrotoxicity of colistin. If the isolate is susceptible in vitro, switch from the colistin-based empiric regimen to the avibactam- or vaborbactam-based regimen.
• NDM (New Delhi metallo-beta-lactamase) — definitive therapy is aztreonam plus avibactam, because NDM is a class B metallo-beta-lactamase that does NOT hydrolyse aztreonam (the only beta-lactam spared), and avibactam protects aztreonam from co-produced ESBLs and AmpC enzymes. Ceftazidime-avibactam alone does NOT work against NDM (avibactam does not inhibit class B enzymes). If aztreonam-avibactam is unavailable, the alternative is colistin (often in combination with a carbapenem or tigecycline), with its attendant nephrotoxicity.
• OXA-48-like — definitive therapy is ceftazidime-avibactam 2.5 g every 8 hours (avibactam inhibits OXA-48), with colistin or tigecycline as alternatives. [1]

The commonest carbapenemase in patients from Egypt and the eastern Mediterranean is OXA-48, often co-produced with other beta-lactamases; NDM is commonest in patients from the Indian subcontinent and the Balkans; KPC is commonest in the Americas and parts of southern Europe (Italy, Greece). The carbapenemase type drives both the therapy and the public-health response, which is why molecular typing is non-negotiable in any CRE case. [1]

(c) Infection-control response (2 marks): [1]

The patient is placed in a single room with contact precautions (gown, gloves, dedicated equipment) and hand hygiene with soap and water before and after every contact. The infection-control team and the clinical microbiologist are notified immediately. Rectal surveillance swabs are taken from ward contacts (patients in adjacent beds and those who shared care) to detect cross-transmission. The ward is reviewed for other CRE cases, and if there is evidence of cross-transmission, an outbreak investigation is launched (enhanced cleaning with a sporicidal or appropriate disinfectant, screening of all admissions to the affected area, and a multidisciplinary review of hand hygiene adherence and equipment cleaning). The patient remains in contact isolation for the duration of this admission and on readmission, because CRE gut carriage is prolonged. [1]

(d) Nephrotoxicity-risk discussion (2 marks): [1]

Colistin carries nephrotoxicity in 40 to 60 per cent of treated patients (acute tubular necrosis, often reversible on cessation but sometimes leaving residual impairment) and neurotoxicity (peripheral neuropathy, respiratory muscle weakness, rarely requiring ventilation). I would discuss with the patient and family that the chosen drug carries a significant risk of kidney injury, that kidney function is monitored daily and the dose adjusted, that the alternative (untreated CRE bacteraemia) carries 40 to 50 per cent mortality, and that kidney recovery usually follows cessation of therapy. I would review the medication chart for concomitant nephrotoxins (NSAIDs, ACE inhibitors, contrast, aminoglycosides, vancomycin, piperacillin-tazobactam) and hold them where possible. If the isolate is KPC or OXA-48 and ceftazidime-avibactam is available, I would switch to it to avoid colistin nephrotoxicity entirely — which is one of the principal advantages of the newer agents. [1]

(e) Source control and microbiology review (1 mark): [1]

I would search for and control the source — the ventilator-associated pneumonia and any line infection (remove central lines, send tip cultures), drain any drainable collection, and debride any infected wound. I would repeat blood cultures every 48 to 72 hours until clearance, monitor the full blood count, C-reactive protein and renal function daily, and conduct a stewardship review at 7 days to assess response and plan the duration — typically 7 to 14 days for uncomplicated bacteraemia with source control, longer for endocarditis, osteomyelitis or undrained deep infection. I would communicate the microbiological findings, the infection-control measures, and the antimicrobial plan to the patient's general practitioner and to any receiving facility on discharge. [1]