Carbapenem-Resistant Klebsiella Outbreak and an Unreliable Colistin Disk
A 16-bed medical intensive care unit notes six patients with carbapenem-resistant Klebsiella pneumoniae (CRKP) isolated from tracheal aspirates over 10 days; four had ventilator-associated pneumonia, two died. Isolates are resistant to meropenem (MIC >16 mg/L by broth microdilution), piperacillin-tazobactam, and cefotaxime; one blood isolate is also resistant to colistin by broth microdilution (MIC 8 mg/L), but the laboratory's routine colistin disk diffusion showed a 12 mm zone, which a trainee recorded as susceptible. Environmental swabs grow CRKP from a sink drain and from a shared ultrasound gel container. Hand hygiene compliance is 48%. The hospital has no antimicrobial stewardship pharmacist and uses colistin plus meropenem empirically for CRKP.
1. Interpret the colistin susceptibility result. Was the trainee's 'susceptible' report correct, and what should the laboratory do?
Read the worked answer
The trainee's interpretation is not correct. Colistin disk diffusion is unreliable because the drug diffuses poorly through agar, producing zones that do not correlate well with MIC. A 12 mm zone should not be reported as susceptible.
The broth microdilution MIC is 8 mg/L. By CLSI criteria, colistin MIC ≥4 mg/L is resistant; by EUCAST, MIC >2 mg/L is resistant. Therefore the isolate is colistin-resistant, and the disk result must be suppressed.
Laboratory action:
- Report the isolate as colistin-resistant based on broth microdilution.
- Do not perform or report colistin disk diffusion for clinical decisions.
- Use broth microdilution with appropriate quality control strains, and consider a second method or reference laboratory confirmation.
- Add a comment that colistin disk diffusion is not predictive and that MIC testing is required.
- Review other CRKP isolates from the unit with the same method.
Clinically, colistin is unlikely to be effective for this blood isolate, so empirical colistin plus meropenem is not appropriate for this patient.
2. What outbreak investigation and infection-control steps should the ICU and laboratory take?
Read the worked answer
Outbreak steps:
- Define a case: a patient with CRKP from any clinical or surveillance site who was in the ICU during the outbreak period.
- Build a line list with admission dates, ventilation, procedures, rooms, shared equipment, and antibiotic exposure.
- Cohort colonised and infected patients, and perform active surveillance cultures such as rectal or tracheal screening according to local policy.
- Audit hand hygiene immediately; 48% compliance is a major driver. Provide feedback and ensure alcohol rub at every bed space.
- Use contact precautions, dedicated equipment, and terminal cleaning with an effective agent.
- Investigate the environment: sink drains, taps, shared ultrasound gel, cleaning cloths, and reusable containers. Discard the shared gel and use single-use packets.
- Review antibiotic use, especially carbapenems and colistin, and introduce stewardship review.
- Communicate findings to the ICU and infection-control committee; repeat environmental and patient sampling after interventions.
Do not stop at identifying the organism; the goal is to interrupt transmission.
3. How should empirical therapy and antimicrobial stewardship be changed for CRKP in this unit?
Read the worked answer
Empirical colistin plus meropenem is problematic when colistin resistance is present and meropenem MICs are >16 mg/L. For a critically ill patient with suspected CRKP, therapy should be guided by susceptibility and, where possible, by rapid carbapenemase testing.
Options depend on the resistance mechanism:
- For KPC-producing CRKP, ceftazidime-avibactam or meropenem-vaborbactam may be appropriate.
- For metallo-β-lactamase producers, ceftazidime-avibactam alone is not active; cefiderocol or a combination regimen may be needed with expert input.
- Source control and removal of infected devices are as important as drug choice.
Stewardship actions:
- Stop using disk diffusion for colistin and report MICs.
- Review all CRKP cases with a pharmacist and infectious-diseases physician; the hospital should appoint stewardship leadership.
- Avoid prolonged empirical broad-spectrum therapy; de-escalate when culture results allow.
- Track carbapenem and colistin use and provide feedback to prescribers.
- Use combination therapy only when justified; do not add colistin to a colistin-resistant isolate.
The unit needs a prospective antimicrobial review, not a single drug change.