KPC-Klebsiella Cluster in a Surgical ICU
A 420-bed tertiary hospital’s surgical ICU has identified four patients with Klebsiella pneumoniae carbapenemase (KPC)-producing K. pneumoniae over six weeks. Two patients had clinical infections (one ventilator-associated pneumonia, one abdominal abscess); two had positive rectal surveillance cultures only. A sink drain in the medication preparation area also grew KPC-K. pneumoniae. The laboratory uses a commercial PCR assay for carbapenemase genes on colonies and has had two invalid runs this month. The infection preventionist is asked whether this is an outbreak, whether the sink is the source, and how to define cases. The ICU has 18 beds and current occupancy is 17; two of the affected patients are still admitted and share a hallway bathroom. There is no prior history of CRE in this unit. The hospital does not routinely perform rectal screening on ICU admission. The lab director reports that one clinical isolate was initially misidentified as Pseudomonas aeruginosa by an automated panel and was corrected by MALDI-TOF.
1. How should the infection preventionist define a case for this investigation, and what distinction should be made between infection and colonization?
Read the worked answer
Use a tiered, time- and place-linked definition, not a single label. A confirmed case could be any patient admitted to or seen in the SICU during the six-week period with KPC-K. pneumoniae isolated from a clinical or surveillance specimen, confirmed by MALDI-TOF and carbapenemase testing. Separate clinical infection from colonization: the two patients with VAP and abscess meet infection criteria; the two rectal-only patients are colonized. Colonized patients still count for transmission investigation because they are reservoirs, but they should not be counted as clinical infections in rates. The environmental sink isolate is not a human case; it is an exposure signal. Before finalizing, exclude duplicates and verify the misidentified isolate with repeat testing and, ideally, whole-genome sequencing. If the lab cannot confirm KPC in all isolates, the case count may change. This approach prevents overcounting while preserving sensitivity for transmission.
2. What steps would you take to determine whether this is a true cluster or a pseudo-outbreak?
Read the worked answer
Start with the laboratory: review the two invalid PCR runs, the misidentified Pseudomonas isolate, and specimen handling for contamination. Re-test archived isolates with MALDI-TOF and a second carbapenemase method; send all four isolates plus the sink isolate for WGS or PFGE. Simultaneously review the clinical micro records for duplicate specimens and changes in culture ordering-an increase in surveillance cultures can create the appearance of a cluster. Plot cases by admission date, unit location, and shared equipment. Calculate the baseline CRE rate for the SICU over the prior 12 months. A true cluster is supported if isolates are clonal and patients overlap in time and place, especially with a positive environmental link. If WGS shows unrelated strains, the cases may be sporadic introductions and the sink may be a bystander. Do not attribute causality to the sink until environmental and clinical isolates match.
3. A new SICU patient has a positive rectal PCR for KPC but a negative culture and no symptoms. How do you interpret and manage this result?
Read the worked answer
A positive PCR with a negative culture is ambiguous. PCR detects carbapenemase genes even when organisms are nonviable or present at low burden; culture is the gold standard for confirming viable colonization. In an ongoing SICU cluster, I would place the patient on contact precautions in a single room if possible, because the cost of missing a colonized patient is high. I would not treat, since there is no infection. I would repeat a rectal culture with an enrichment step and ask the lab to hold the isolate for typing. If repeat cultures are negative and the patient has no epidemiologic link, I would discuss discontinuation with the lab and IP team; some programs require two negative cultures. Document the result as probable colonization, not infection, and avoid using it as a clinical outcome.