Six CLABSIs in Nine Months: Cluster, Coincidence, or Measurement Artifact
Mercy West's 24-bed adult ICU recorded 6 central line-associated bloodstream infections in the last 9 months (1.9 per 1,000 central line days) against an NHSN risk-adjusted baseline of 0.6. Bundle compliance audits show 91% overall, but head-of-bed elevation and dressing integrity are audited only by self-report. Three of the six patients had lines inserted in the emergency department rather than the ICU. In March the system switched to a new supplier of chlorhexidine-impregnated dressings, and in April the laboratory moved to a more sensitive blood culture system. Two of the six cases occurred in the same 4-bed pod, and the ICU nurse educator position has been vacant for five months. The infection preventionist wants to declare a cluster and report it to the board; the CNO is worried about public reporting and the pending Joint Commission survey.
1. Before concluding that this is a true increase in infection, what data validation and epidemiologic steps are needed?
Read the worked answer
Treat the rate as a hypothesis, not a finding.
- Verify case classification. A more sensitive blood culture system increases detection of low-grade bacteremia and of organisms that must meet stricter NHSN criteria, such as commensal skin flora requiring two cultures or clinical signs. Have a second infection preventionist independently re-review all six charts against the current protocol; a single reviewer's six can become three or four.
- Verify the denominator. Central line days should come from device documentation, not a fixed estimate; a census or device-use change can move the rate without any change in infections.
- Use statistical process control. With roughly 24 lines and nine months of data, the expected count is about 1.5 to 2. Plot a u-chart before declaring special-cause variation; small denominators produce unstable rates.
- Rule out a pseudo-outbreak. Check the chlorhexidine dressing lot numbers and any manufacturer notices, laboratory contamination, and the skin-prep product.
- Look for a common cause. The pod clustering, three ED insertions, and the vacant educator position are plausible system contributors pointing to insertion and maintenance practice, not hand hygiene alone.
Do not pause basic prevention while analyzing; it is safe to strengthen obvious controls in parallel with the review.
2. Assuming the increase is real, what should the improvement approach be and who should own it?
Read the worked answer
Run a two-layer analysis and fix the bundle where it is actually failing.
Analysis. Perform an RCA-squared on each case to identify causal and contributing factors, then a common-cause analysis across all six: insertion site (ED versus ICU), operator, kit, dressing, bathing compliance, and line duration.
Interventions:
- Target ED insertions. Standardize an ED insertion kit, a checklist read aloud, a trained inserter, and stop-the-line authority; audit ED insertions daily for a month.
- Replace self-reported audits with direct observation of maintenance: dressing integrity, tubing changes, hub disinfection, and daily necessity review. Self-report is the most likely reason the bundle reads 91% while infections continue.
- Re-establish bathing and dressing standardization after the supplier change, retrain with return demonstration, and verify product compatibility.
- Fill or cover the educator role or assign unit-based champions; the vacancy maps onto the period of increase.
- Use the Model for Improvement: an aim of zero CLABSI over 12 months, a family of measures (observed insertion compliance, observed maintenance compliance, CLABSI per 1,000 line days), weekly PDSA cycles, and a multidisciplinary team with ED representation reporting to the quality committee.
Ownership should sit with a physician-nurse pair with an executive sponsor, not with infection prevention alone, because most of the controllable failure points sit in clinical workflow.
3. What are the reporting and disclosure obligations, and how should the board's quality committee be briefed?
Read the worked answer
Separate the three obligations, because the CNO's concern conflates them.
- Internal reporting: report now. A possible cluster is a board-level quality issue regardless of external reporting status, and the committee should see the case list, the validation results, the improvement plan, and the resources required.
- External reporting: NHSN CLABSI reporting feeds CMS quality programs and is required in most states. Confirm the state requirement and submit per protocol. Deliberately under-reporting to protect a survey is a compliance risk larger than any citation.
- Disclosure to affected patients: the six patients or their families should be told what happened and what is being done, consistent with the organization's disclosure policy. Several states require this, and it is the ethical baseline regardless.
- Survey readiness: Joint Commission surveyors look for evidence that the organization detected, analyzed, and acted. A documented cluster with a disciplined response is a strength; an undocumented one is the finding.
Brief the committee in four parts: what is known and what remains uncertain, the validation result, the action plan with owners and dates, and the resources required. Requesting the educator position at that same meeting is more persuasive than requesting it later, and it ties the request to a documented gap rather than a general appeal.