The Quality Dashboard That Improved Without Improving

I treat this as a measurement validity question, not a loyalty question. The nurse manager may be right, and the board may be about to celebrate a denominator artifact. I would investigate quietly and quickly:

Confirm/refute tests:

• Denominator check: Plot catheter-days and device utilization ratio alongside the rate. If catheter-days fell sharply, the rate change may be real (fewer devices = fewer infections) or artifact (fewer at-risk days).
• Numerator check: Plot absolute CAUTI counts, not just rates. If counts are flat while the rate fell, the denominator moved.
• Culture-collection audit: Pull 20-30 charts from the pre- and post-period with urinary symptoms and compare whether cultures were sent. If the lab criteria change reduced culturing of symptomatic patients, that is a measurement shift, not a care improvement.
• Case review: Review every CAUTI case in both periods for appropriateness of culture and application of the surveillance definition (NHSN). Misapplied definitions inflate or deflate rates.
• Cross-check: Look at other device-related measures (CLABSI, VAP) and antibiotic days of therapy for UTI. If UTI treatment did not fall, infections likely did not fall.

I would ask infection prevention and the lab to co-present the methodology, not to defend it. The framing to them: 'We need to be sure the board is celebrating something real.'

I separate the internal correction from the external communication and move fast on both.

1. Brief the CEO first, privately, within 48 hours of my preliminary finding - before the board meeting and before any further public statements. The message: 'I have a measurement concern about the CAUTI number we cited. I am validating it now. I want you to hear it from me before the board does.' This protects the CEO and the relationship.
2. Do not retract publicly until validated. If the investigation confirms a denominator/culture shift, the correction is a factual update ('our surveillance methodology changed; the underlying infection rate is being re-baselined'), not a confession. If it refutes the concern, no correction is needed and I have protected the CEO from a false alarm.
3. Prepare a board-ready explanation with the methodology change disclosed, re-baselined trend, and a corrective action (standardized culture criteria, surveillance validation).
4. Address the newsletter with the CEO and communications: a brief, non-defensive clarification if warranted. The reputational risk of being seen to hide a measurement problem is larger than the risk of a correction.

Principle: transparency about measurement is a credibility asset, not a liability. The board and community forgive a corrected metric; they do not forgive a concealed one.

Sepsis mortality is a patient-safety signal and outranks a measurement-quality issue in urgency, even though the CAUTI item has a board deadline. I would run them in parallel with different owners and different clocks:

• CAUTI: I own the validation, three-week board timeline, as above.
• Sepsis: I ask the quality director to convene a rapid case review within two weeks - every sepsis death in the worsening period, with a standardized tool (timeliness of recognition, lactate, antibiotics, fluids, escalation). I add sepsis as a standing agenda item immediately, not next quarter.

Sequencing logic: The sepsis index at 1.15 may reflect case-mix, coding, or care. I must know which before I speak to the board. If it is care, the board needs to hear it from me with an action plan, not discover it in a chart pack. If it is coding/case-mix, I need that documented so the committee does not chase a phantom.

I would tell the committee chair: 'We have two signals - one about whether our CAUTI number is real, one about whether our sepsis care is safe. I am treating the second as the higher priority and will bring both to the board with a clear distinction between measurement and outcome.' That framing keeps the committee focused on patients rather than on the newsletter.