Reading Your Own Run Chart: A CAUTI Project with an Uncomfortable Balancing Measure

Statistically, the process changed. Clinically, the claim is still open - and the balancing signals may be telling you why.

Chart interpretation:

• Eight consecutive points below the baseline median is a "shift" - a recognized special-cause rule. It is unlikely to be random variation.
• One point below the lower control limit is a second special-cause signal, but a single extreme point can also reflect a data artifact.
• Together these support a real reduction in catheter utilization.

Before claiming success, verify:

1. Data integrity and definitions. Did the denominator change (patient days, observation patients)? Were catheter-days counted consistently? Did the removal protocol introduce under-documentation rather than true removal?
2. Case ascertainment. Is the unit culturing less often because fewer catheters exist, or missing CAUTIs in patients whose catheters were removed and reinserted?
3. Outcome lag. CAUTI counts are few and noisy; process improvement should lead outcome improvement by months.
4. The balancing measures are the story. A 9-point drop in responsiveness plus nurses being overridden by an uninformed surgeon suggests protocol-driven removal without clinical context - possible urinary retention, reinsertion, immobility, or unmanaged toileting needs. Reinsertion rate and retention/bladder-scan data are essential.

Say it plainly: The intervention is reducing catheter exposure as intended. We have not yet demonstrated a CAUTI reduction, and we have a signal of harm that must be addressed before scaling.

Treat this as a sociotechnical design flaw, not a discipline problem.

Framing: the protocol gave nurses authority without giving the surgeon information or giving nurses a safe path when removal is clinically wrong. Both sides are behaving reasonably inside a badly designed system.

Actions:

1. Meet the surgeon individually, not in a group email. Bring the utilization data and ask what went wrong clinically. Likely he encountered a patient who needed the catheter and was never consulted.
2. Co-design exceptions with him. Named criteria that require surgeon or urology sign-off before removal - recent urologic or pelvic surgery, epidural/spinal anesthesia with planned monitoring, strict output requirement, acute retention history, sacral wound with contamination risk.
3. Keep nurse authority but add a shared decision step - the necessity huddle should include the surgeon's documented plan; if the plan is absent, the nurse huddles with the charge nurse before removal.
4. Check the nurses' experience of speaking up. Were they punished, ignored, or just overruled without discussion? Debrief both nurses, protect them from retaliation, and ask what would have made speaking up workable.
5. Add the missing measures: reinsertion within 48 hours, bladder-scan retention events, and patient-reported toileting assistance.

If the drop in HCAHPS responsiveness tracks with patients waiting for toileting help after removal, that is a staffing/workflow problem the protocol exposed - not proof the protocol was wrong.

Lead with the balanced picture and the caveats. Committees trust the presenter who names the weakness before the audience does.

Presentation structure:

1. The problem and the aim - one sentence, with baseline (SIR 1.4, three CAUTIs in six months).
2. The intervention, dated and annotated on the SPC chart. Never show a pre/post bar chart when you have an SPC chart.
3. The result: a shift in catheter utilization - stated as process improvement, not as "we cut CAUTIs by X%."
4. The outcome measure: report CAUTI counts with the explicit caveat that they are too few and too lagged to conclude anything yet; state the date you will re-evaluate.
5. The balancing measures, presented with equal prominence - HCAHPS responsiveness, reinsertion rate, retention events, staff-reported barriers, the surgeon's concern.
6. The corrective actions taken - exception criteria, surgeon engagement, huddle redesign.
7. The ask - resources, a data analyst, or a decision to spread. Be specific.

Practical tips:

• Bring an annotated chart with the intervention and the balancing-measure dip marked, so the relationship is visible rather than defensible.
• Avoid "the data show nurses were removing too many catheters." Say "the protocol lacked a consultation step."
• Pre-write the one-sentence takeaway you want in the minutes: "Utilization improved with special-cause signal; outcome not yet demonstrated; a balancing-measure signal has been identified and addressed."