Reducing Central Line-Associated Bloodstream Infections in the ICU: A Nurse-Led Evidence Based Quality Improvement Initiative.

The design was pre-post without a concurrent control group, so other temporal trends or hospital-wide initiatives cannot be excluded. It was done in one academic medical centre, so results may not transfer to other hospitals. The intervention included several components at once, so the contribution of the flowchart, training, audits, standardised kits or handover changes cannot be separated. Observers may have influenced nurse behaviour through the Hawthorne effect. Inter-rater reliability was not formally assessed for compliance audits. Knowledge tests were anonymous and compared at aggregate level, so individual nurse changes cannot be tracked.

Do not read the 63.2% reduction as proof that the flowchart alone prevented infections. The flowchart was introduced together with training, bedside reminders, audits, standardised kits and handover changes, and there was no concurrent control group.