Background This is a "proof-of-concept" study aiming to evaluate the impact of a multistep bundles intervention in the management and outcomes of patients with gram-negative bloodstream infections (GN-BSIs). Methods This was a single-center, quasi-experimental design study. In the pre-phase (January 2019 to May 2020), patients were retrospectively enrolled. During the post-phase (June 2020 to September 2021), all patients were prospectively enrolled in a nonmandatory 3-step bundles intervention arm including (i) step 1: imaging to detect deep foci of infection, follow-up blood cultures and procalcitonin monitoring; (ii) step 2: early targeted antibiotic treatment and surgical source control; (iii) step 3: discontinuation of antibiotics within 7-10 days in case of uncomplicated BSI. Patients were followed up to 28 days from BSI onset. The primary outcome was 28-day mortality. Results A total of 271 patients were enrolled: 127 and 144 in the pre- vs post-phase, respectively. Full application of step 1 (67% vs 42%; P < .001), step 2 (83% vs 72%; P = .031), and step 3 (54% vs 2%; P < .001) increased in the post-phase. Overall, the intervention reduced 28-day mortality (22% vs 35%, respectively; P = .016) and the median duration of total (11 vs 15 days; P < .001) and targeted (8 vs 12 days; P = .001) antibiotic therapy. Finally, the multivariate Cox regression confirmed the independent protective effect of adherence to step 1 (adjusted hazard ratio [aHR], 0.36; 95% CI, 0.20-0.63) and step 2 (aHR, 0.48; 95% CI, 0.29-0.81) on risk of 28-day mortality. Conclusions Clinical management and outcomes of patients with GN-BSIs may be improved by providing a pre-established multistep bundles intervention. A multi-step bundles strategy for management of Gram-negative bloodstream infections was established in our hospital. It was effective in reducing mortality, duration of antibiotic treatment and adverse events to antibiotics, without increasing the risk of recurrence/relapse of infections
