Early oral stepdown antibiotic therapy versus continuing intravenous therapy for uncomplicated Gram-negative bacteraemia (the INVEST trial): study protocol for a multicentre, randomised controlled, open-label, phase III, non-inferiority trial

Background: The incidence of Gram-negative bacteraemia is rising globally and remains a major cause of morbidity and mortality. The majority of patients with Gram-negative bacteraemia initially receive intravenous (IV) antibiotic therapy. However, it remains unclear whether patients can step down to oral antibiotics after appropriate clinical response has been observed without compromising outcomes. Compared with IV therapy, oral therapy eliminates the risk of catheter-associated adverse events, enhances patient quality of life and reduces healthcare costs. As current management of Gram-negative bacteraemia entails a duration of IV therapy with limited evidence to guide oral conversion, we aim to evaluate the clinical efficacy and economic impact of early stepdown to oral antibiotics. Methods: This is an international, multicentre, randomised controlled, open-label, phase III, non-inferiority trial. To be eligible, adult participants must be clinically stable / non-critically ill inpatients with uncomplicated Gram-negative bacteraemia. Randomisation to the intervention or standard arms will be performed with 1:1 allocation ratio. Participants randomised to the intervention arm (within 72 h from index blood culture collection) will be immediately switched to an oral fluoroquinolone or trimethoprim-sulfamethoxazole. Participants randomised to the standard arm will continue to receive IV therapy for at least 24 h post-randomisation before clinical re-assessment and decision-making by the treating doctor. The recommended treatment duration is 7 days of active antibiotics (including empiric therapy), although treatment regimen may be longer than 7 days if clinically indicated. Primary outcome is 30-day all-cause mortality, and the key secondary outcome is health economic evaluation, including estimation of total healthcare cost as well as assessment of patient quality of life and number of quality-adjusted life years saved. Assuming a 30-day mortality of 8% in the standard and intervention arms, with 6% non-inferiority margin, the target sample size is 720 participants which provides 80% power with a one-sided 0.025 alpha-level after adjustment for 5% drop-out. Discussion: A finding of non-inferiority in efficacy of oral fluoroquinolones or trimethoprim-sulfamethoxazole versus IV standard of care antibiotics may hypothetically translate to wider adoption of a more cost-effective treatment strategy with better quality of life outcomes.