Antibiotic Therapy Duration for Multidrug-Resistant Gram-Negative Bacterial Infections: An Evidence-Based Review.

Determining the optimal duration of antibiotic therapy for infections caused by multidrug-resistant Gram-negative bacteria (MDR-GNB) is a critical challenge in clinical medicine, balancing therapeutic efficacy against the risks of adverse effects and antimicrobial resistance. This narrative review synthesises current evidence and guidelines regarding antibiotic duration for MDR-GNB infections, emphasising bloodstream infections (BSI), hospital-acquired and ventilator-associated pneumonia (HAP/VAP), complicated urinary tract infections (cUTIs), and intra-abdominal infections (IAIs). Despite robust evidence supporting shorter courses (3-7 days) in uncomplicated infections caused by more susceptible pathogens, data guiding optimal therapy duration for MDR-GNB remain limited, particularly concerning carbapenem-resistant Enterobacterales (CRE), difficult-to-treat Pseudomonas aeruginosa (DTR-Pa), and carbapenem-resistant Acinetobacter baumannii (CRAB). Current guidelines from major societies, including IDSA and ESCMID, provide explicit antimicrobial selection advice but notably lack detailed recommendations on the duration of therapy. Existing studies demonstrate non-inferiority of shorter versus longer antibiotic courses in specific clinical contexts but frequently exclude critically ill patients or those infected with non-fermenting MDR pathogens. Individualised duration decisions must integrate clinical response, patient immunologic status, infection severity, source control adequacy, and pharmacologic considerations. Significant knowledge gaps persist, underscoring the urgent need for targeted research, particularly randomised controlled trials assessing optimal antibiotic duration for the most challenging MDR-GNB infections. Clinicians must navigate considerable uncertainty, relying on nuanced judgement and close monitoring to achieve successful outcomes while advancing antimicrobial stewardship goals.