Antibiotic Treatment of Common Infections

More Evidence to Support Shorter Durations

Benjamin J. Smith; George Heriot; Kirsty Buising


Curr Opin Infect Dis. 2020;33(6):433-440. 

In This Article

Sepsis and Bacteraemia

Clinician surveys have identified significant heterogeneity in self-reported treatment durations for bacteraemia.[2–4] This variability may reflect the paucity of evidence to support decision-making. In general, the treatment approach should consider the pathogen, the likely source of infection and any metastatic complications. Particular caution should be exercised in patients with the potential for seeding to prosthetic material.

S. aureus bacteremia (SAB) is typically stratified into 'uncomplicated' and 'complicated' infection, with minimum treatment durations of two and four weeks, respectively.[5] Trials are currently underway to investigate the safety of shorter durations of intravenous therapy for SAB in low-risk patients but at present there is insufficient evidence to support such an approach.[6,7] Evidence for treatment duration of other gram-positive bloodstream infections (e.g. enterococci) is limited to expert opinion.

The optimal treatment duration for gram-negative bacteraemia is dependent on the underlying source of infection. A recent, multicentre RCT in patients with source-control and an early clinical response to therapy demonstrated that seven days of antibiotic therapy is noninferior to 14. Mortality at 90 days exceeded 10% in both trial arms but reflected the inclusive selection criteria (including both elderly and immunocompromised patients).[8] These findings have been supported by a Swiss RCT with similar inclusion criteria that compared 7, 14 and CRP-guided treatment duration.[9] Of note, the median treatment duration in the CRP-guided arm (in which patients received antibiotics until afebrile for 48 h and CRP had decreased 75% from its peak) was seven days. Patients with bacteraemia due to Pseudomonas aeruginosa were not well represented in the first of the above studies, and excluded in the second; therefore, short-course therapy should be considered cautiously for these individuals. Evidence for treatment of Pseudomonas bloodstream infections is limited to observational data.[10]

Current recommendations for treatment of invasive meningococcal disease (encompassing bacteraemia and meningitis due to Neisseria meningitidis) are for a minimum of five days of therapy. However, there have been a number of published observational studies suggesting that treatment durations of three or four days are as effective as longer courses and are associated with very low rates of relapse.[11,12]

Specific guidelines exist for the management of bloodstream infection associated with intravascular catheters. Recommended treatment durations are based on the isolated bacteria, presence or absence of complications and the approach to management (i.e. catheter removal or retention). Advice is based on expert opinion as no randomized trials have evaluated antibiotic treatment durations.[13]

For patients with fever during chemotherapy-induced neutropaenia, a definitive focus of infection is often not determined, and broad-spectrum antimicrobial therapy is typically continued for the duration of neutropaenia. Current guidelines recommend treatment until neutrophil recovery to greater than 0 × 5 × 109 cells/l.[14] However, a recent Spanish RCT in patients with high risk febrile neutropaenia has shown that stopping treatment when patients have resolution of signs and symptoms of infection (regardless of neutrophil count) can reduce total antibiotic exposure (mean 16.1 versus 13.6 antibiotic-free days, p = 0.026) with no difference in mortality or adverse events.[15]