Should I take the full course of antibiotics?

The ubiquitous belief that a course of antibiotics should be taken until finished has its origins in the Nobel prize speech in 1945 by John Fleming when he advocated: “If you use penicillin, use enough!” This mantra has become embedded in national and international public awareness antibiotic programs. However, the research does not support this decree in all clinical cases.

For example, uncomplicated community-acquired pneumonia (CAP) in children treated with either cephalosporins or penicillins showed no difference in treatment failure with a median 6 days versus a median 10 days and readmission rates for CAP for 5 versus 10-day therapy were significantly lower with the short-course.

Similarly, a median treatment of 9 days for P. aeruginosa bacteraemia showed no significant difference in the 30-day reoccurrence risk, compared to a median treatment of 15 days (HR = 0.68, 95% CI =0.34-1.36, P = 0.28) whilst a short course (6-10 days) treatment in low-risk methicillin-susceptible Staphylococcus aureus bacteraemia (n=141) showed no significant difference in 90-day mortality when compared to long course treatment (11-16 days).

For uncomplicated gram-negative bacteraemia (GNB), a randomised controlled trial (n=604) showed that a 7-day course was non-inferior to 14-day treatment when assessed for clinical failure, readmissions, extended hospitalisation at 90 days or mortality. Similarly, a systematic review and meta-analysis by Li et al. (2021) reported no significant differences for 30-day mortality or recurrent bacteraemia, 90-day mortality or recurrence, adverse events, Clostridium difficile infection or the development of resistance for short versus long-term treatment for uncomplicated GNB. As a summary of antibiotic use for GNB, Le Fevre has noted that the personalising and individualising of antibiotic treatment duration is “… a promising approach, in need of further research.”

A more nuanced approach is required for certain types of infections, with a shorter duration of therapy for uncomplicated intra-abdominal infections such as appendicitis, but a longer duration for complex infections such as faecal peritonitis.

A problem with prolonged antibiotic use is the induction of opportunistic overgrowth. Tan et al. (2021) reported that superinfection rates occur more regularly when patients received long-term treatment of non-ventilator associated hospital-acquired pneumonia, compared to short-term treatment (6.3% v. 18.2% p= 0.027).

Exceptions to the “shorter is better” approach involve persistent infections such as M. tuberculosis, H. pylori, Salmonella Typhi, T. pallidum, S. aureus (cystic fibrosis patients with bronchiectasis or pneumonia, or device-associated infections) and M. leprae. There are also some infections, for example osteomyelitis, where a shorter duration increases the risk of relapse. Intravenous antibiotics are often recommended for at least 4-6 weeks after surgical debridement, followed by 3-6 months of oral therapy.

Bacterial resistance is multilayered and can occur either through natural resistance or acquired resistance. The latter involves horizontal gene transfer (HGT) via transformation, transposition, and conjugation. HGT enables bacteria to rapidly change in response to shifting environmental circumstances, thereby providing an enhanced survival advantage. For example, HGT of vanB determinants occurs in hospital E. faecalis and E. faecium, and multi-drug resistant vancomycin-resistant enterococci (VRE) can transfer genetic determinants to methicillin-resistant S. aureus (MRSA). The inappropriate use of antibiotics has also allowed P. aeruginosa to acquire resistance via chromosomal mutations and acquisition of antibiotic resistance genes via HGT.

In summary, improper antimicrobial stewardship, including sub-therapeutic dosing, has resulted in the current problem of universal antimicrobial resistance. Therefore, reducing unjustified antibiotic use is vital to mitigating bacterial resistance. Longer duration of exposure to antibiotics for opportunistic bacteria (E. coli, E. faecium, S. aureus, K. pneumoniae, Pseudomonas spp, Acinetobacter spp and Enterobacter spp) increases both resistance and transfer of resistant strains between asymptomatic patients. Paradoxically and counter-intuitively, less is better for most patients, as a reduced use of antibiotics lowers rather than increases resistance.