The association between antibiotic use and risk of subsequent infection: a systematic review.

Problem

The microbiota lining human mucosal surfaces is thought to play an important role in defense against pathogens and in immune function. Antibiotic induced disruption of the microbiota could predispose to overgrowth and invasion by pathogens. Recent evidence of the association between antibiotic use and subsequent increased risk of meningitis raises questions about the possibility that antibiotics have long term effects which are the opposite of their intended purpose. We hypothesised that antibiotic use is associated with increased subsequent risk of infection. We carried out a systematic review of the literature for evidence of other associations between antibiotic use and subsequent infections in the community setting.

Approach

Two independent investigators systematically reviewed the literature to collect observational studies of association between antibiotic use and subsequent infection risk. Eligible studies were published before October 31, 2015, and of any design. The authors searched MEDLINE, screened titles and abstracts and extracted data in duplicate using a predefined search strategy. Studies focusing on infections caused by Clostridium difficile, drug resistant organisms and fungal organisms were excluded as these have been the subject of previous reviews and their association with prior antibiotic use is widely recognised.

Findings

The search retrieved 5580 studies of which 7 met the inclusion criteria, including 4 case-control studies, 2 cohort studies and 1 cross sectional study. In the community setting, the association between antibiotic exposure and subsequent risk of infection was reported with respect to otitis media with effusion (exposure in previous 3 months), diarrhoeal illness due to Salmonella typhimurium (exposure in previous 30 days), Staphylococcus aureus skin infection (exposure in previous 6 months), typhoid fever (exposure in previous 2 weeks), infectious mastitis (exposure during breastfeeding), upper respiratory tract infection (prior exposure of >6 weeks) and recurrent boils (exposure in 6 months period prior to first presentation with boil). Although not the focus of our review, we also found 15 studies that reported an association between prior antibiotic use and a diverse range of infections in the hospital setting.

Consequences

We have found an association between prior antibiotic use and subsequent risk of a range of infections in both the community and hospital setting. Mechanisms underlying these associations should be sought, particularly whether antibiotic induced disruption of the microbiota could be responsible. Our findings are suggestive of a previously unrecognised antibiotic related harm which is distinct from the risk of colonization and infection with drug resistant organisms. The association between antibiotic use and subsequent infection adds further weight to calls for rational prescribing of antibiotics, particularly in primary care where most antibiotic prescribing takes place.