Factors associated with antibiotic use in children hospitalized for acute viral gastroenteritis and the relation to rotavirus vaccination

Evidence on unnecessary antibiotic use in children with acute viral gastroenteritis (AGE) is scarce. We characterized the extent and correlates of antibiotic use among children hospitalized with viral AGE. A single-center study enrolled children aged 0-59 months hospitalized for AGE between 2008 and 2015 in Israel. Information was collected on laboratory tests, diagnoses, antibiotic treatment, and rotavirus vaccination. Stool samples were tested for rotavirus antigen, GII-norovirus, and stool cultures were performed for bacterial enteropathogens. Data from 2240 children were analyzed. Rotavirus vaccine was given to 79% of eligible children. Rotavirus test was performed on 1419 (63.3%) children. Before the introduction of universal rotavirus vaccination (2008-2010), rotavirus positivity in stool samples was 37.0%, which declined to 17.3% during the universal vaccination years (2011-2015). Overall, 1395 participants had viral AGE. Of those, 253 (18.1% [95% CI 16.1-20.2]) had unnecessary antibiotic treatment, mostly penicillin 46.6%, ceftriaxone 34.0% and azithromycin 21.7%. A multivariable analysis showed an inverse association between rotavirus vaccination and unnecessary antibiotic treatment (odds ratio = 0.53 [95% CI 0.31-0.91]), while positive associations were found with performing chest-X-ray test (3.00 [1.73-5.23]), blood (3.29 [95% CI 1.85-5.86]) and urine cultures (7.12 [3.77-13.43]), levels of C-reactive protein (1.02 [1.01-1.02]) and leukocytes (1.05 [1.01-1.09]). The results were consistent in an analysis of children with laboratory-confirmed rotavirus or norovirus AGE, or after excluding children with CRP > 50 mg/L. In conclusion, antibiotic prescription was common among hospitalized children with viral AGE, which was inversely related to rotavirus vaccination, possibly due to less severe illness in the vaccinated children.