The Impact of Prescribing Antibiotics to Children

Abstract & Commentary

By Hal B. Jenson, MD, Chief Academic Officer, Baystate Health Professor of Pediatrics and Dean of the Western Campus of Tufts University School of Medicine. Dr. Jenson is on the speaker's bureau for Merck.

Synopsis: Prescribing β-lactam antibiotics for children with upper respiratory tract infections is associated with in a 3-fold increase in the MIC for ampicillin and 2-fold increase of the ICEHin1056 resistance gene of oropharyngeal Haemophilus isolates. The effect in individuals is transient (less than 12 weeks) but is sustained in the population.

Source: Chung A, et al: Effect of antibiotic prescribing on antibiotic resistance in individual children in primary care: Prospective cohort study. BMJ. 2007;335:429.

An observational cohort study in the united Kingdom of 119 children 6 months to 12 years of age with upper respiratory tract infections (either otitis media or presumed viral infections) compared antibiotic resistance of oropharyngeal Haemophilus isolates among 71 children who received a β-lactam antibiotic (amoxicillin, 70; cephradine, 1) and 48 who did not receive any antibiotic. Throat swabs were obtained at 0, 2, and 12 weeks, and plated onto 2 Haemophilus selective media plates. Antibiotic resistance was measured by 1) the geometric mean minimum inhibitory concentration (MIC) for ampicillin; and 2) the presence of the integrative and conjugate element ICEHin1056, which encodes β-lactamase among nasopharyngeal Haemophilus species.

The antibiotic and no antibiotic groups were similar in mean age (each 5.4 years), mean number of children in the household (1.16 vs 1.13), and previous exposure to antibiotics at any time (87% vs 85%), or within 3 months (13% vs 20%). Daycare or school attendance was more common among children who received an antibiotic (96% vs 85%, P = 0.05). The initial MIC for ampicillin was 2.4 m/mL in the antibiotic group and 4.1 µ/mL in the no antibiotic group (P = 0.24), and the proportion of children initially with ICEHin1056 was 32% vs 38%, respectively (P = 0.52). The presence of ICEHin1056 was high even among children who had never received an antibiotic (8 of 15, 53%).

At 2 and 12 weeks, the MIC for ampicillin in the no antibiotic group was 2.7 µg/mL at each visit, which was not significantly different from the baseline of 4.1 µg/mL. Among children who had received an antibiotic, the MIC increased to 9.2 µg/mL at 2 weeks (P = 0.005), 3.5 times higher than for the no antibiotic group, and at 12 weeks, the MIC fell to 5.7 µg/mL (P = 0.06).

There was no significant difference in the prevalence of ICEHin1056 homologues at 2 and 12 weeks in the no antibiotic group, which was 36% and 37% compared to 38% at baseline. In the antibiotic group, the prevalence doubled to 67% at 2 weeks (P = 0.002) and fell at 12 weeks to 36%, close to the baseline of 32%. However, ICEHin1056 was identified in Haemophilus isolates from most children of both groups (83%, 95% confidence interval, 76-89%) on at least one occasion.


This community study confirms that the short-term impact of prescribing antibiotics to children includes a 3-fold increase in the MIC for ampicillin among oropharyngeal Haemophilus species and a 2-fold increase, to 67%, in the prevalence of the ICEHin1056 resistance element. These changes are transient, and the prevalence of resistant Haemophilus returns to baseline within 12 weeks. These changes indicate that a β-lactamase resistant antibiotic, such as amoxicillin-clavulanate, should be prescribed if antibiotic therapy is indicated for a child who has received a β-lactam antibiotic within 12 weeks.

The greater concern is the high equilibrium of resistant Haemophilus species in the community, which in this study was at a baseline of 32-37%. Significantly, resistant Haemophilus species were recovered from 83% of all children at some time. It appears that although the adverse impact of antibiotics on an individual child is transitory, there is sufficient pressure to sustain a high level of antibiotic resistance in the population. Two potential counter responses include developing guidelines for prescribing shorter courses of antibiotics, and more importantly, not prescribing antibiotics for children with presumed viral upper respiratory tract infections.