Factors predicting amoxicillin prescribing in primary care among children: a cohort study

Background: Antibiotic prescribing during childhood contributes to antimicrobial resistance, which is a major public health concern. Antibiotics are most commonly prescribed to children for respiratory tract infections (RTI). Aim: To identify factors associated with amoxicillin prescribing and RTI consultation attendance in primary care in young children. Design and Setting: Cohort study in Bradford with data from pregnancy to age 24-months collected between 2007-2013, linked to electronic primary care and air pollution data. Methods: We calculated amoxicillin prescribing rates/1, 000 child-years, and fitted mixed-effects logistic regression models, with general practice (GP) surgery as the random effect, to establish risk factors for amoxicillin prescribing and RTI consultation during the first two years. Results: Among 2, 493 children, the amoxicillin prescribing rate was 710/1, 000 child-years during the first year (95% CI: 677-744) and 780/1, 000 (745-816) during the second year. Odds of amoxicillin prescribing during year one were higher for infants who were male (adjusted OR 1.4 (1.1-1.6)), socio-economically deprived (1.4 (1.0-1.9)), and with a Pakistani ethnic background (1.4 (1.1-1.9)). Odds of amoxicillin prescribing during the second year were higher for infants with a Pakistani ethnic background (1.5 (1.1-2.0)/1.6 (1.2-2.0)) and pre-/early-term infants (1.2 (1.0-1.5)). Additional risk factors included caesarean delivery, congenital anomalies, household overcrowding, birth season, and formal childcare attendance. GP surgery-level variation explained 7-9% of variation in amoxicillin prescribing. Conclusions: Socio-economic status and ethnic background are strongly associated with amoxicillin prescribing and RTI consultations during childhood. Interventions reducing RTI spread in household and childcare settings may reduce antibiotic prescribing in primary care.