Penicillin V versus amoxicillin for pneumonia in children-a Swedish nationwide emulated target trial

Abstract

Objectives: Although most countries recommend amoxicillin for paediatric pneumonia, there is a long tradition of treatment with penicillin V (PcV) in Sweden, thus not empirically covering Haemophilus influenzae. There are, however, large regional differences in treatment practice. The aim was to compare clinical outcomes (treatment failure and severe complications), in children aged 1-59 months treated with PcV vs. amoxicillin for pneumonia.

Methods: This population-based emulated target trial included all children born in Sweden between 2001 and 2021, using national health, sociodemographic, and population registers. All pneumonia cases from hospitals and paediatric outpatient clinics in children aged 1-59 months treated as outpatients with PcV or amoxicillin between July 2005 and December 2021, were identified. Adjusted odds ratios (aORs) and 95% confidence intervals (CIs) for treatment failure (newly dispensed antibiotic prescription or pneumonia-associated hospitalization day 1-14) and severe complications (lung complications, an invasive bacterial disease, admission to intensive care unit or death day 1-28) were calculated with logistic regression analysis.

Results: PcV was prescribed in 14 766 cases and amoxicillin in 10 566. Treatment failure occurred in 7.7% with PcV vs. 4.7% with amoxicillin, aOR 1.76 (95% CI: 1.54-2.00). Severe complications were rare, with no significant difference between PcV and amoxicillin (0.3% vs. 0.2%, aOR 0.96, 95% CI: 0.53-1.73). Sensitivity and interaction analyses showed consistent results.

Discussion: PcV treatment compared with amoxicillin, was associated with an increased risk for treatment failure but not for severe complications. The absolute risks for adverse outcomes were low in both groups suggesting a minor role of H. influenzae in paediatric pneumonia.

Keywords: Amoxicillin; Antibiotics; Children; Paediatric infectious diseases; Penicillin; Pneumonia.