Background Term and preterm neonates are at higher risk for serious adverse drug reactions than older children and adults. To date, no study has investigated spontaneous reports of adverse drug reactions (ADRs) following direct drug exposure of neonates.

Methods This is a retrospective study of all spontaneous reports of ADRs registered in the French Pharmacovigilance Database from 1986 to 2012. All reports concerning individuals from birth to ≤30 days, 1 month or 4 weeks were retrieved. ADRs classified as related to ‘pregnancy’ or ‘breastfeeding’ were discarded. Reports were described with regards to characteristics of the infant, the reported ADR(s) (registered according to the Medical Dictionary for Regulatory Activities [MedDRA]) and the reported suspected medicine(s) (registered according to the Anatomical Therapeutic Chemical [ATC] Classification) Causality assessment was performed using the French Causality assessment method.

Results A total of 1688 unique reports were analysed. More than half of these reports included at least on serious ADR (n=995; 59%). Median age at ADR occurrence was 9 days and a slight predominance of male neonates was observed (male/female ratio of 1.3). Fourteen percent of neonates presented a medical history of preterm birth, infection or genetic and/or congenital abnormality. Overall, 3127 ADRs were described in these reports. The most commonly reported system organ classes (SOCs) were injury, poisoning and procedural complications (16%), general disorders and administration site conditions (12.5%), blood and lymphatic system disorders (12%), gastrointestinal disorders (8%), investigations (8%) and nervous system disorders (8%). In the majority of ADRs reported (n=2279, 73%), infants fully recovered without sequelae but 4% of neonates deceased as a consequence of the reported ADR. Approximately 1/3 of the reported ADRs fell under 10 categories of ADRs the MedDRA classification and 1 out of 5 ADRs was coded as drug maladministration, medication error, overdose or drug maladministration without clinical consequences. A total of 2238 medicinal products out of the 2797 reported were described as suspect or interacting. Therapeutics classes must commonly incriminated were antiinfectives for systemic use and nervous system drugs. Active substances most frequently related to serious adverse reactions were zidovudine, ibuprofen and nevirapine. A total of 4103 drug-ADR pairs were described in the database; the global intrinsic causality score was ‘doubtful/unlikely’ for 35% of them and the extrinsic causality score was ‘never described before’ for only 6% of these pairs. Among the 10 most frequently encountered drug-ADR pairs, 2 active substances were mainly implicated: zidovudine and phytomedione (Vitamin K).

Conclusions We have provided a unique overview of ADR notifications after direct drug exposure in the neonates. Reporting of ADRs in neonates is scarce. The majority of ADRs are classified as serious and a subsequent proportion is associated to drug administration errors. The number of commonly incriminated drugs is limited and causality assessment is particularly challenging. There is still a need to increase awareness and education of health professionals on neonatal drug safety and to promote detailed ADR reporting in this paediatric population.