Avaliação de gravidade e evitabilidade de reações adversas a medicamentos suspeitas em neonatos: desempenho das ferramentas de Hartwig e Bracken

Abstract

Introduction: Neonates in neonatal intensive care units (NICUs) are highly susceptible to adverse drug reactions (ADRs) due to the complexity of pharmacotherapy, physiological immaturity, and extensive off-label drug use. Therefore, it is crucial to characterize ADRs in terms of severity and avoidability to ensure the health and safety of neonates. However, the reliability of ADR classification tools in the NICU context remains insufficiently established. Objective This study aimed to evaluate the inter-rater reproducibility of tools used to characterize the severity and avoidability of ADRs in NICUs. Methods: An observational and prospective study was conducted between January 2019 and December 2020 in a referral NICU for high-risk pregnancies in Natal, Brazil. The Hartwig tool was employed for assessing severity, while the Bracken tool was used to evaluate avoidability. Three experienced clinical pharmacists independently assessed all ADR cases. Inter-rater reproducibility was measured using Cohen's kappa coefficient (k) with corresponding 95% confidence intervals (CI). Results: Among 79 ADR cases, the mean gestational age was 29.7±4.4 weeks, and the birth weight averaged 1,446.0±1,179.3g. The assessment of ADR severity using the Hartwig tool revealed a significant overall correlation (overall k = 0.573; 95% CI 0.409-0.737) with an 86.5% agreement rate and 68.3% discordance among raters. However, no statistically significant correlation was observed for determining avoidability using the Bracken tool (overall k = 0.007; 95% CI -0.046-0.022), with an inter-rater discordance rate of 83.3%. Conclusion: The Hartwig tool demonstrates good reproducibility among different evaluators in determining the severity of ADRs, unlike the Bracken method for assessing avoidability.