Roadmap to optimal pharmacovigilance practice in neonatal intensive care units.

In the United States, critically sick neonates treated in neonatal intensive care units (NICUs) contribute to less than 3% of total hospital occupancies and less than 2% of total annual health care cost for their immediate hospital stay after birth.1 Despite these apparently low numbers, the long-term challenges in the care of surviving sick neonates with their complications, from birth to adult life, result in much larger economic consequences. Current data suggest that the total cost of preterm birth, taking into consideration the entire childhood period, could be as high as $4.5 billion.2 Besides the individual challenges, available evidence has shown that parents of children with neurodevelopmental impairment have a significantly higher rate of mental health illness, with up to 80% risk of marital discord further complicating the societal burden. Multiple factors have an impact on the outcome of the critically ill neonate of which many are hard to mitigate. Yet promotion of translation of evidence into clinical practice in NICUs is at least one critical step forward so that improvement in the outcome can be made and a meaningful benefit at an individual, family and societal level can be achieved. Among all areas of evidence-to-practice gaps in the field of neonatal medicine, effective and safe drug therapy stands out. For over two decades, regulatory agencies around the globe have put initiatives in place for the provision of incentives or mandates on the production of rigorous therapeutic information for children.3 Despite these efforts, data related to neonatal population, both for novel or existing therapeutics, remains very limited, with subsequent failure in provision of effective therapy or occurrences of drug-related adverse events (AEs). Lack of proper pharmacotherapy data in the neonatal population is a known and important factor contributing to poor short- and long-term outcomes and possible lifelong physical and mental health compromise.4 If the journey to achieve a neonatal specific drug development program is long and not immediately feasible, closing the gap in detection, assessment, understanding and prevention of drug-related AEs appears an urgent and attainable goal. Many factors put the sick neonate treated in the NICU at a significant risk for drug-related AEs. The complex processes of growth and development affect both drug pharmacokinetic (PK) and pharmacodynamic (PD) processes. Prescription of off-label and unlicensed drugs is common practice in the NICU. Lack of proper dosing guidelines and labelling information for even the most frequently used medications in the NICU results in considerable risk of drug-dose-related AEs.5 The Generally Regarded as Safe (GSA) status of excipients does not apply to neonates with rapid maturational and physiological changes. Excipients of high concern, such as ethanol and propylene glycol, are present in many commonly prescribed medications in the neonatal population, exposing them to risk of serious AEs.6 The risk of drug-drug interaction (DDI) in sick neonates with high rate of polypharmacy is well known and while DDI adult data is usually sourced for the NICU population, the ontogeny of cytochrome P450 enzyme activity extremely limits the value of this approach.7 While it is essential to make every effort to increase available PK and PD data in the neonatal population through prospective studies and innovative quantitative approaches, identifying the limitations of the current practice, such as identification and quantification of adverse drug reactions, appears to be an urgent need. For years, the high background rate of mortality and morbidity and difficulty in differentiation of patient's primary disease signs and symptoms from a potential avderse drug reaction have resulted in reluctance of NICU clinicians to include drug safety evaluation in daily practice. However, if recent developments in the analytical and structural approach to drug AE detection and assessment in the neonatal population can be translated to the bedside, the provision of pharmacovigilance practice can be potentially addressed in a rather straightforward manner. In this commentary we will provide our recommendations for an optimal roadmap to address this important gap in the state-of-the-art care of neonatal patients. For timely and effective detection of drug-related AEs in neonatal populations, with a high background rate of mortality and morbidity and also an extensive variability in commonly used biomarkers, access to standard terminology for the description of the same kind of event is the first key step. In 1983, the National Cancer Institute implemented the first system for standardised terminology and grading of drug-related AEs of cancer treatments.8 Since then, this system has substantially evolved with the development of a specific paediatric AE terminology list with a unique temporal focus on the paediatric population, including the foetal and neonatal period. In 2019, the International Neonatal Consortium (INC) implemented a consensus process that resulted in the development of neonatal generic adverse event standard severity scaling (NAESS) applicable to any drug-related AE in the neonatal population and also specific severity criteria for 35 most commonly experienced neonatal AEs.9 These efforts provide an excellent platform for the development of a comprehensive but focused foetal and neonatal standardised terminology list that includes all applicable AE terminology, specific severity gradings and links to the corresponding MedDRA Lowest Level Terms. Regulatory agencies have provided strict definitions for the “seriousness” of a drug-related AE, which can be adapted to the neonatal population in their current format. In contrast, assessment of a causal relationship between a drug and an AE has been broadly variable from clinical judgment to algorithms and probabilistic methods. An effective causality assessment of captured drug-related AEs is an important part of any robust pharmacovigilance system. For neonatal patients, the traditional casualty assessment tools have shown suboptimal intra-rater reliability, with the need for further development and adaptation to this specific population. In 2012, Du et al reported the development of a new algorithm using actual patient data to examine the relatedness of the captured drug-related AEs in the NICU population.10 This new algorithm, although in need of further validation, showed higher reliability than the available traditional tools. Although confirmation of the reliability and validity of this newly developed tool is of utmost importance, until such data exist it appears prudent to make the best use of this readily available knowledge for the postmarketing surveillance of at least the most frequently used drugs in the NICU. For the large majority of critically ill neonates, the list of prescribed drugs is limited to 50 medicinal products that have been on the market for a number of years.11 Most of these drugs, however, are prescribed in an off-label manner with a scarcity of clinical trials on efficacy, dosage and safety. Implementation of a drug-related AE monitoring and reporting system, using the best currently available tools, will at a minimum provide the basics of good pharmacovigilance practice for these most vulnerable patients. Furthermore, unless pharmacovigilance practice becomes a standard of care in the NICU, many key healthcare professionals will remain unaware of the necessity and the process of adverse drug reaction reporting, creating further challenges to the identification of the barriers to quality and possible strategies for improvement of the currently available approach. To achieve safe use of therapeutic products in the NICU, the establishment of a sturdy pharmacovigilance system that overcomes the recognized barriers to quality is the key approach. The initial step should target successful promotion of the awareness, knowledge, attitude and practice of drug-related AE reporting among all stakeholders. Subsequently, barriers to spontaneous reporting and a passive approach, which is the basis of postmarketing drug safety surveillance, should be identified. Under-reporting of drug-related AEs, specifically in the NICU environment where busy healthcare professionals are the primary, if not sole sources for reporting suspected drug-related AEs, is a well-recognized major challenge. Beside enhancement of knowledge, it is critical to facilitate the practice of AE reporting through the establishment of user-friendly platforms, hence encouraging the practice in the midst of busy working hours. Use of a standards-based AE reporting system, integrated with the Epic electronic health record (EHR) is already reported in cancer centers of excellence, with improvement in the consistency and the efficiency of reporting.12 The development of neonatal standardised AE terminology and the specific standard severity scaling will make it feasible to build neonatal drug AE assessment as part of the EHR system (Figure 1). Postmarketing surveillance of medicinal products frequently used in the NICU is a critical but rather neglected need of the fragile neonatal patient. Vigorous identification and quantification of adverse drug reactions in the sick neonate can play a considerable role in the improvement of short- and long-term outcomes as well as health-related individual, family and societal burdens. Tremendous effort has already put into building an infrastructure for the provision of a robust and age-appropriate postmarketing surveillance system for neonatal patients. Moreover, a multistakeholder collaborative effort is required to increase awareness among parents, healthcare professionals, regulatory agencies and pharmaceutical companies so that timely support for the optimal development of a robust surveillance system can be provided. The authors report no conflicts of interest.