Off‐label drug use in neonates and infants in Spain: A five‐year observational study

To provide information about the off‐label rate of all drug prescriptions in neonates and infants up to 1 year in Spain. Also, to analyse the off‐label prescription of medicines under current practice in this age group according to different evidence sources.


| INTRODUCTION
There are several reasons why paediatric medicine development has been largely neglected for a long time (e.g., protection from clinical research for ethical reasons, economic considerations related to investigation, and complexity of age subgroups). Currently, there is broad consensus that childhood deserve access to medicines that have been specially researched and developed for them. [1][2][3] Despite this, the development and testing of paediatric medicines has been far from satisfactory and the need for clinical trials has frequently been underscored, 2,3 especially for neonates and infants, which are the most vulnerable age groups. 1 Consequently, paediatricians must rely on their own experience and extrapolation to prescribe rather than on the results of robust clinical research. 2 Moreover, medicines are often not available in a suitable pharmaceutical form and paediatricians must turn to medicines authorised for adults or children with different age ranges and adapt dosing and pharmaceutical form. 4 This off-label use of medicines comes with the risk of ineffectiveness and adverse drug reactions (ADRs) in children. 1,2,5 Despite the attempt of international harmonisation in paediatric drug development and authorization, each regulatory body in the world has its own complex pathway, which is also subject to local regulation and subsequent barriers. This fact has often been controversial in many countries like the United States (US) or the European Union (EU), due to the increasing need of moving forward faster in paediatric drug regulation from the perspective of clinical practice and real-world evidence. 6 Very few studies, such as Suarez-Castañ on et al., 7 have analysed the evidence of use for off-label medicines in primary health care setting. According to studies conducted in the EU member states, there is a large variation in off-label prevalence and no clear patterns were identified. 1 A recent systematic literature review addressed unlicensed and off-label prescription of drugs to children in primary health care, 8 which included six studies conducted between 2000 and 2016, indicated that the incidence of off-label use varied from 29.5% to 51.7% of total drugs prescribed (prospective analysis). The prevalence of off-label drugs ranged from 31.7% to 93.5% in relation to the total number of drugs prescribed (retrospective analysis). The studies included patients from Brazil, France, Malta, Scotland and The Netherlands.
A novel joint policy statement, recently published in Europe, aimed to fill the gap between medicines licensed for paediatric use (regulatory authorisation) and real-world evidence in neonates, infants, children, and adolescents. Its purpose is to offer guidance for health care professionals on when and how to prescribe off-label medicines to children and to provide recommendations for future European policy. 9 Thus, the main objectives of this study were first to provide information about the off-label rate of all drug prescriptions in neonates and infants up to 1 year in Spain, and secondly to analyse the time trends in off-label medicines prescription under current practice in this age group according to different clinical evidence sources (national, UK, non-EU). The average births per year in Spain during the study period was close to 400 000. Neonates and infants up to 1 year accounted for 2% of the total population in the country. 10 The primary health care workforce in Spain is organised around a multidisciplinary team, with a gatekeeping function which provides free access to prevention and promotion of health, acute and chronic care, home care and community care activities. The core of the team is made up of general practitioners (GPs), who are family and community medicine specialists, paediatricians, nurses, auxiliary nurses, social workers, dentists and administrative staff.

| Inclusion criteria
All prescriptions issued by paediatricians in neonates and infants (0-1 year) in primary health care (outpatients). Due to the high number of prescriptions (per year >5 000 000) expected for this age group according to preliminary results, it was decided not to broaden the paediatric population under study, 13 such as infants and young children (1-4 years), children (5-11 years) and adolescents (12-15 years). 14

| Exclusion criteria
All prescriptions issued for neonates and infants up to 1 year at hospitals (inpatients). Prescriptions from GPs, nurses and other professionals different from paediatricians in primary health care. Repeated prescriptions from specialists were excluded too.

| Data sources
Study data collection included the prescription records from a private database (complete national scope) sorted by months of age and drug (branded name, active substance, strength, package and dosage form). The database did not provide information about the type of prescription (paper format or electronic) and the number of patients (neither aggregated nor by age group). Since it was not linked to medical records, both first and repeated prescriptions were considered but undifferentiated. 15 The Summary of Product Characteristics (SmPC), the legal document approved as part of the marketing authorisation of each medicine, was the primary source to determine off-label prescriptions. 16 Additional data sources were used for further analysis: Pediamécum (a Spanish documentary database of active substances used in Paediatrics), 17 British National Formulary (BNF), 18 DailyMed 19 and the Food and Drug Administration (FDA). 20 These sources were visited at different points in time, accordingly to each year of study.

| Variables
Variables collected were quantitative: demographic (months of age; gender could not be derived from the dataset) and resource use (number of total prescriptions; number and percentage of off-label prescriptions and evidence-based; number and percentage of off-label over-the-counter A prescription is the form used to prescribe a medicine. According to the Spanish regulation, one single prescription could be used to prescribe several packages of the same medicine in specific situations (i.e., insulins and some systemic antibiotics).
The age grouping of study was neonates and infants up to 1 year. 14 In addition, four subgroups of age were established (0-2 months, 3-5 months, 6-8 months and 9-11 months) to carry out a sub-analysis about general results on prescription rate according to months of age.
In the present study, off-label drugs were defined as those that were not authorised for this age group according to the information on the SmPC. 16 Otherwise, they were on-label.
Off-label prescriptions were subsequently classified as evidencebased and clinically unjustified (not evidence-based), according to the additional data sources listed for data analysis. Sources like Pediamécum, 17 BNF 18 and DailyMed 19 include updated evidence of real clinical practice and the FDA labelling 20 may provide different information than the European or Spanish SmPC. 16 In summary, these additional sources may include variations from the SmPC information depending on age. In case any of them pointed out information for use in less than oneyear old children, the drug was therefore considered off-label on a regulatory basis but evidence-based according to current clinical practice. OTC drugs are usually aimed for mild symptoms and self-care.
Although they are medicines dispensed directly to a consumer without the need of a prescription from a GP in Spain, they may appear in prescriptions as a reminder to the patient.

| Data analysis
Total number of prescriptions were studied according to labelling, evidence of use in real clinical practice and drug characterisation. A T A B L E 1 Number of prescriptions issued, percentage of total by age group, average percentage per age year in each age group and breakdown by year F I G U R E 2 Off-label and evidence-based off-label rate per year. Legend: The figure represents total off-label rate by year, which were further classified depending on whether off-label use was evidence-based or could be clinically justified any way. In this regard, evidence-based off-label prescriptions were divided depending on the source of evidence: Pediamécum and/or British National Formulary (European) and DailyMed According to the preliminary analysis conducted by months of age during the first 2 years of the study (2015 and 2016), a single drug could be on-label for a specific month of age, but off-label for another; and in the same way, it could be an off-label prescription justified by clinical practice for a specific month of age, but not evidence-based for another. 13 3 | RESULTS

| Prescription rate according to months of age
The number of prescriptions issued in neonates and infants (0-1 year) split by months of age is shown in Figure 1. Note that neonates and infants up to 2 months accounted for an average of 34.15% of total prescriptions in the age group up to 1 year. The highest average prescription rate was in the first months of age and progressively dropped as age increased, as shown in Figure 1.
The statistical analysis showed that subgroups followed normal distribution for the number of yearly prescriptions and that subsequent decreases in this number along the age subgroups were all statistically significant (p < 0.05) in the 5-year study period.  Year 2016 Year 2017 Year 2018 Year 2019

| Off-label prescriptions characterisation
During the 5 years of study, a total of 9 271 296 prescriptions were

| Off-label prescriptions and justification by clinical practice
The average off-label prescription rate over the 5-year period was  scriptions, respectively. Note that total sum may exceed 100% since off-label use could be defined in more than one SmPC section.

| Drugs and therapeutic groups most prescribed off-label
The top 10 drugs and therapeutic groups that were most prescribed off-label between 2015 and 2019 are included in Table 2 and Supplementary Table SS2. The leading group of medicines prescribed off-label was selective beta-2-adrenoreceptor agonists, being salbutamol the most prescribed active substance (>6% of annual off-label prescrip-   The therapeutic groups that were mostly prescribed off-label accounted for more than 23% of total off-label prescriptions per study year (Supplementary Table SS2  In this regard, sub-analysis by age groups would be also a must.

| Off-label prescriptions characterisation
Different studies conducted have shown that neonates and infants receive the greatest proportion of off-label drugs, meaning that they have the highest exposure to drugs that are insufficiently documented with regards to efficacy, safety and dosage, even at hospital setting. 25 Therefore, this group of age seems to be particularly vulnerable.
Many off-label drug studies concluded their results by determining the prevalence of off-label use and drug characterisation in childhood. 26,27 However, studies in neonates and infants were mostly developed at hospital setting (e.g., intensive care units), so comparisons with our novel study become complex since the most relevant medicines described differ from those mostly used in community practice. 25 The recent systematic literature review from Almeida Andrade et al. 8 pointed out that the highest proportion of unlicensed prescription was related to children under 2 years of age. Another systematic literature review (2018) on off-label medication use in children 28 identified 31 studies, with off-label prescription rates from 3.2% to 95%.
Only 7 studies were specifically in the neonate population, with off-label drug use ranging from 26% to 95%. 1 In the US, Yackey et al. 27  off-label use, as it was rather the combination of different items (e.g., therapeutic indication, strength, route of administration and contraindications). In contrast, in the current study the off-label use concerns were most often related to the posology section.
As far as the authors know, the current study is the only one that has evaluated use of OTC drugs in very young children in primary care.
The authors found that on average, almost 13% of total off-label prescriptions were issued for OTC drugs. However, the % of off-label OTC use could be even higher, as these drugs do not require prescription to be dispensed and the risk of self-medication should not be neglected.
Few studies reported that children in general receive substantial amounts of OTC drugs including natural remedies, which are often regarded as harmless agents. A Swedish review 32 estimated that children were likely to receive at least one OTC drug during any given 3-month period and that the evidence related to paediatric use was non-existent. However, those results are only indicative and could not be compared to the present study, as it did not include population under 11 years of age. According

| Off-label prescriptions and justification by clinical practice
In this study, the prescriptions that were off-label in the SmPC were analysed twice in order to assess the evidence of use in daily practice according to several international databases. and endocrinology/metabolic diseases (9%) at the forefront. 39 However, the diseases with highest number of completed paediatric investigation plans could not be correlated to the real burden of disease in this population group and on the contrary be subrogated to the developing clinical trials in adults. 39 In case of infants, evidence in primary health care (outpatients) was wider, although disaggregated results by year of age were scarce. 1 For instance, the review carried out by Kimland et al. 32 pointed out that the highest proportion of off-label drug use in children was among topically administered drugs and eye drops, although no further details about younger age groups were disclosed.
A recent publication from our investigation group 40 did not find a decrease of the annual off-label prescription rate in the primary health care setting of the Spanish paediatric population, even though the latest Paediatric Regulation (1901/2006/EC) 36 publication, and more than 10 years after its adoption. In general, prescription of innovations for outpatients in primary health care setting in Spain is scarce, which is usually due to highest prices than the standard of care and uncertainty about effectiveness in the short term. This scenario leads to using medicines with broad experience in the market despite not being authorised for paediatric population in the SmPC, which may pose a safety risk for the most vulnerable paediatric population groups, like neonates and infants (<1 year old).
In addition to the specific regulation that encourages paediatric drug development, big data (i.e., patient registries, real-world data and digital health) will bring an expanded knowledge about paediatric medication use, which will definitely help improving the safety of this population group. 41

| LIMITATIONS OF THE STUDY
This study has some limitations that could lead to underestimation of off-label use. In addition, there is an imbalance in the evidence available of off-label use between hospital and primary health care setting in Spain. Thus, the results of most studies found in the literature could not be compared to those presented in this study.
First, this study was focused on the off-label use of authorised medicines and did not include medical devices, compounded formulas, cosmetics, food supplements and medicinal plants. Second, only prescriptions issued by paediatricians at primary health care were considered. This situation could be underestimating the potential total number of off-label prescriptions in the study population, as repeated prescriptions from other medical specialties (i.e., hospital) were not accounted. Third, non-prescription medicines were not accounted (i.e., self-medication and many OTC drugs), although 13% of drugs prescribed off-label were identified as OTC. Fourth, as the data analysed was not linked to medical records, it was not possible to assess if the off-label use was justified by evidence-based information provided by the GPs. Finally, results show potential off-label use from prescriptions, as dispensation data was not provided by the database used.

| CONCLUSION
The off-label use of drugs remains an important public health concern for the paediatric population, especially for neonates and infants up to 1 year, who receive the greatest proportion of off-label drugs as demonstrated in this study. Therefore, to achieve the ideal therapeutic practice, more new labelling information must be added, as practitioners still rely on the evidence-in-use when off-label drugs are required for their patients. Nonetheless, substantial part of all offlabel prescriptions is clinically justified by clinical evidence sources beyond the SmPC.
To sum up, even though expectations for the future are focused in adding more and more clinical trials for new medicines on the market, there is also a need to add more research and more safety information on some older drugs, which are commonly used in paediatric patients. As disclosed in this study, these activities would be needed primarily for dermatological corticosteroids and antifungals, ophthalmological anti-infectives and dosage forms for respiratory diseases, which show a high rate of off-label prescriptions.