Off-Label and Unlicenced Medicine Use among Hospitalised Children in South Africa: Practice and Policy Implications

Background: Information regarding off-label and unlicensed medicine use among South African children is limited. This is a concern as the prescribing of off-label and unlicensed medicines can lead to issues of effectiveness and safety as well as raise liability issues in the event of adverse events. This potentially exposes physicians to legal penalties. Consequently, we sought to determine the prevalence of off-label and unlicensed medicine use among paediatric patients in South Africa to provide future direction. Methods: This study retrospectively examined the use of medicine in a point-prevalence survey study (PPS) involving paediatric patients aged (0–2 years) admitted to selected public hospitals in Gauteng Province, South Africa. Data were collected per hospital over two days between February 2022 and July 2022. Demographics, duration of treatment, diagnosis, and medicines prescribed were collected from patient medical records using a mobile application. Prescribed medicines were reviewed against the medicine formularies and other databases to assess their appropriateness. Results: From three academic hospitals, 184 patient records were reviewed. A total of 592 medicines were dispensed, of which 379 (64.0%) were licensed and 213 (36.0%) were used off-label/unlicensed for paediatric patients 0–2 years of age. The most prevalent off-label and unlicensed medicines were multivitamins (n = 32, 15.0%) and ampicillin injections (n = 15, 7.0%). Conclusion: The frequency of unlicensed and off-label medicine prescribing shown in this study is consistent with the literature and can be considered high. This practice can pose a risk because it adversely affects patients if not properly regulated. Attention is needed to ensure future high-quality, safe, and effective use of medicines.


Introduction
Medicine prescribing and the use of medicines in paediatric patient care have been a global issue for a number of years, with high rates of off-label prescribing seen in paediatric patients in a number of studies [1][2][3][4][5][6].In general, studies indicate that the global level of off-label or unlicensed use of medicines among hospitalised children ranges from 12 to 70% for prescriptions and can reach up to 100% in some studies [7,8].Recent reviews also suggest that the unlicensed use of medicines can account for up to 75% of medicine use among hospitalised children in some studies [8].Having said this, Oshikoya et al. (2017) reported off-label prescriptions of only 7.7% among children with chronic diseases attending specialty paediatric clinics in Nigeria [9].However, the potential for drug-drug interactions was higher among paediatric patients in an earlier study [10].In addition, there were considerable concerns with the off-label use of pentazocine among paediatric surgical patients in Nigeria, and most children experienced between two and seven adverse Pharmacy 2023, 11, 174 3 of 17 to determine the type and extent of medicines (off-label and unlicensed) prescribed to children (0-2 years) in academic hospitals in South Africa.Initially, the intention was to conduct this research in four academic hospitals in Gauteng Province, South Africa.However, whilst permission was obtained for all four hospitals, access was ultimately granted to only three.Gauteng Province was selected for this initial study because of the ease of access.Academic hospitals were selected because they provide specialised care, e.g., paediatric, neonatal intensive care, oncology, and paediatric surgery.The combined bed capacity of the four conveniently selected academic hospitals for the survey was 927.As per PPS study designs, the total number of beds was used as the population size determinant for this study.

Sample Size and Strategy
Files of paediatric patients aged 0-2 years hospitalised and available in the ward during the study period were included.The required sample size of 234 was calculated using a 95% confidence interval with a 50% proportion and a margin of error of 3%.A systematic sampling strategy was used whereby a patient in the ward was selected for every file of 0-2-year-olds until the sample size was achieved.

Data Collection
The data collected were all the medicines prescribed among paediatric patients aged 0-2 years who were admitted and available in the ward on the day of data collection from 08:00-17:00 h.Data were collected from patient medical records with the aid of a skilled data HM collector.Patient information was recorded on the PPS information sheet, accessible as a mobile application.The following information was collected from the patients' files: their age, weight, and gender; length of therapy; diagnosis; route of administration (oral, intravenous, inhalation, topical, and rectal); and prescribed medications.
The following definitions were used in the PPS forms [31]: • Off Label use: Defined as the administration of a drug/medicine in a manner that differs from that recommended in marketing authorisation with respect to age, dose, frequency of administration, route of administration, formulation, and/or indication.
Similarly, an approved medicinal product is a medicinal product prescribed and administered in accordance with its marketing authorisation.

•
Un-licensed use: Refers to the use of a medicinal product that has not been approved for marketing by the country's medicine regulatory authority (the South African Health Product Regulatory Authority (SAHPRA)) [32].
The categories of off-label use included age, weight, absence of paediatric information, lack of paediatric clinical data, contraindication, route of administration, and formulation/dosage form of administration as stated in the literature insert or official compendium.The categories of unlicensed use included medicines not approved by the national medicine regulator [32].
The dates were deidentified, and all patient identification details (names, identity numbers, and patient file numbers) were not recorded to completely delink the patient from the data.Data were collected from February 2022 to July 2022.This period was chosen as it includes the winter, a flu season for the young and old, with anecdotal evidence suggesting an increase in medicine use during this period.
In terms of developmental differences and medicine use, the children in the PPS forms were categorised into preterm new-born infants (born before 37 weeks of pregnancy), term new-born infants (0-28 days), infants (28 days up to 12 months), and paediatrics (1 year up to 2 years) [33].

Data Analysis
The collected data were extracted from a mobile application and imported into Microsoft Excel.An unbiased arbitrator checked and cleaned the collected data to ensure consistency.The information was then entered and analysed using (IBM) SPSS Version 28.0.Descriptive statistics were used to calculate the frequencies and percentages of patient demographics and for all variables relevant to the study objectives.Medicines were described at different levels of the World Health Organisation's Anatomical Therapeutic Chemical Code (ATC) classification.Subsequently, the frequency of medicines used off-label and unlicensed was calculated using the WHO's (ATC) classification [34].The conditions under which these medicines were prescribed were categorised as per the International Classification of Diseases 10th revision (ICD10) classification system [35].A medicine is characterised by its distinct active ingredient, often known by its International Non-proprietary Name (INN).Consequently, various formulations with the same active ingredient, such as paracetamol syrup and paracetamol drops, were recorded as identical medicines.Binary logistic regression was used to evaluate the association among demographic variables with off-label/unlicensed medicine use at a 95% CI and p-value ≤ 0.05 significance level.

Ethics Approval and Consent to Participate
Ethical clearance (SMUREC/P/128/2020) was obtained from the Sefako Makgatho Health Sciences Research Ethics Committee, University of the Witwatersrand (clearance certificate M210426), and the National Department of Health (GP_202011_0470).The study was conducted after obtaining official permission from the hospital administration.No consent to participate was required because this study was a retrospective data review of patient records with no direct contact with the patient, and the data were delinked from the patient's details.This is similar to other PPS studies in South Africa [36][37][38][39][40].

Results
After receiving ethical approval from all the potential academic hospitals, access to data was denied by the clinicians at one of the four academic hospitals as they feared that the study findings might be used as evidence during litigation despite the anonymity of the data.Overall, 184 (78.6%) of the envisaged 234 samples of paediatric patient files were subsequently accessed from three of the four academic hospitals.

Demographics
Among the 184 paediatric patient files that were reviewed, 592 medicines were prescribed, translating to an average of 3 medicines per patient file.A total of 79.3% of the files were for children between the ages of 0 and 1, as indicated in the study population (Table 1).The mean weight of the study population was 4.62 kg (SD ± 3.63 kg), with a little more than half (51.6%) of the patients being male and the majority, 96.7%, African.The prevalence of off-label or unlicensed medicine use in children aged 0-2 years included in this study was 36% (213/592).Overall, 177 medicines (29.9%, n = 592) were prescribed off-label, while 36 (6.1%, n = 592) were categorised as unlicensed.Off-label prescribing was most prevalent among neonates aged 0-28 days, with 84 cases (39.4% of the 213 cases).On the other hand, unlicensed medicine use was most prevalent among infants aged between 29 days and 1 year, with 17 cases (8.0% of the 213 cases), as illustrated in Figure 1.The prevalence of off-label or unlicensed medicine use in children aged 0-2 yea included in this study was 36% (213/592).Overall, 177 medicines (29.9%, n = 592) we prescribed off-label, while 36 (6.1%, n = 592) were categorised as unlicensed.Off-label p scribing was most prevalent among neonates aged 0-28 days, with 84 cases (39.4% of t 213 cases).On the other hand, unlicensed medicine use was most prevalent among infan aged between 29 days and 1 year, with 17 cases (8.0% of the 213 cases), as illustrated Figure 1.

The Top 10 Most Prescribed Medicines at ATC Level 5, Chemical Substance, or INN
Six of the top ten most prescribed medicines were off-label or unlicensed (Table 2)   Six of the top ten most prescribed medicines were off-label or unlicensed (Table 2).
It is worth noting that the use of off-label and unlicensed medicines was not mutually exclusive.In fact, every single paediatric patient (n = 184) included in this study received at least one off-label or unlicensed medicine during their hospital stay.The most frequently prescribed off-label medicine was intravenous caffeine (5.2%, n = 31 of the 592).On the other hand, the medicine that was predominantly used in the unlicensed category was multivitamin syrup/drops, accounting for 6.4% of the cases.In children aged 0-28 days, ampicillin injections were the most frequently prescribed off-label medicine, with a total of 15 cases, accounting for 7.0% of all medicines (n = 512).Caffeine citrate injections were the second most prevalent off-label medicine, with 11 cases (5.2%).Additionally, gentamicin injections were identified as a prevalent off-label medicine exclusively among the 0-28-day age group, with a total of nine cases (4.3%).
Injections were the most common (80%) dosage form of off-label medicines used in this age group.In contrast, oral dosage forms-tablets, capsules, and syrups-were most common (90%) in infants.Consequently, most medicines were not established for use in neonates or manipulated in infants.

Unlicensed Medicine Use by Age Group at the ATC Level 4 Chemical Subgroup
The most used unlicensed medicine across all ages was multivitamins (A11), and probiotics (A07) were the medicines mostly used unlicensed in infants aged from 29 days to two years.In children aged 0-28 days, multivitamins were mostly used unlicensed, with 12 cases (5.8% of 213).Again, in children aged 29 days to one year, were the most common unlicensed medicine with 11 cases (5.3% of 213), followed by probiotics in 2 cases (0.9% of 213).Lastly, in children aged 1 to 2 years, multivitamins were the most common unlicensed medicine with five cases (2.3% of 213), followed by probiotics with two cases (0.9% of 213), as indicated in Table 4.

Conditions Associated with Off-Label and Unlicensed Medicines by ICD-10 Codes
The data presented in Table 5 highlight the most common medical conditions, as indicated by their corresponding ICD-10 codes, in which medicines were used off-label or unlicensed.These were identified based on their frequency within the dataset and comprised 81.7% of diagnosed medical conditions in the sample with off-label or unlicensed medicine use.Bacterial infections (ICD-10 code A49.9) were the most prevalent, accounting for 26.8% of the cases.Within this category, the majority of cases were prescribed off-label or unlicensed medicines in all age groups, with the highest proportion in the 0-28-day age group (61.4%).

and Unlicensed Medicine Use by Therapeutic (ATC Level 2) and Age Groups
The most common (81.8%) off-label and unlicensed medicines are shown in Table 6.Anti-bacterial for systemic use (ATC code J01), used off-label, was the most prevalent medicine category, comprising 26.3% (56/213) of all off-label and unlicensed medicines (n = 213).The off-label use of this category was highest in the 0-28-day age group (62.5%).The next most frequent medicine group, accounting for 17.4% (37 cases) of all off-label/unlicensed medicines, were vitamins (ATC code A11).Vitamins were unlicensed and mostly prescribed to infants aged 29 days-1 year.
No statistically significant associations were found between patient demographics/ health-related variables and off-label/unlicensed use using binary logistic regression analysis.The Chi-Square Tests between the reason for off-label/unlicensed use and patient demographics (age categories and patient weight) were found to be statistically significant with a p-value of <0.001.

Discussion
To the best of our knowledge, this is the first study to report the prevalence of offlabel use and unlicensed medicine use among the paediatric population aged zero to two years in South Africa.The frequency of unlicensed and off-label medicine prescribing in our study is consistent with some of the published literature and can be considered high [21,32,33,[41][42][43][44][45][46][47][48][49][50][51].We have seen in published studies among LMICs that off-label use among hospitalised paediatric patients can account for up to 99.5% or more of prescriptions [4,[6][7][8]15,35,43,[50][51][52][53][54].However, whilst the percentage of off-label prescribing in our study was appreciably lower than the rates seen in a number of LMICs at 36% for off-label/unlicensed use, this does not negate potential concerns in a number of these very young infants.This practice can pose a risk because it can adversely affect young infants if, for instance, doses and their implications are not properly regulated in hospitals [4,6,7].Very young infants are more susceptible to side effects when prescribed off-label medicines due to differences in pharmacokinetics and pharmacodynamics compared to adults [43].This prescribing trend is particularly prevalent in infants aged 0-28 days, with 44% of infants in our study receiving off-label or unlicensed medicines.This rate is higher than in older infants and is consistent with global studies [54][55][56].The heightened risk stems from the fact that neonates' renal and hepatic functions are not fully developed [13,21,50,53,[56][57][58]. Furthermore, there is a limited evidence base for the safety and efficacy of these medicines in very young infants [35,[59][60][61][62].
We saw an average of three medicines prescribed per infant in our study, similar to other studies [31,53,63].However, this was much higher than in Norway (0.8) and Spain (1.5) [64], and slightly lower than in Italy (3.7) [65], Malaysia (where the median number was 4 [63]), and Indonesia (where the median number of medicines prescribed was 9 [54]).Overall, paediatric patients are exposed to a high number of off-label and unlicensed medicines, which could lead to suboptimal clinical efficacy and unanticipated side effects.This needs to be addressed going forward, enhanced by regularly reviewing the evidence base for their use [66].This is because for many medicines typically prescribed in neonatal and young infant ICUs, safety and efficacy data for neonatal pharmacotherapy are lacking, with an appreciable number of neonates in ICUs being prescribed medicines that are not approved or are used off-label [4,7,43,67].Neonatal pharmacotherapy and prescribing practices require special attention, primarily because, as mentioned, neonates have unique pharmacokinetic and pharmacodynamic profiles compared to older children and adults.These differences can influence drug absorption, distribution, metabolism, and elimination [68,69].Consequently, the efficacy and safety of medications in neonates can vary significantly from other populations [29,70].Accurate dosing, vigilant monitoring, and a comprehensive understanding of the drug's effects are essential to prevent potential adverse reactions [71] and to ensure therapeutic efficacy in this very young population.Consequently, healthcare professionals must remain updated on the latest research, guidelines, and recommendations related to neonatal pharmacotherapy [17,69,72].This is very important for tertiary hospitals in the public healthcare system in South Africa and beyond going forward.
In our study, similar to others [13,35,50,[73][74][75][76], systemic antibiotics were the most frequently prescribed medicines, especially among neonates.This high use of antibiotics reflects the fact that the top indication for off-label or unlicensed use in medicines in our study was for bacterial infections at 26.8%, higher than studies in Spain at 12.0% [42], France at 22% [45], and Uganda at 18.9% [77].However, this study found lower use compared to studies conducted in the Western Cape, South Africa, which had a rate of 39% [78], and Jordan, where the rate of bacterial infections was 54.1% [79].This high rate of antibiotic prescribing is perhaps not surprising, as academic hospitals typically treat more premature children with low birth weight and sepsis than secondary or community hospitals [80].However, it is important to fully monitor the prescribing of antibiotics in this population because sepsis is the leading cause of neonatal death globally, killing more than 1 million neonates worldwide each year, with appreciably higher mortality rates in LMICs [81][82][83].This results in antibiotics being among the most commonly prescribed drugs in neonatal intensive care units [84][85][86].Proper dosing of antibiotics is critical as under-or over-dosing can increase antimicrobial resistance (AMR) [56,87], which is a concern with mortality from AMR growing globally, with the highest mortality rates from AMR currently seen in sub-Saharan Africa [88].In addition, an estimated 31.0% of neonatal sepsis deaths are currently due to AMR and are rising [89].In addition, under-or over-dosing medicines (including antibiotics), which may result from their off-label or unlicensed use, is a concern, as this poses a risk of reduced effectiveness, increased adverse reactions, or both, along with potentially increasing AMR [75,87].We have seen antimicrobial stewardship (ASP) programmes effectively introduced in hospitals across Africa in recent years to improve antimicrobial prescribing, with hospital pharmacists playing a key role [90][91][92][93].These ASP exemplars should provide guidance to key stakeholder groups in South Africa and beyond to address the inappropriate use of antibiotics alongside rising AMR on the continent.
There was also appreciable prescribing of caffeine in our study.Thomas (2014) classified caffeine as unlicensed in all dosage formulations [49].However, at the time of our study, caffeine was registered with the regulators in South Africa as an injectable but was prescribed and administered orally.It is worth noting that there is currently no commercially available oral solution specifically formulated for caffeine in South Africa.Nevertheless, the Department of Health strongly recommends the oral route of administration for caffeine as per EML guidance [94][95][96][97][98][99].Vitamins were also the most common medicine used unlicensed or unapproved in our study.This was due to a lack of marketing authorisation for vitamins from the regulatory body in South Africa.Currently, the South African Health Products Regulatory Authority classifies vitamins as dietary supplements.Previously, they were considered food/dietary supplements and did not undergo the same registration process as conventional medicines.Consequently, this might be the justification for unlicensed medicine status.This again needs to be looked at in light of their considerable use among this patient population in South Africa.
Age, route of administration, and dosage were the most common reasons for off-label prescribing in our study, similar to previous studies [7,32,49,100].Due to the lack of suitable oral drug forms for neonates and young children, tablet splitting and dissolution in sterile liquids before administration are common, as seen in other studies [101].However, both within and outside hospitals, the practice of tablet splitting or dissolution can place nurses and caregivers in the difficult situation of having to prepare and administer the medications according to current recommendations [102].Furthermore, child acceptance of these manipulated medicines could be compromised [103].It is also difficult for physicians to adjust dosages over time to ensure the adequate safety and efficacy of the prescribed medicines because the predominant method of manipulation is mixing with liquid and food.Alongside this, for certain medicines, food-drug interactions can appreciably affect their bioavailability and therapeutic efficacy, which needs to be considered when administering them [101,102].Drug delivery and uptake can also be influenced by the medicine's swallowability, taste, smell, texture, and appearance, which can be altered when adjusting or diluting doses [103,104], potentially leading to worse outcomes as a result.Consequently, this again needs careful monitoring.
The role of Drug and Therapeutic Committees (DTCs) in promoting rational and evidence-based prescribing practices is also pivotal, especially in addressing the current lack of rigorous regulatory evaluation of paediatric formulations [105].We have seen the role of DTCs grow in South Africa compared with other African countries; however, more needs to be accomplished going forward [106][107][108][109][110]. Standard Operating Procedures (SOPs) can aid DTCs in streamlining and regulating inappropriate medication use, particularly concerning dosing for young infants, and preventing adverse drug reactions [110].Monitoring adherence to these SOPs elevates the quality of care and strengthens the feedback loop, which is essential for continuous improvement.
Collaboration between DTCs and regulatory authorities, including SAHPRA, is paramount to bolstering the impact of these endeavours.Such a partnership would ensure up-to-date information on drug safety, efficacy, and quality for paediatric populations.By prioritising clinical trials tailored to the paediatric population and addressing issues related to off-label and unlicensed medication use, a comprehensive understanding of drug effects and risks can be established for this vulnerable group.Ultimately, by harnessing the synergy of DTCs and SAHPRA, South Africa has the potential to be a beacon for evidence-based paediatric medicine administration, extending best practices across the African continent.This is similar to the situation seen with the implementation of national action plans to reduce AMR across Africa, with ongoing activities in South Africa being more advanced than those seen in a number of other African countries and providing direction [90,105,[110][111][112][113][114][115][116][117].
We are aware of a number of limitations with this study.Firstly, as this was a PPS study design, we only recorded medicines prescribed that day.Consequently, we were unable to link any off-label or unlicensed medicine with any contribution to any adverse drug reaction.Furthermore, this pilot study was conducted in only one province.However, despite these limitations, we believe the findings are robust.

Conclusions
In this study, off-label and unlicensed medicine use appear very common among paediatric patients aged 0 to 2 years of age admitted to public, academic hospitals in South Africa, similar to other studies.Addressing the issue of off-label and unlicensed drug use in paediatric patients is vital for promoting patient safety and improving healthcare outcomes going forward.In the first instance, this includes a greater role for DTCs in hospitals to develop appropriate standards and monitor their implementation, as well as encouraging greater understanding of neonatal pharmacology.In addition, seeking to instigate clinical trials, where possible, tailored to the paediatric population should be initiated.We will be following up on these suggestions in the future.
Author Contributions: H.M., with the assistance of M.M., developed the study concept.H.M. was responsible for conducting the data collection.Data analysis and manuscript preparation were carried out collaboratively by H.M., M.M., N.S., S.O. and B.G.The initial draft of the manuscript was authored by H.M., and subsequent versions underwent review and comments from all authors.The final manuscript received approval from all authors after careful reading and evaluation.All authors have read and agreed to the published version of the manuscript.
Funding: There was no funding provided for this study.
Institutional Review Board Statement: Ethical clearance (SMUREC/P/128/2020) was obtained from the Sefako Makgatho Health Sciences Research Ethics Committee, University of the Witwatersrand (clearance certificate M210426), and the National Department of Health (GP_202011_0470).This study was conducted after obtaining official permission from the hospital administration.No consent to participate was required because this study was a retrospective data review of patient records with no direct contact with the patient, and the data were delinked from the patient's details.All procedures were conducted in compliance with the appropriate guidelines and regulations.
Informed Consent Statement: Not applicable.

Figure 1 .
Figure 1.Prevalence of off-label and unlicensed medicine use in different age categories withi sample of 184 children aged 0-2 years old across three academic hospitals located in Gauteng Pr ince, South Africa.

Figure 1 .
Figure 1.Prevalence of off-label and unlicensed medicine use in different age categories within a sample of 184 children aged 0-2 years old across three academic hospitals located in Gauteng Province, South Africa.

3. 3 .
The Top 10 Most Prescribed Medicines at ATC Level 5, Chemical Substance, or INN
3.4.Off-Label Use by Age GroupOff-label medicine use varied by age group.Table3provides an overview of the top 10 off-label medicines stratified by age group.

Table 3 .
Categories of the top 10 off-label medicines (ATC Level 5) in children under 2 years of age.

Table 4 .
Most prevalent unlicensed medicines used.

Table 5 .
Most common medical diagnoses or conditions per ICD-10 codes for which medicines were prescribed off-label or unlicensed in three different age groups.

Table 6 .
Categories of off-label or unlicensed medicines used by age group (ATC Level 2).