Off‐Label and Unlicensed Prescribing of Medicines in Paediatric Populations: Occurrence and Safety Aspects

Off-label (OL) and unlicensed (UL) use of medicines in children and adolescents is common as only few medications are licensed for use in the paediatric population. Hence, lack of knowledge about adverse drug reactions (ADRs) from paediatric OL/UL prescribing makes medicine use in children of public interest and concern 1, 2. This topic was also the starting point for the study by Hart et al.: ‘Paediatric Off-Label Use of Melatonin - A Register Linkage Study between the Norwegian Prescription Database and Patient Register’ which was published in this journal in April 2015 3. Hartz et al. studied the prevalence of melatonin users in 4- to 17-year-olds for treatment of sleeping problems in the Norwegian Prescription Database 3. The analyses showed that the prevalence of melatonin use increased in both boys and girls from 2004 to 2012, and the prevalence was highest in the 4 to 8-year-olds 3. Among recurrent users, approximately 90% of users had a diagnosis of either psychiatric or neurological disorder, and therefore, the study confirmed the hypothesis that melatonin, licensed for treatment of sleeping problems in adults, is being prescribed off-label for the treatment of sleep problems in children and adolescents with psychiatric and neurological disorders 3. OL prescription includes prescribing outside the recommended dosage, indication and route of administration as well as age of the patient, and UL prescriptions include drugs not licensed for paediatric use and non-licensed drugs that are being prescribed on special permissions 1-4. Lack of knowledge about possible long-term safety aspects from OL/UL prescribing in children is of concern to both prescribers and parents, particularly with respect to harmful events 5, 6. Worldwide, several empirical studies have examined the extent of OL/UL prescribing in the paediatric population in both in- and outpatients, as well as in reports submitted to national and international ADR and claims databases 7-26. Across studies, the overall OL/UL prescription rates ranged between 5% and 70% of all prescriptions, and the rates were highest in children up to 2 years of age and in adolescents (11 to 17-year-olds) 7-28. In all studies, OL/UL prescribing was predominantly due to unapproved indication, dosage and dosing schedule 7-26. Hence, the OL/UL prescription varied across studies and settings with respect to dosage, indication, route of administration and therapeutic groups 7-26. Anti-infectious agents (amoxicillin/clavulanic acid), analgesics (paracetamol, propofol, morphine, fentanyl, ibuprofen), bronchodilators (salbutamol), topical drugs, sex hormones, cardiovascular drugs and psychotropic drugs were among those medicines most frequently prescribed OL/UL 7-26. Occcurence of ADRs in children are common, and although some are serious, empirical studies on this topic are scarce. Information about ADR occurrence in children from prescribing of medicines is limited 27. A qualitative review of empirical studies reporting paediatric ADRs occurring in hospital settings, by general practitioners or as ADR reports submitted to national ADR databases showed large variations in incidence and prevalence across studies and populations 28. The average ADR incidence for inpatients was 42% (range 1–179%), for inpatients hospitalized due to an ADR 9% (range 1–28%), for outpatients 14% (range 2–68%) and below 1% in national ADR databases. The average prevalence rate for inpatients was 24% (range 1–72%), for patients hospitalized due to ADRs 4% (range 1–9%), for outpatients 4% (range 1–17%) and below 1% in national ADR databases 28. The ADRs were mainly reported for the therapeutic groups: vaccines, antibiotics and psychotropic medicine. Due to lack of information about paediatric licensing status for the reported medicines across countries, it was not possible to distinguish between ADRs reported for OL/UL medicine use and ADRs reported for medicines with paediatric indications 28. As clinical trials in the paediatric population are limited, clinicians and health authorities have to rely on spontaneous ADR reports as the major source of information about previously unknown and rarely occurring ADRs, although the spontaneous reporting systems suffer from different barriers such as incomplete recognition of suspected ADRs and administrative barriers to reporting, which may result in under-reporting of clinically important ADRs 29. A study, which retrospectively analysed 4388 ADRs reported in 0- to 17-year-olds to the national Danish ADR database, showed that 17% of the reported ADRs were associated with OL/UL prescribing 30. More than one half of off-label ADRs were reported in adolescents, and serious ADRs were more likely to be reported for hormonal contraceptives, anti-acne preparations and allergens 31. In the study, the definition of OL prescribing was to the children's age, as we did not have any data on indication and dosage schemes 31. Despite the relatively large number of studies monitoring paediatric OL/UL prescribing with respect to occurrence, only few have systematically reported ADRs associated with OL/UL prescribing in the included cohorts. In a prospective intensive surveillance study conducted in a paediatric ward, OL/UL prescriptions were responsible for 67% of all reported ADRs 32. Posthumus et al. (2012) demonstrated that for 683 patients aged 0-18 years, 7% of their hospital admissions were due to ADRs. Of the children admitted to hospital due to ADRs from OL/UL prescribing, 33% of these ADRs were reported for chemotherapy, particularly methotrexate, doxorubicin and vincristine 30. Hence, other studies have monitored ADR occurrence from OL/UL prescribing in selected patient cohorts and for specific therapeutic groups, for example the use of antiretroviral drugs in HIV-infected children 33, prescribing of topical drugs in paediatric outpatients 34 and sildenafil use in paediatric pulmonary patients 35. There are huge gaps in the evidence on the safety of medicines in children, as only few medicines prescribed for children have been systematically tested in the clinical development trials for this patient group. However, if a child's health condition requires treatment, and no other alternative treatments exist, OL/UL prescribing must be appropriate 1-5. However, there are OL/UL prescriptions that need to be restricted or even penalized in the case of compassionate prescriptions or unjustified prescriptions or prescriptions not based on any scientific grounds 1-5. OL/UL prescriptions are not easy to track down because if the prescriber is to add ‘off-label’ on the prescription, then clearly, in practice, the prescriber will only do so in exceptional cases because of fear of reprisals 36. Thus, to identify the OL/UL prescriptions leading to long safety problems, healthcare professionals must be able to cross-reference the national drug utilization systems as well as ADR databases in order to map the patients’ actual medicine use. The shared computerized electronic patient record may resolve this problem, but only few countries have implemented such national IT systems, for example Denmark 37. However, the Danish electronic patient record is only able to provide information about OL/UL prescribing for a period of 2 years back in time, and therefore, additional information sources must be included, for example patient interviews and/or medication lists issued by general practitioners 37. In the future, there is a need for studies exploring specific cases of OL/UL prescribing through questionnaire-based surveys, additional studies of spontaneous reports submitted to national and international ADR databases and ADR monitoring studies in hospitalized children where health records are being scrutinized 31. To detect OL/UL prescriptions in children during hospital admission, a web-based electronic messaging interface linked to drug formularies with information about licensing status that can be implemented in existing knowledge and decision support systems must be conducted, preferably combined with a tool for detecting of OL/UL prescriptions leading to ADRs and drug–drug interactions in children and adolescents 31. In conclusion, several empirical studies have monitored the occurrence of OL/UL prescribing in children and adolescents; however, only few studies have investigated the safety implications of OL/UL prescribing, particularly with respect to serious and potentially life-threatening ADRs. Until more clinical studies have been conducted in the paediatric population, prescribers and caretakers must be aware of the risk of potential serious ADRs from OL/UL prescribing, and the safety of OL/UL prescribing in this patient group must be closely monitored 38, 39. Therefore, clinical pharmacists must be present in hospital wards with the purpose to monitor for OL/UL prescribing in order to reduce medication risk and thus contribute to patient safety and reduction of costs related to prolonged hospitalization due to ADRs. Lise Aagaard has received travel funds from Pfizer and Swedish Orphan BioVitrum. No sources of funding were used to assist in the preparation of this study.