The study aims to investigate the prevalence of off-label prescribing in the general paediatric ward at a major teaching hospital in Tasmania, Australia.
The study aims to investigate the prevalence of off-label prescribing in the general paediatric ward at a major teaching hospital in Tasmania, Australia.
The drug charts and medical records from two groups of 150 consecutive paediatric patients, admitted 6 months apart in July 2009 and January 2010, were studied retrospectively. Patients were required to spend at least one night in hospital and be aged less than 12 years. Each prescribed drug was compared with the approved product information to determine if the usage was off-label. Data concerning documented informed consent and adverse drug reactions were also recorded.
Three hundred patients were prescribed a total of 887 medicines. Of these, 31.8% were off-label and 57.3% of children received an off-label medication. There was no significant seasonal variation in patient characteristics or prescriptions. Drugs were most commonly off-label due to their dosage or frequency of administration. Of the 106 different drugs used, the use of 51 was off-label on at least one occasion, and for 30 drugs their use was off-label on more than 75% of occasions. The drugs most commonly used off-label were oxycodone, salbutamol and paracetamol. No informed consent documentation was identified, and two of five recorded adverse drug reactions were associated with off-label drug use.
Off-label use of medicines occurred frequently in paediatric inpatients. The available evidence often supported off-label medication use. An improved system for the revision of approved drug information and an Australian guideline for paediatric prescribing are needed.
A key objective of the Australian National Medicines Policy is that medicines meet appropriate standards of quality, safety and efficacy. In order to achieve this objective, the Australian Register of Therapeutic Goods exists, a register of therapeutic goods that have been evaluated and approved by the Therapeutic Goods Administration (TGA). The terms and indications for a ‘labeled’ drug are listed in its approved product information (PI). In Australia, this is also found in MIMS, the Monthly Index of Medical Specialties.
‘Off-label’ prescriptions involve an unregistered use for a registered drug. This is not to say that the use is incorrect, as the use of off-label medications may represent the most logical, evidence-based therapy. Medication approval and registration is not often sought for paediatric patients and, as a consequence, they may receive drugs that have only been approved for adults. A paediatric patient's age or weight may therefore result in the use of the drug being off-label. In other cases, the drug itself may be approved for administration to children, but one or more aspects of the way it is used may differ from those that have been approved, in terms of dosage, administration frequency or route of administration. While drugs may be initially registered for a specific indication, there are often circumstances where it is found to be effective in other conditions not originally listed in the approved PI. The PI may not be updated to reflect this, resulting in off-label indications.
Other Australian studies have examined the extent of off-label prescribing to outpatients in general, inpatients at a dedicated paediatric hospital and within a neonatal ICU, with rates of off-label prescribing of 26%, 16% (including unregistered items) and 47%, respectively. To date, however, no study has examined the extent of off-label prescribing on a general paediatric ward of a General Australian hospital, a setting where significant numbers of children are treated. Also, there have been no studies looking at the general trends in off-label use in Australian children for almost 10 years.
A retrospective review was performed of medical records and drug charts for patients admitted to the general paediatric ward of the Royal Hobart Hospital (a 500-bed acute care teaching hospital). To reduce potential bias caused by seasonal variations, data were collected for two groups of 150 patients, admitted 6 months apart. The first group comprised 150 consecutive patients admitted from 1 July 2009 until 5 August 2009 inclusive. The second group comprised 150 consecutive patients admitted from 1 January 2010 until 22 February 2010 inclusive. Only patients who spent at least one night on the paediatric ward were included in the study. Patients who were aged 12 years or older on admission were excluded, as this is the age cut-off often used by manufacturers in their PI. Patients admitted for social (non-medical) reasons were also excluded. A further four patients within the relevant periods were excluded as large portions of their medical records were unavailable.
Data recorded included gender, age, weight on admission, reason for admission, diagnosis, length of time in hospital and outcome. Details for each medication that was administered to the patient were recorded: name of drug, maximum single dose, maximum daily dosage, type of formulation, route of administration, indication and dose calculation (where recorded). The indication of each drug was ascertained from either the medication chart or elsewhere in the patient's medical record. Medical records were also examined for the presence of written informed consent that may have related to off-label uses of medications, as well as documentation of suspected or confirmed adverse drug reactions (ADRs). The following items were excluded from analysis: normal saline solutions and heparin used to maintain the patency of intravenous lines, blood products, total parenteral nutrition, oxygen therapy, vaccinations, medications given in the Department of Emergency Medicine and unscheduled over-the-counter medications. The later were excluded, as due to the lack of a PI, their appropriateness in children could not be assessed.
Following data collection, analysis was performed to classify each prescription as off-label, unregistered or registered. Drugs considered to be unregistered were excluded at this stage. Comparison was made between the prescription details and the approved Australian PI, as contained in the edition of eMIMS that was current at the time of prescribing. The following categories of off-label prescribing were used:
A prescription was considered off-label if at any time during administration, it met at least one of these four criteria. Prescriptions were allocated to multiple off-label categories where applicable. Categorised levels of dosing information, as devised by Tan et al., were used to determine off-label use regarding the patient's age.
Ethics approval had been obtained from the Human Research Ethics Committee (Tasmania) Network.
Of the 300 paediatric inpatients, 56% were male, and the median age was 2 years, 6 months (range: 1 day to 11 years, 11 months). The median length of hospital stay was two nights. There were no statistically significant differences in patient demographics, admission diagnoses or pattern of medication use between the two data collection periods; consequently, all data were combined. Twenty patients did not receive any medications while in hospital. Medications fitting the inclusion criteria were prescribed 887 times to patients in the study, accounting for 106 different drugs. The most commonly prescribed medications were paracetamol, ibuprofen and oxycodone.
Of the 887 prescribed medications, 283 (32%) were found to be off-label for one or more reasons. One hundred and seventy-two of the 300 (57%) paediatric inpatients received at least one off-label medication. Drugs were most likely to be off-label because they were given at a dosage or frequency above that approved for patient's particular age or weight, as shown in Table 1. Twenty prescriptions were off-label according to more than one of the categories used. For the 94 drugs in the age/weight off-label category, the level of dosing information was assessed according to the categories devised by Tan et al., and is summarised in Table 2.
|Off-label category||Occurrences (% of all prescriptions)|
|The dosage given or frequency of drug use is greater than that sanctioned in the PI for that patient. This often relied on the individual patient's age and weight, and was based on the maximum individual dose and the maximum daily dose.||116 (13.1)|
|The drug is not approved for any use at the patient's particular age or weight.||94 (10.6)|
|The indication for prescribing the medication is different from those listed in the PI.||85 (9.5)|
|The route of administration differs from the approved route for that formulation in the PI.||8 (<1)|
|Level of dosing information (as defined by Tan et al.)||Occurrences (% of prescriptions off-label for age)|
|PI states or implies that safety and effectiveness of product is not established (e.g. ‘not recommended for use’, ‘limited experience with use’)||31 (33.0)|
|No suggestion that product can be used in age group||12 (12.8)|
|PI states product not approved for use in age group||45 (47.8)|
|PI states product contraindicated for use in age group||6 (6.4)|
The frequency with which each drug was used and the percentage of these uses that was considered to be off-label are shown in Table 3. Of the 106 drugs used, 51 were used in an off-label way at least once (48%). Thirty drugs were used off-label in more than 75% of prescriptions. The drugs most commonly used off-label relative to their number of prescriptions were oxycodone, salbutamol and paracetamol.
|Amoxycillin and Clavulanate||12||2||17|
|Trimethoprim and Sulfamethoxazole||9||7||78|
No written consent documents were found for any patient relating to off-label medication use. Five ADRs were identified among the studied patients; two of these reactions involved drugs used in an off-label way.
Historically, the need for drug registration has come about because of adverse incidents. In particular, incidents involving diphtheria toxoid in 1901, sulfanilamide in 1937 and thalidomide in 1961 prompted controls on quality, safety and efficacy, respectively. In each of these disasters, those who suffered or died were mostly children. It is therefore poignant that children often receive off-label medicines. Shirkey was the first to describe the problem of off-label prescribing, likening children to ‘therapeutic orphans’ due to the lack of availability of paediatric medicines.
More than half of the children in this study received at least one off-label medication while in hospital. Studies overseas also found that large numbers of hospitalised children received off-label medicines.[8-12] In the Australian context, the figure in this study is higher than that found in a dedicated paediatric hospital, where 36% of patients received off-label or unregistered drugs, but less than the 80% of neonatal intensive care unit patients who received off-label medicines.
The percentage of paediatric inpatient prescriptions that were considered off-label in this study (32%) was within the range of 19–60% found in overseas general hospitals.[8-12] However, international comparisons are difficult given varying drug registration requirements and prescribing trends. Our level of off-label prescribing was broadly comparable with the results of the three other Australian studies: 26% in outpatients, 16% in inpatients at a specialist paediatric hospital and 46% in a neonatal intensive care unit.
The drug most commonly used off-label in this study was oxycodone. This was due to the statement that the drug ‘should not be used in patients under 18 years’, found in the PI of both the liquid and tablet dosage forms. However, there are several randomised trials that have shown that oxycodone can provide effective pain relief for children,[13, 14] and dosing instructions are included in several reference texts.[15, 16] In contrast, Pokela et al. reported that the pharmacokinetics of oxycodone varied greatly in infants, and that the dose of oxycodone should be titrated individually. As a result, it would seem that a large-scale clinical trial of oxycodone in children is warranted, and the PI be subsequently amended to reflect the results.
Gentamicin was also off-label in all 12 uses of the drug, often due to single daily dosing, whereas the PI only details divided daily doses. In spite of this, single daily dosing has been described for over 15 years, with some studies finding that nephrotoxicity and ototoxicity occur less with once-daily administration in children. Other reference texts list both higher doses and single daily dosing.[15, 16] It is possible that off-label use is related to a lack of incentive or obligation for the manufacturers to update the PI for gentamicin.
The use of salbutamol was commonly off-label because it was given by metered dose inhaler at dosages considerably greater than those approved. The use of salbutamol in high doses for acute asthma is supported by reference texts and the National Asthma Council's Asthma Management Handbook, again suggesting that the PI has not been updated to reflect current clinical guidelines.[16, 19]
Ondansetron was used in an off-label way in 51% of prescriptions. Thirteen cases were off-label because the drug was used at dosages greater than the 0.1 mg/kg/dose specified in the PI for post-operative nausea and vomiting (PONV) or greater than 5 mg/m2 specified in the PI for chemotherapy-induced nausea and vomiting (CINV). Ondansetron is approved in the United States for CINV at this higher dosage. Three children under 2 years of age were given ondansetron, for whom it is not approved in Australia. However, ondansetron has been recently been approved in the United States for CINV and PONV in paediatric patients from 6 months and 1 month of age, respectively, highlighting the international inconsistency of PIs. Five patients received ondansetron for nausea and vomiting unrelated to chemotherapy or surgery, an off-label use. One of these uses was for nausea and vomiting related to gastroenteritis; a Cochrane review has found some evidence supporting ondansetron use for this indication.
A number of uses of paracetamol were at approximately 20 mg/kg/dose and up to 80 mg/kg/day. This is not listed in the approved PI; however, standard references state that paracetamol can be used in children in dosages up to 90 mg/kg daily under medical supervision, to be reviewed after 48 h.[15, 16] Elsewhere, it has been reported that liver injury from paracetamol can occur at dosages greater than 75 mg/kg/day when used for more than 2 days, and that it is not clear whether individual doses higher than 15 mg/kg will have a better therapeutic effect.
The cephalosporins, cephalexin and cephazolin, were both commonly off-label due to their prophylactic use (approval is for treatment of infection only). Ten of the off-label cephalexin uses were surrounding urinary tract surgery such as for hypospadias; eight of them while a catheter was in situ. Therapeutic Guidelines: Antibiotic recommends the use of cephalexin for recurrent urinary tract infections but does not suggest its prophylactic use in catheterised patients. Amoxycillin was off-label in 71% of uses, most of which were due to post-operative prophylaxis following tonsillectomy. Studies have investigated the use of amoxycillin following tonsillectomies, with inconsistent results regarding its benefit.
Most of the said examples demonstrate that pharmaceutical companies are unlikely to go to the expense of a submission to the regulatory agency (TGA) in order to update a medication's approved information. This is particularly significant where there is little incentive for change and the financial returns are small.
No informed consent documents were identified during the study. The need for informed consent is a contentious area. Before prescribing or administering any medication to a child, the carer and child should have the benefits and risks explained to them regardless of the registration status of the medication. Therefore, in general, it should not be necessary to take any additional steps to obtain consent when prescribing or administering an off-label medication beyond those taken when prescribing or administering a licensed one. Where there is no high-quality evidence for the use of a medication, obtaining written informed consent is recommended, together with the approval of a hospital drug committee.
An increased potential for ADRs is one of the assumed consequences of off-label prescribing, as one fifth of all ADRs have been associated with off-label prescribing due to age alone. Five ADRs were documented in this study. Two of these reactions involved drugs used in an off-label way, suggesting that off-label use did not markedly increase the incidence of ADRs. The incidence of ADRs with off-label prescribing was lower for patients in this study (1.7%) relative to other studies overseas. However, our study was retrospective and not designed to detect ADRs. Neubert et al. found that an ADR occurred in 28% of German paediatric patients receiving an off-label prescription, and Santos et al. reported a figure of 16% in Brazil.
The results of this study indicate that the debate around medications for children and their safe use needs to move away from focussing on off-label use, and towards their evidence-based use. The Orphan Drugs programme aims to encourage drug companies to bring essential products for rare conditions to Australia, and a Paediatric Medicines Advisory Group has been established to provide advice on issues relating to paediatric medicine. A priority list of medicines requiring approval for paediatric use in Australia has been created, and an Australian paediatric prescribing guide may eventuate in the coming years. Nonetheless, initiatives to improve paediatric drug testing, registration and utilisation are still required.