Objective To determine the type and frequency of manipulations of drug dosage forms required to administer smaller doses for children and the drugs involved.
Methods An experienced paediatric clinical pharmacist estimated the requirement to manipulate a medicine to achieve accurate dose administration from prescription data in all neonatal and paediatric inpatients collected over 5-day periods and information on drug dosage form availability in a regional children's hospital (RCH) and regional paediatric intensive care unit (RPICU), a regional neonatal intensive care unit (RNICU) and paediatric and neonatal wards of a district general hospital (DGH) using paper-based prescribing systems. Doses were expressed by weight. Ward stock supply with some intravenous drugs ready-to-administer was provided. The main outcome measures were the estimated requirement for dosage form manipulation, nature of the manipulation and drug name.
Results Of 5375 evaluated drug administrations, 542 (10.1%) were judged to require manipulation or measurement of a small volume (<0.2 ml). The most frequent manipulation was measurement of oral dose in volumes of 0.1 to <0.2 ml in the DGH. Requirement to measure doses of <0.1 ml (oral and intravenous) accounted for 25.2% of all manipulations, with the need to measure intravenous doses of <0.1 ml being most frequent in the RNICU and RPICU (60.4% and 31.9% of manipulations, respectively). Hydrocortisone was the drug most frequently judged to require manipulation with both measurement of small volumes for intravenous injection (RPICU and RNICU) and segmentation of tablets (RCH).
Conclusions Manipulation of medicines (including measurement of very small volumes) to provide accurate smaller doses for children is common in the hospital setting.
Statistics from Altmetric.com
Many of the medicines used for babies and children have not yet been licensed for this purpose and their use is considered ‘off-label’.1 The dosage form (eg, tablet, capsule, injection) is usually designed for adult use and may not be ‘age-appropriate’ for a child. For example, a dose required for a child may be contained in a portion of a tablet or the high concentration of drug in an injection may mean that the paediatric dose cannot be measured accurately without prior dilution.2 Even when a drug is authorised for babies and children, the dosage form may not be age-appropriate.3 Thus, manipulation of the dosage form by, for example, segmenting tablets or diluting injections may be undertaken in an attempt to administer accurately the prescribed dose for a child. There is a risk that the dose delivered will be inaccurate and that medication errors may occur when calculating and preparing doses.4 There is little information about the extent of manipulation in paediatric practice, the accuracy with which dosage forms can be manipulated or the drugs commonly involved.5 ,6
The Manipulation of Drugs Required in Children (MODRIC)2 suite of projects was established to review systematically the relevant literature, to observe and enquire about practice, to assess risk and to develop guidelines. This study forms part of the MODRIC investigation and aims to estimate the type and frequency of manipulations and the drugs involved in a regional children's hospital (RCH), a regional neonatal intensive care unit (RNICU) and in paediatric and neonatal wards of a district general hospital (DGH).
In this study manipulation is defined as the physical alteration of a drug dosage form at the point of administration for the purpose of obtaining and administering a proportion of the dose originally contained within the dosage form. Descriptions of common manipulations are given in table 1.
The manipulation of a dosage form for convenience of administration only (eg, crushing a whole tablet, dispersing in water or mixing with food and administering orally) was not considered in this study. Extemporaneous dispensing or compounding undertaken in the hospital pharmacy (often to avoid manipulation at the point of administration) was similarly not considered to be manipulation.
The measurement of very small volumes of injections or oral liquids (classified as <0.1 ml and 0.1 to <0.2 ml) were also included since the dosage form (injection or oral liquid) is frequently diluted to allow an accurate measurement of volume.
All inpatient prescriptions written on paper-based systems were reviewed during three 5-day periods in a RCH (309 beds including 15 high dependency beds), a DGH with paediatric wards (26 beds) and a neonatal unit (16 cots, including three intensive care cots) and in a RNICU (48 cots, including 18 intensive care cots) between February and June 2011. Ward stock models of supply were in use with supply of non-stock medicines from the pharmacy to individual patients. Some intravenous injections (eg, antibiotics in the RCH) and infusions (eg, dopamine, dobutamine in a regional paediatric intensive care unit (RPICU)) were prepared ready-to-administer by the pharmacy. Research pharmacists and nurses recorded data on the drugs prescribed and administered including drug name, dosage form, dose, route and frequency of administration. An experienced paediatric clinical pharmacist (AN) determined whether a manipulation would be required to deliver the dose prescribed from the dosage forms and concentrations known to be available in the hospitals. If an age-appropriate formulation (including extemporaneous or compounded formulations) was known to be available in the hospital, it was assumed that it was used.
Data for the RPICU (care level 4; 23 beds) of the RCH were recorded and analysed separately.
If there was insufficient information to determine whether a manipulation would be required, the administration was recorded as unassessable. All data were anonymised.
A total of 5495 drug administrations were recorded during the 5-day periods; 120 administrations (2.2%) were unassessable leaving 5375 for evaluation. Of these, 542 (10.1%) were judged to require manipulation as defined.
The number of manipulations in each of the different care settings is shown in table 2. The RCH (RCH + RPICU) accounted for 339 manipulations in 3633 assessable doses (9.3%). The largest proportion of manipulations was judged to be required in the DGH. One type of administration (oral dose 0.1 to <0.2 ml domperidone, ranitidine and tocopherol) accounted for 107 (71%) of the manipulations anticipated in this setting.
The types of manipulation judged to be required in the different neonatal and paediatric care settings are shown in table 3. Overall, the most frequent manipulation was measurement of an oral dose in volumes of 0.1 to <0.2 ml, and this was required most frequently in the DGH. The requirement to measure doses of <0.1 ml (oral and intravenous) accounted for 25.2% of all manipulations, with the need to measure intravenous doses of <0.1 ml being most frequent in the RNICU and RPICU (60.4% and 31.9% of the manipulations in these settings, respectively). There was also a frequent need to measure volumes for intravenous injection of 0.1 to <0.2 ml in these two setting (34.0% and 40.4% of manipulations, respectively).
The cutting of transdermal patches was not required during the 5-day data collection periods.
table 4 shows the drugs most commonly judged to require manipulation and the type of manipulation. Drugs are only included in the table if more than four manipulations (of all types) were recorded from the care settings. Consequently, 20 different drugs involved in 54 manipulations of five different types have been excluded from table 4. Hydrocortisone was the drug most frequently judged to require manipulation with both measurement of small volumes for intravenous injection (RPICU and RNICU) and segmentation of tablets (RCH).
The most frequent manipulation in the RCH was dispersion of diclofenac dispersible tablets and administration of a proportion of the resulting dispersion. In the RPICU the most frequent manipulation was measurement of <0.1 ml intravenous fentanyl, in the RNICU the most frequent manipulation was measurement of <0.1 ml intravenous phenobarbital and in the DGH measurement of small volumes (<0.2 ml) of oral domperidone was the most frequent manipulation undertaken.
The results of this study suggest that manipulation of dosage forms to allow administration of an accurate dose to children is common in a variety of paediatric care settings. In a small observational study on paediatric wards, Skwierczynski and Conroy7 showed that 10% of drug administrations required manipulation but they included manipulations such as crushing tablets and administering all of the dosage form and did not include PICU, NICU or DGH settings.
For many types of manipulation there is no published scientific evidence to support the processes.8 The general evidence suggests the potential for significant variability in the dose delivered when preparations are manipulated, but there is little evidence about outcomes in children.9 Pharmacopoeia and other regulatory requirements expect that a dosage form will consistently deliver 85–115% of the stated dose but, for most types of manipulation, the dose delivered has not been investigated and, where information is available (mainly for splitting tablets), there is evidence that this standard is not achieved and is dependent on many different factors including the tablet brand, the apparatus used for splitting and the experience of the manipulators.10 ,11
Broadhurst et al investigated dispersion of aspirin tablets in water and demonstrated that aliquots contained 23–147% of the intended dose.12 In this study the most frequent manipulation of a drug in tablet form was the dispersion of diclofenac to obtain smaller doses by measuring aliquots. Diclofenac dispersible tablets (Voltarol Dispersible tablets 50 mg; Novartis, UK) are authorised in the UK but not for children, and there is no information in the product literature about the accuracy of the manipulation or the volume of water in which diclofenac would completely dissolve.
Splitting tablets into halves or quarters may deliver up to 86% deviation from the intended dose, with segments cut from scored and unscored tablets with a proprietary tablet splitter achieving greater accuracy than manual breaking or cutting with scissors.13
Paracetamol suppositories of 80, 120 and 325 mg strengths yielded 69–195% of the intended half dose when split by anaesthetists.14 Only four manipulations of suppositories were required during the study period. In the hospitals concerned a variety of strengths of paracetamol suppositories was available, some of them unlicensed, reducing the requirement for manipulation.
In this study 25% of drugs manipulated would require measurement of <0.1 ml or a dilution (usually 1:10 parts) to permit a larger volume to be measured. Most of these measurements were for low therapeutic index intravenous drugs such as fentanyl and phenobarbital in the NICU or PICU. It is difficult to measure volumes of <0.1 ml accurately using standard 1 ml syringes. Misinterpretation or miscalculation during measurement or dilution can produce tenfold overdoses.15 The dead space in the hub of the syringe contains a large volume in relation to the dose which, if rinsed and injected or included in the dilution, can introduce substantial error. There may be similar problems when measuring volumes of 0.1–0.2 ml, and a lack of familiarity with the markings for hundredths of 1 ml have also been associated with medication error. Many liquid medicines (intravenous and oral) are produced in concentrations that make for easy calculation and measurement for older children and adults. When used for neonates and infants, the doses required are contained in very small volumes so there is a need for lower strengths and amounts in vials (for intravenous drugs) that would not allow measurement of a tenfold overdose.16 17 A specific warning about intravenous paracetamol has been issued by the regulatory agency in the UK.18
When prescribing medicines, paediatricians have successfully been persuaded to use the doses quoted in formularies such as the British National Formulary for Children.19 However, prescribing of accurate doses based on mg/kg or mg/m2 without knowledge of the drug preparations available can lead to attempts to manipulate dosage forms to deliver the prescribed dose. Unless the therapeutic index of the drug is low (eg, fentanyl, morphine, phenobarbital, vecuronium), sensible rounding of the dose may mean that the whole of a dosage form can be safely administered or that the volume to be administered is easily calculated and measured.20 It may be safer to use dosage bands corresponding to appropriate dosage forms or easily measured volumes when the therapeutic index is large and for public health campaigns.21
Babies and children should have access to dosage forms which allow accurate administration of the required dose in a form which the child finds acceptable. Recent European regulations on ‘better medicines for children’22 requires manufacturers to study new drugs in children and to provide age-appropriate dosage forms if there is likely to be a therapeutic benefit. However, this has done little to increase the age-appropriate dosage forms of existing drugs which are used off-label for babies and children. It seems inevitable that manipulation of dosage forms to provide accurate smaller doses for children will continue. In some circumstances no age-appropriate dosage form will be available and, in others, children will choose manipulation over an age-appropriate alternative (eg, preferring to swallow half a tablet over the taste of a liquid medicine). It may not be of concern to the prescriber if there is some inaccuracy of dose delivery, the pragmatic view being ‘better a less accurate dose than no dose at all’. In our experience many caregivers would share this view when rationalising the practical difficulties of administering medicines to uncooperative children. At present, however, there is little information to guide those who manipulate medicines, and the MODRIC programme will provide guidance based on evidence, risk assessment and pragmatism. Clearly more research evidence about dosage accuracy would help determine which drugs, dosage forms and types of manipulation should be supported and which might cause harm.
The limitations of this study should be recognised. Actual manipulation of dosage forms was not observed in this part of the study but deduced by a single experienced assessor from the dose prescribed and dosage forms known to be available. While direct observation should be considered the ‘gold standard’ for determining manipulation of dosage forms, the logistics of conducting such a study for large numbers of patients are difficult and would require many observers (R Richey, personal communication). In practice it is possible that additional manipulations are undertaken because the appropriate dosage form to deliver the dose was not available at the point of administration, because of patient choice or because manipulation was unnecessary as an extemporaneous preparation was compounded by the pharmacist but not recorded on the prescription.
Manipulation of dosage forms (including measurement of very small volumes) in an attempt to provide accurate doses to babies and children is common in tertiary and secondary care practice. There is little published evidence concerning the accuracy of the dose to support the practices, but currently the lack of availability of specific paediatric medicines and the pragmatic approach of carers means that manipulation will continue. Further research is required to determine which types of manipulation and drug can be supported and which might harm children.
Manipulation of dosage forms to obtain accurate smaller doses for babies and children is common in UK hospitals.
There is little evidence to support the accuracy of dosing when dosage forms are manipulated.
Attempts to measure very small volumes are made when delivering doses to neonates and patients in the paediatric ICU. This may be inaccurate or lead to miscalculation.
The drug concentration and volume of injection ampoules and vials may lead to undetected medication error because miscalculated doses can be administered from single containers usually intended for older children or adults.
The authors acknowledge the assistance of Christine Donnell and the ADRIC (Adverse Drug Reactions in Children) research team in collecting prescribing and administration data.
Contributors AN, RR, US, CB, JC, MP and MT conceived and designed the study. RR organised and collected the data. AN interpreted the data and prepared the article. RR, JF and MT assisted in data interpretation. All authors contributed to critical revision of the article and agreed the final draft.
Funding Funding for researcher salaries and overheads was provided by the UK National Institute for Health Research (NIHR) under its Research for Patient Benefit (RfPB) programme (grant reference number PB-PG-0807-13260).
Disclaimer This paper presents independent research commissioned by the NIHR. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.
Competing interests None.
Ethics approval All MODRIC studies were considered by the local ethics committee and were judged to not require formal approval or informed consent. Local R&D approval was obtained where required.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement Additional data about less frequently manipulated drugs, types of manipulation and ages of patients can be made available to researchers by application to the corresponding author.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.