Globally, antibiotics are the most common drugs used to treat acutely ill children.Antibiotic doses should be selected to achieve optimal killing of bacteria with minimal undesirable effects, such as drug toxicity or the selection of resistant pathogens. Given the challenges associated with increasing antibiotic resistance, optimal use of antibiotics for children is important for global antimicrobial stewardship
J A Bielicki and colleagues compare common strategies for selecting antibiotic dose for children and discuss how best to balance usability with accuracy
Globally, antibiotics are the most common drugs used to treat acutely ill children.Antibiotic doses should be selected to achieve optimal killing of bacteria with minimal undesirable effects, such as drug toxicity or the selection of resistant pathogens. Given the challenges associated with increasing antibiotic resistance, optimal use of antibiotics for children is important for global antimicrobial stewardship.
Challenges of achieving optimal dosing in children
Drug dosing in children is more complex than in adults. As the organs and immune system develop throughout childhood, the way in which drugs are absorbed, transported, and eliminated by the body (pharmacokinetics) changes, which in turn affects the drug’s action on the body (pharmacodynamics).4 During the first two years of life, the evolution of renal function and hepatic metabolism have an important effect on the optimal antibiotic dose. Inaccurate dosing can lead to problems because higher antibiotic doses potentiate undesirable side effects, especially diarrhoea, and may promote the selection of resistant bacteria.5
Finding suitable formulations for children is also more challenging than for adults. Antibiotic doses need to be titrated to achieve optimal effectiveness and be adapted to maturational changes. Achieving the correct dose may require splitting or crushing tablets, for example.Although liquid antibiotic preparations are the most flexible to dose, measuring small volumes can be inaccurate. Generally, syringes have been found to be more accurate than dosing cups or spoons, and the use of pictures to demonstrate correct dosing can further improve accuracy.Nevertheless, in one US study more than two thirds of preventable adverse drug events outside hospital were due to parental errors in administering drugs, indicating the difficulties of providing easy to follow instructions.Antibiotics accounted for a quarter of preventable adverse drug events and more than half of non-preventable events outside hospital.
Antibiotic administration to children can be further complicated because many medicines are unpalatable. In addition, liquid formulations are cumbersome to transport and store, particularly since they may need to be refrigerated. Practically, therefore, dosing of liquids in children is dictated by the smallest volume that can be reliably measured by parents with the provided spoon or syringe. This must be balanced against the largest acceptable single dose volume for an often unpleasant tasting medicine.
Aminopenicillins: case study in dose selection
Aminopenicillins, such as amoxicillin and amoxicillin/clavulanate (co-amoxicillin), are the most commonly used oral antibiotics in children, and amoxicillin is probably the commonest single medicine given to children worldwide.Approaches to selecting oral doses for aminopenicillins illustrate the problems found for all β-lactams and many other classes of antibiotics given orally.
There are over 50 generic licensed amoxicillin products in Europe alone, and more products are available in individual countries.With such a variety of products available, selection of the appropriate formulation and dosing scheme for children of different ages is complicated. Furthermore, high quality data on the effect of selected formulations and doses of amoxicillin for children of different ages are lacking. A recent review by the European Society of Paediatric Infectious Diseases concluded that substantial uncertainty remains about optimal dosing of antibiotics for childhood pneumonia because of a lack of pharmacokinetic data. For acute otitis media, it has been shown that for many children a daily dose of 40 mg/kg per day would be inadequate to treat an infection with penicillin resistant pneumococcus.
The summary of product characteristics for amoxicillin gives the target dose as 40-90 mg/kg/day in two or three divided doses for children weighing less than 40 kg; the standard adult dose (500 mg three times daily) is recommended for children weighing more than 40 kg. This wide ranging recommended daily dose represents a pragmatic approach to treatment that is believed to result in a relatively low rate of serious negative outcomes. Although there are no robust data to support this belief, the mortality associated with childhood infections commonly treated with oral amoxicillin is very low in most high income countries. However, the effect on development of resistant bacteria is unclear.
Standardised dose selection
Regardless of the target dose, prescribers need a simple and rapid approach to identify the appropriate dose for each child. International guidance is inconsistent, with the US and much of continental Europe favouring exact weight based dosing, the UK applying age banded dosing, and the World Health Organization recommending weight banded dosing.
We compared the three main dose selection methods by simulating the accuracy of amoxicillin exposure using data on age and body weights of children in hospital from the Antibiotic Resistance and Prescribing in European Children (ARPEC) project point prevalence survey of antimicrobial prescribing. The data were collected in 61 UK paediatric units (including 44 district general hospitals and 17 tertiary units) and five African paediatric units (two in the Gambia, one in Ghana, one in South Africa, and one in Malawi).
For our simulation we defined the smallest deliverable dose as 25 mg (1 mL of 125 mg/5 mL suspension) based on existing measuring spoons and applied a maximum daily dose of 1500 mg, corresponding to the standard recommended adult dose for non-severe infection. We excluded children weighing >40 kg because they would have received the fixed (adult) doses under all three selection approaches. Infants <6 days old or weighing <3 kg were also excluded because they are unlikely to receive oral amoxicillin in the community.
For exact weight based dosing the total daily target dose was defined as 70 mg/kg. For weight banded dosing we applied the standard WHO recommendations (rather than the higher pneumonia dosing). For age banded dosing, we used the recommendations of the British National Formulary for Children.
Which dose selection approach is best?
Table 1 compares the doses that would have been received by children in the survey under the three approaches to dosing. Clearly, the most accurate way of selecting the amoxicillin dose is to weigh the child and prescribe an exact mg/kg dose. Using this approach, all children would receive a dose within the desired range, but such accurate dosing may be difficult to achieve. Solid formulations, such as the WHO recommended dispersible tablets, cannot be divided sufficiently to produce exact doses for all children and dosing of small liquid volumes may become very inaccurate and increase the risk of medication errors.
Table 1
Total daily doses of amoxicillin in simulation of three dose selection approaches in 1037 children from UK and 252 from Africa with bodyweight ≤40 kg
These problems can be overcome by using weight bands. When we applied the WHO weight bands to our sample 8% of UK children and 4% of African children would have received a total daily dose below the recommended range (table 1). This may be a particular problem in settings with high incidence of invasive bacterial infections or high prevalence of antimicrobial resistance, especially pneumococcal penicillin resistance. Daily doses at the upper end of the recommended dose spectrum may be needed to adequately treat infection when there is a high prevalence of pneumococcal penicillin resistance.
The simplest method of dose selection is to use age as a proxy measure for weight, based on the strong association between age and weight as reflected in growth charts. The British National Formulary has recommended this approach for 50 years, with dosing below the target range widely observed in community based prescribing.In our simulation, this was the least accurate method, with 18% of UK children receiving doses outside the recommended range. Eleven per cent would receive more than the recommended daily dose, and the maximum dose was 179 mg/kg. Common side effects, such as diarrhoea, occur more often with higher daily doses, which is likely to affect adherence.
In the absence of clear data to support the use of one dose selection method, practical considerations, such as those listed in table 2⇓, may influence preference for a specific approach. Although the WHO has commented on the advantages of using age banded dose selection for drugs with a wide therapeutic index, particularly in situations where a recent weight is unavailable, this approach has some major difficulties. These include defining the age bands so that they reflect rapid changes in weight (especially during the first six months of life and from 8 to 9 years of age), accounting for weight being normally distributed around the 50th centile for age, and ensuring that locally relevant weight for age standards are used. In our simulation use of UK age bands would lead to every third African child being prescribed a dose above 90 mg/kg per day, with the maximum expected daily dose being 250 mg/kg per day (table 1). These higher doses may be appropriate given the higher antibiotic resistance in the region.
Table 2
Features of different dose selection approaches for oral antibiotic treatment in childhood
Future options
Dosing is increasingly recognised as an important aspect of appropriate use of medications in childhood, especially for antibiotics. Published dose selection recommendations are now widely accessible and therefore potentially used both within the contexts they were designed for and outside of them. Consequently, international agreement on the best approaches is needed. This has been achieved for antiretroviral medications for HIV infection, for which United States, WHO, and European guidelines generally recommend weight banded dose selection.
Our data suggest that weight bands are relatively reliable in delivering a specific dose range to children and are not influenced by weight for age patterns. They also allow use of fixed dose formulations, such as dispersible tablets. However, they require an up to date weight. Although age banded approaches are appealing for their simplicity, they are influenced by regional weight for age patterns and a substantial proportion of children receive very low or high total daily doses. This may be acceptable in some settings, such as in the UK where serious bacterial infection rates and antibiotic resistance are low, and for drugs with a wide therapeutic index, where the negative effect of daily doses outside the recommended range is thought to be small. It remains to be seen whether antibiotic dose is an important factor in selection of resistant bacteria in vivo. If this were the case, a strong argument could be made for adopting weight banded dose selection to ensure selection pressure is minimised. Combined age and weight bands may be a partial solution, but they would still need to be tailored to the local context to account for variations in weight for age patterns.
Key messages
- Methods for selecting dose of oral antibiotics in childhood must balance simplicity with accuracy
- Although using exact weight is most accurate, it requires a recent weight and may not fit with fixed dose formulations
- Age bands result in a substantial proportion of children receiving doses outside the recommended range and are context specific
- Weight banded dose selection is transferable across populations and the most practical choice when up to date weight is available
Notes
Cite this as: BMJ 2015;351:h5447
Footnotes
- We thank all partners and collaborators of the ARPEC point prevalence survey for their data contributions, particularly Ann Versporten and Herman Goossens.
- Contributors and sources: This manuscript was prepared on behalf of the improving Children’s Antibiotic Prescribing collaboration, a multidisciplinary group with an interest in optimal antibiotic use in children. JAB and MS led the Antibiotic Resistance and Prescribing in European Children (ARPEC) project, which investigated optimal antibiotic use and antimicrobial resistance surveillance methodologies for children. CISB and MS lead the Neonatal and Paediatric Pharmacokinetics of Antimicrobials (NAPPA) study, which is investigating the pharmacokinetics of penicillins in neonates and children. SS is a general practitioner with a strong interest in the role of general practice in managing acute infections in children. PFL and ICKW are pharmacology researchers with a longstanding interest in the safety and use of paediatric medicines.
- Competing interests: We have read and understood BMJ policy on declaration of interests and declare MS chairs the Department of Health expert advisory committee on antimicrobial resistance and healthcare associated infection. The ARPEC project was cofunded by the European Commission Directorate General for Health and Consumers DG Sanco through the Executive Agency for Health and Consumers (agreement number 2009 11 01). CISB is funded as a clinical research fellow by the Global Research in Paediatrics Network of Excellence (GRiP), part of the European Union seventh framework programme (FP7/2007–2013, grant agreement number 261060). SS is funded by a National Institute for Health Research Career Development Fellowship (NIHR CDF-2011-04-048). The views expressed are not necessarily those of the NHS, NIHR, or Department of Health. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
- Provenance and peer review: Not commissioned; externally peer reviewed.
References
- Sturkenboom MC, Verhamme KM, Nicolosi A, et al. Drug use in children: cohort study in three European countries.BMJ2008;337:a2245.
2. Holstiege J, Schink T, Molokhia M, et al. Systemic antibiotic prescribing to paediatric outpatients in 5 European countries: a population-based cohort study. BMC Pediatrics2014;14:174.
3. Bielicki J, Lundin R, Patel S, et al. Antimicrobial stewardship for neonates and children: a global approach. Pediatr Infect Dis J2015;34:311-3.
4. Anderson BJ, Holford NHG. Understanding dosing: children are small adults, neonates are immature children. Arch Dis Child2013;98:737-44.
5. Martiney MN, Papich MG, Drusano GL. Dosing regimen matters: the importance of early intervention and rapid attainment of the pharmacokinetic/pharmacodynamic target. Antimicrob Agents Chemother2012;56:2795-805.
6. Batchelor HK, Marriott JF. Formulations for children: problems and solutions. Br J Clin Pharmacol2015;79:405-18.
7. Le Doare K, Barker CI, Irwin A, et al. Improving antibiotic prescribing for children in the resource-poor setting. Br J Clin Pharmacol2015;79:446-55.
8. Yin HS, Wolf MS, Dreyer BP, et al. Evaluation of consistency in dosing directions and measuring devices for pediatric nonprescription liquid medications. JAMA2010;304:2595-602.
9. Yin HS, Mendelsohn AL, Wolf MS, et al. Parents’ medication administration errors: role of dosing instruments and health literacy. Arch Pediatr Adolesc Med2010;164:181-6.
10. Katz MG, Kripalani S, Weiss BD. Use of pictorial aids in medication instructions: a review of the literature. Am J Health Syst Pharm 2006;63:2391-7.
11. Kaushal R, Goldmann DA, Keohane CA, et al. Adverse drug events in pediatric outpatients. Ambul Pediatr2007;7:383-9.
12. Head of Medicines Agencies. Human mutual recognition and product index.
13. Esposito S, Cohen R, Domingo JD, et al. Antibiotic therapy for pediatric community-acquired pneumonia: do we know when, what and for how long to treat? Pediatr Infect Dis J2012;31:e78-85.
14. Canafax DM, Yuan Z, Chonmaitree T, et al. Amoxicillin middle ear fluid penetration and pharmacokinetics in children with acute otitis media. Pediatr Infect Dis J1998;17:149-56.
15. Madhi SA, De Wals P, Grijalva CG, et al. The burden of childhood pneumonia in the developed world: a review of the literature. Pediatr Infect Dis J2013;32:e119-27.
16. Huijben S, Bell AS, Sim DG, et al. Aggressive chemotherapy and the selection of drug resistant pathogens. PLoS Pathogens2013;9:e1003578.
17. Paediatric Formulary Committee. British national formulary for children 2014-2015. BMJ, Royal Pharmaceutical Press, RCPCH, 2014.
18. FDA. Amoxicillin suspension—FDA prescribing information, side effects and uses.
19. World Health Organization. Pocket book of hospital care for children. 2nd ed. WHO, 2013.
20. Versporten A, Sharland M, Bielicki J, et al. The antibiotic resistance and prescribing in European children project: a neonatal and pediatric antimicrobial web-based point prevalence survey in 73 hospitals worldwide. Pediatr Infect Dis J2013;32:e242-53.
21. Chiou CC. Does penicillin remain the drug of choice for pneumococcal pneumonia in view of emerging in vitro resistance? Clin Infect Dis2006;42:234-7.
22. Saxena S, Ismael Z, Murray ML, et al. Oral penicillin prescribing for children in the UK: a comparison with BNF for children age-band recommendations. Br J Gen Pract 2014;64:e217-22.
23. Ahmed U, Spyridis N, Wong IC, et al. Dosing of oral penicillins in children: is big child=half an adult, small child=half a big child, baby=half a small child still the best we can do? BMJ2011;343:d7803.
24. Easton J, Noble S, Perry CM. Amoxicillin/clavulanic acid: a review of its use in the management of paediatric patients with acute otitis media. Drugs2003;63:311-40.
25. Make Medicines Child Size. Technical consultation on the use of pharmacokinetic analyses for paediatric medicine development. 2009. www.who.int/childmedicines/progress/Pharmacokinetic_June2009.pdf.