The Worldwide Antibiotic Resistance and Prescribing in European Children (ARPEC) point prevalence survey: developing hospital-quality indicators of antibiotic prescribing for children

Abstract

Objectives

Previously, web-based tools for cross-sectional antimicrobial point prevalence surveys (PPSs) have been used in adults to develop indicators of quality improvement. We aimed to determine the feasibility of developing similar quality indicators of improved antimicrobial prescribing focusing specifically on hospitalized neonates and children worldwide.

Methods

A standardized antimicrobial PPS method was employed. Included were all inpatient children and neonates receiving an antimicrobial at 8:00 am on the day of the PPS. Denominators included the total number of inpatients. A web-based application was used for data entry, validation and reporting. We analysed 2012 data from 226 hospitals (H) in 41 countries (C) from Europe (174H; 24C), Africa (6H; 4C), Asia (25H; 8C), Australia (6H), Latin America (11H; 3C) and North America (4H).

Results

Of 17 693 admissions, 6499 (36.7%) inpatients received at least one antimicrobial, but this varied considerably between wards and regions. Potential indicators included very high broad-spectrum antibiotic prescribing in children of mainly ceftriaxone (ranked first in Eastern Europe, 31.3%; Asia, 13.0%; Southern Europe, 9.8%), cefepime (ranked third in North America, 7.8%) and meropenem (ranked first in Latin America, 13.1%). The survey identified worryingly high use of critically important antibiotics for hospital-acquired infections in neonates (34.9%; range from 14.2% in Africa to 68.0% in Latin America) compared with children (28.3%; range from 14.5% in Africa to 48.9% in Latin America). Parenteral administration was very common among children in Asia (88%), Latin America (81%) and Europe (67%). Documentation of the reasons for antibiotic prescribing was lowest in Latin America (52%). Prolonged surgical prophylaxis rates ranged from 78% (Europe) to 84% (Latin America).

Conclusions

Simple web-based PPS tools provide a feasible method to identify areas for improvement of antibiotic use, to set benchmarks and to monitor future interventions in hospitalized neonates and children. To our knowledge, this study has derived the first global quality indicators for antibiotic use in hospitalized neonates and children.

Introduction

Antimicrobial resistance is a major public health problem and continuing progress in the treatment of many infections is threatened by the growing resistance of pathogens to antimicrobial agents.¹ Judicious use of antibiotics is essential to slow the emergence of antibiotic resistance in bacteria and extend the useful lifetime of effective antibiotics.² The 2011 European Commission Action Plan against the rising threats from antimicrobial resistance emphasized the importance of surveillance data with regard to antimicrobial use and resistance at local and national levels and the role of antibiotic stewardship.³ Also, President Obama of the USA has recently increased efforts to combat and prevent antibiotic resistance, among which is the core aim of improving the quality of antibiotic prescribing.^4,5

Access to standardized and validated antibiotic surveillance data is essential to assess the appropriateness of antimicrobial prescriptions and to set and monitor outcomes of interventions. There is as yet no consensus regarding the use of indicators to monitor trends of antibiotic use in hospitals.⁶ The European Surveillance of Antimicrobial Consumption (ESAC) project developed and validated a web-based point prevalence survey (PPS) tool to evaluate hospital antimicrobial prescribing rates at local and national level. The ESAC-PPS identified inappropriate use linked to specific agents or specialties, proposed quality indicators of prescribing practices and identified targets to improve antimicrobial prescribing.^7–9 The ESAC-PPS tool was used to set quantitative targets for improved antibiotic prescribing in adults and to measure the effectiveness of interventions to reach these targets through repeated PPSs.^10,11

Currently, no comparable detailed information on antibiotic use in hospitalized neonates and children is available. Longitudinal surveys of antimicrobial use in neonates and children are in part lacking because there is no consensus on the indicators that should be used to monitor trends of antibiotic use. PPS tools were not specifically designed for capturing antimicrobial prescribing data in this population. As part of the Antibiotic Resistance and Prescribing in European Children (ARPEC) project,¹² the ESAC-PPS tool was specifically adapted to survey antimicrobial use in hospitalized neonates and children (ARPEC-PPS; work package 5). The feasibility of carrying out the ARPEC-PPS method in a large number of hospitals was piloted in 73 hospitals from 23 different countries in 2011.¹³ After a successful pilot study, this survey was subsequently rolled out as a global ARPEC-PPS in a very large set of hospitalized neonates and children worldwide.

The study aimed to describe antibiotic prescribing practices among hospitalized neonates and children worldwide and to determine the feasibility of adapting existing adult quality indicators of optimal prescribing to neonates and children.

Materials and methods

Ethics

This study was a completely anonymized audit of current antimicrobial prescribing practices. No unique identifiers were entered into the database. Every patient record was given a unique non-identifiable survey number, which was automatically generated by a computer program specifically designed for anonymous data entry. Formal ethics approval for this study depended on the country and was taken care of by each participating hospital if required.

Countries and hospitals

Paediatric infectious disease specialists, clinical microbiologists and pharmacists, and other interested healthcare professionals from established networks such as the European Society of Paediatric Infectious Diseases¹⁴ and Global Research in Paediatrics,¹⁵ were invited to participate in this worldwide ARPEC-PPS conducted in October–November 2012. We collected data from 226 hospitals from 41 countries belonging to the six United Nations (UN) regions (Africa, Asia, Oceania, Latin America, North America and Europe).¹⁶ For Europe, four geographical UN sub-regions were defined:¹⁶ Eastern Europe (Hungary, 1 hospital; Poland, 1 hospital; Romania, 2 hospitals), Northern Europe (Denmark, 4; Estonia, 2; Finland, 1; Latvia, 10; Lithuania, 1; UK, 65), Southern Europe {Croatia, 2; Greece, 5; Italy, 7; Kosovo [in accordance with UN Security Council resolution 1244 (1999)], 3; Former Yugoslav Republic of Macedonia, 1; Malta, 1; Portugal, 3; Slovenia, 4; Spain, 13}, Western Europe (Belgium, 14; France, 9; Germany, 22; Luxembourg, 1; the Netherlands, 1; Switzerland, 1). Hospital numbers for other regions were as follows: Africa (Gambia, 2; Ghana, 2; Malawi, 1; South Africa, 1); Asia (Bahrain, 1; Georgia, 6; India, 8; Iran, 3; Kuwait, 3; Oman, 1; Saudi Arabia, 2; Singapore, 1); Oceania (Australia, 6); Latin America (Argentina, 1; Colombia, 6; Mexico, 4); and North America (USA, 4).

Hospitals were classified as (i) primary, (ii) secondary and (iii) tertiary, specialized care and infectious diseases hospitals.¹⁷

Four main paediatric ward types were defined: general paediatric medical wards (GPMWs); paediatric surgical wards (PSWs); paediatric intensive care units (PICUs); and specialized paediatric medical wards (SPMWs). The last type included haematology/oncology special paediatric medical wards (HO-SPMWs), cardiology wards (C-SPMWs), transplant wards (bone marrow transplant/solid) (T-SPMWs) and all others (other-SPMWs). Neonatal wards included three aggregated neonatal intensive care unit (NICU) levels providing special care (NICU level 1), high-dependency care to low birth weight neonates (NICU level 2) or tertiary referral care to very low birth weight neonates (NICU level 3), and a general neonatal medical ward (GNMW).¹⁸

Data collection

As described in full detail elsewhere,¹³ participants were asked to conduct a 1 day cross-sectional PPS during which all paediatric and neonatal wards had to be audited once within a fixed period of time. The PPS included all neonates (<30 days) and children <18 years old present in the ward at 8:00 am at least since midnight on the day of the survey. Data collection was based on the standardized ESAC-PPS protocol. For each patient receiving at least one antimicrobial, mandatory data included prescribed antimicrobial agent (dose per administration, number of doses per day and route of administration), age and gender, indication for treatment (community-acquired, hospital-acquired, surgical prophylaxis and medical prophylaxis), reason for treatment according to a predefined list, speciality (medicine, surgery or intensive care) and whether the reasons for treatment were documented in the patient notes. Information on surgical prophylaxis was captured for the previous 24 h, indicating 1 dose, 1 day or >1 day. The ESAC-PPS protocol was further adapted for the purpose of collecting data on children and neonates. Additional mandatory variables included: underlying diagnosis according to predefined lists of common medical and surgical conditions in children and neonates; whether the treatment was targeted (based upon microbiological culture and sensitivity testing) or empirical; and current weight and ventilation status. For premature neonates, birth weight and gestational age were also collected. Denominators included the total number of patients present on the ward at 8:00 am and the total number of beds by the predefined six different paediatric and four neonatal department types.

Drugs were classified according to the standardized and internationally recognized WHO Anatomical Therapeutic Chemical (ATC) classification system classifying drugs according to their main therapeutic use.¹⁹ Antibiotics were grouped into antibacterials for systemic use (ATC J01), oral metronidazole (ATC P01AB01), intestinal anti-infectives (oral antibiotics vancomycin, paromomycin and colistin, ATC A07AA) and antibiotic drugs for the treatment of tuberculosis (rifampicin, ATC J04AB02). Antimycotics for systemic use (ATC J02) were merged with antifungals classified under the intestinal anti-infectives (ATC A07AA: oral nystatin, amphotericin B, miconazole). We further collected data on antivirals for systemic use (ATC J05), nitroimidazole derivatives (ATC P01AB) and antimalarials (ATC P01B).

Data collection was performed using paper forms, after which data were entered into a central database using a web-based application for data entry, validation and reporting. All data were mandatorily entered online. The software was designed to avoid missing data (e.g. it was not possible to enter incomplete patient and denominator files online and complete the submission). The data validation procedure involved extra checks on erroneous data entry, e.g. extremely high antibiotic prevalence rates, possibly involving wrong denominators, and extremely high dosing. All data were completely anonymously entered into the database and safeguarded at the University of Antwerp, Belgium. A helpdesk as well as a frequently asked question list were available in support of the participants. Participation was exclusively on a voluntary basis and the numbers of hospitals and patients were not intended to be representative of a country or region. Depending on the countries' legal requirements, hospitals had to comply with local ethics approval. A data privacy excerpt document was made available for this purpose. Informed consent was not needed because the survey did not require direct involvement or contact with the patient, treatment or other intervention.

Data analysis

For this paper, antimicrobial use is reported as the number of treated patients and the number of therapies. Therapy was defined as the use of one substance in one route of administration. Antimicrobial prescribing rates and the derived potential quality indicators (Table 1) are expressed as percentages (proportional use), means and/or ranges aggregated at UN regional level,¹⁶ by ward type, by indication (therapeutic or prophylactic antibiotic prescribing) or according to age category (children aged >1 month and neonates aged <30 days). We also ranked the number of antibacterials for systemic use (ATC code J01) accounting for 90% and 75% of (antibiotic) drug utilization (DU90% and DU75%, respectively).

Table 1.

Overview of the suggested antibiotic quality indicators

	Potential antimicrobial quality indicators for hospitalized neonates and children
1.	Documentation of the reason for antimicrobial prescribing in the notesa
2.	Targeted therapeutic antibiotic prescribinga
3.	Parenteral administration of antibioticsa
4.	Number of antibiotic combination therapiesa
5.	Broad-spectrum antibiotic prescribingb
6.	Antibiotic prevalence rates for hospital-acquired infectionsc
7.	Targeted broad-spectrum antibiotic prescribing for hospital-acquired infectionsc
8.	Empirical broad-spectrum antibiotic prescribing for community-acquired infectionsc
9.	Broad-spectrum antibiotic prescribing for surgical prophylaxisd
10.	Prolonged antibiotic prescribing for surgical prophylaxisd

^aSee Table 2.

^bSee Figures 1 and 2 and Tables 3 and 4.

^cSee the Results section ‘Therapeutic prescribing’ and Tables 6 and 7.

^dSee the Results section, last paragraph.

Results

General overview

The 2012 ARPEC-PPS was performed in 226 hospitals from 41 countries and included 22 primary (515 patients, 3%), 81 secondary (3074 patients, 17%) and 123 tertiary (14 104 patients, 80%) care hospitals. Data on admissions to 1482 wards (17 693 patients) were recorded, and included 444 GPMWs (5298 patients), 327 SPMWs (3584 patients), 240 NICUs (3515 patients), 201 PSWs (2273 patients), 135 PICUs (1062 patients) and 135 GNMWs (1961 patients). Overall, 56% of treated children were males (range, from 49% in Africa to 59% in Latin America). Among antimicrobial prescriptions for children, antibiotics represented 85.7% (n = 7987), followed by antimycotics (9.6%, n = 891) and antivirals (4.7%, n = 446). Among antimicrobial prescriptions for neonates, antibiotics represented 89.2% (n = 2298), followed by antimycotics (8.8%, n = 227) and antivirals (1.9%, n = 50).

Potential indicators of antibiotic prescribing

Table 2 provides antibiotic prevalence rates and selected quality indicators by ward type. Out of 17 693 patients, 6499 (36.7%) received at least one antimicrobial, with the highest rates observed among PICUs (61.3%) and SPMWs (46.0%). Overall baseline recording of the reason for antibiotic therapy or prophylaxis was 73.3% (range, from 81.7% in Africa to 53.3% in Latin America). Overall proportional targeted therapeutic antibiotic prescribing was 22.3% (Latin America, 28.9%; North America, 27.9%; Asia and Australia, 27.8%; Africa, 21.8%; Europe, 19.5%). At least one parenteral antibiotic was recorded in 78.6%, with substantial regional variation among children aged >1 month (Asia, 88.2%; Latin America, 81.2%; Africa, Europe and Australia, 68.7%; North America, 57.5%), while being higher for neonates (up to 98.2%). In total, 322 different antibiotic combinations at ATC4 level (chemical subgroup; see e.g. Figure 1 for a subset of the classification categories) were recorded worldwide.

Table 2.

Overview of antibiotic prevalence rates and quality indicators by ward type

Ward type	Patients (n)	Patients treated [n (%)]	Prescriptions (n)	Reason in notes (%)	Targeted treatment (%)	Parenteral administration (% of patients)	Antibiotic combination (% of patients)
Paediatric intensive care unit (PICU)	1062	651 (61.3)	1113	74.1	25.7	89.7	50.7
Specialized paediatric medical ward (SPMW)	3584	1648 (46.0)	2662	67.9	27.4	66.1	40.7
General paediatric medical ward (GPMW)	5298	2091 (39.5)	2929	76.4	19.7	74.3	31.6
Paediatric surgical ward (PSW)	2273	803 (35.3)	1135	63.3	19.0	78.1	32.6
Neonatal intensive care unit (NICU)	3515	1065 (30.3)	1933	82.7	18.7	95.6	71.2
General neonatal medical ward (GNMW)	1961	241 (12.3)	424	68.2	27.9	96.7	70.1
Total	17 693	6499 (36.7)	10 196	73.3	22.3	78.6	43.8

All indicators were calculated at patient level, with the exception of Reason in notes and Targeted treatment, which were calculated at prescription level. For targeted treatment, selection has been made for therapeutic antibiotic use only (HAI and CAI).

Gambia, with outlying results for targeted use, was removed from analyses.

Figure 1.

Proportion of prescribed antibiotics (ATC4 level) among children (>1 month) by region. The pale grey part of the stacked bars represents other antibacterial subgroups. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Antibiotic drug utilization at regional level in children

Figure 1 reports the most commonly prescribed antibiotics among children. In all regions except North America, β-lactams made up more than half of all antibiotic prescriptions. Third-generation cephalosporins were most frequently prescribed in Eastern Europe (35.7%) and Asia (28.6%), fourth-generation cephalosporins (cefepime) in North America (7.8%) and carbapenems in Latin America (13.3%). A considerable amount of narrow-spectrum antibiotics was prescribed in Africa and Northern Europe (e.g. β-lactamase-sensitive penicillins, 11.0% and 4.3%, respectively) and Australia (e.g. first-generation cephalosporins, 9.6%).

Table 3 reports the specific antibiotics prescribed to children by UN region, ranked at overall DU90%. In total, 77 different systemic antibiotic substances (ATC J01) were administered to children. Ceftriaxone (8.5% of total use of antibiotics) ranked first in Eastern Europe, Asia and Southern Europe (31.3%, 13.0% and 9.8%, respectively). In Western Europe and Australia, cefotaxime was more popular than ceftriaxone (4.8% and 5.6%, respectively), while cefepime was more commonly prescribed in North America (7.8%). The second most prescribed antibiotic was sulfamethoxazole/trimethoprim (7.2% of total antibiotic use), ranking first in Western Europe (11.4%), North America (11.2%) and Australia (10.0%). Vancomycin ranked second in Latin America (12.9%), North America (10.6%) and Asia (8.5%). Meropenem ranked first in Latin America (13.1%).

Table 3.

Most prescribed antibiotics (ATC J01; 5th level) to children (≥30 days old) (%) by UN region^a, ranked at overall drug utilization 90% (DU90%)

Eastern Europe (n = 112)		Northern Europe (n = 1806)		Southern Europe (n = 1558)		Western Europe (n = 1157)		Africa (n = 491)		Asia (n = 1094)		Australia (n = 502)		Latin America (n = 528)		North America (n = 463)
ceftriaxone	31.3	amoxicillin/inhib.	13.6	ceftriaxone	9.8	sulfa./trimethoprim	11.4	gentamicin	16.3	ceftriaxone	13.0	sulfa./trimethoprim	10.0	meropenem	13.1	sulfa./trimethoprim	11.2
ampicillin	10.7	amoxicillin	7.1	amoxicillin/inhib.	7.6	amoxicillin/inhib.	8.1	ceftriaxone	14.1	vancomycin	8.5	piperacillin/inhib.	7.6	vancomycin	12.9	vancomycin	10.6
sulfa./trimethoprim	8.0	ceftriaxone	6.9	sulfa./trimethoprim	7.4	cefuroxime	6.7	benzylpenicillin	10.2	cefotaxime	8.2	gentamicin	7.6	ceftriaxone	11.6	cefepime	7.8
cefuroxime	5.4	cefuroxime	5.5	piperacillin/inhib.	5.5	amoxicillin	5.6	cefuroxime	9.8	amikacin	7.4	cefazolin	7.6	clindamycin	7.6	azithromycin	7.6
ampicillin/inhib.	4.5	sulfa./trimethoprim	5.3	meropenem	4.7	vancomycin	4.8	sulfa./trimethoprim	8.1	meropenem	7.1	amoxicillin	6.2	amikacin	6.6	clindamycin	6.5
procaine benzylpenicillin	4.5	flucloxacillin	5.3	gentamicin	4.6	cefotaxime	4.8	ampicillin	5.1	piperacillin/inhib.	5.4	flucloxacillin	6.2	piperacillin/inhib.	5.1	cefazolin	6.5
amikacin	3.6	gentamicin	5.1	teicoplanin	4.6	meropenem	4.6	cloxacillin	4.1	ceftazidime	4.5	ticarcillin	6.0	cefalotin	4.7	piperacillin/inhib.	5.8
azithromycin	3.6	azithromycin	4.5	metronidazole	4.5	piperacillin/inhib.	4.5	chloramphenicol	3.5	cefuroxime	4.2	cefotaxime	5.6	ciprofloxacin	3.2	amoxicillin	5.0
levofloxacin	3.6	cefotaxime	4.4	cefotaxime	4.3	gentamicin	4.2	amoxicillin/inhib.	3.3	metronidazole	4.2	vancomycin	5.2	sulfa./trimethoprim	3.0	ceftriaxone	4.3
vancomycin	2.7	piperacillin/inhib.	4.0	ceftazidime	4.0	colistin	3.8	ciprofloxacin	3.3	sulfa./trimethoprim	3.7	amoxicillin/inhib.	3.8	ampicillin	3.0	gentamicin	3.5
piperacillin/inhib.	2.7	clarithromycin	3.9	amikacin	3.9	metronidazole	3.5	metronidazole	2.9	ampicillin/inhib.	3.1	metronidazole	3.8	trimethoprim	3.0	ampicillin	3.5
gentamicin	2.7	metronidazole	3.7	ampicillin	3.9	ampicillin	3.0	meropenem	2.4	clindamycin	2.9	azithromycin	3.8	amoxicillin	2.8	tobramycin	3.5
cefoperazone comb.	2.7	meropenem	3.5	cefuroxime	3.6	teicoplanin	2.9	amikacin	2.4	clarithromycin	2.7	ceftriaxone	3.2	metronidazole	2.7	metronidazole	2.4
nalidixic acid	2.7	benzylpenicillin	3.0	vancomycin	3.5	ceftriaxone	2.9	flucloxacillin	2.4	amoxicillin/inhib.	2.7	tobramycin	2.8	ampicillin/inhib.	2.7	ciprofloxacin	2.4
meropenem	1.8	vancomycin	2.4	ciprofloxacin	2.8	ceftazidime	2.8	amoxicillin	1.8	cefazolin	2.3	ciprofloxacin	2.6	cloxacillin	2.7	cefalexin	2.4
		ceftazidime	2.4	clarithromycin	2.5	cefazolin	2.2			ampicillin	2.2	benzylpenicillin	2.2	cefotaxime	2.1	trimethoprim	2.2
		ciprofloxacin	2.1	clindamycin	2.3	amikacin	2.1			cefoperazone. comb.	2.0	cefalexin	2.0	clarithromycin	1.9	phenoxymethylpenicillin	1.9
		tobramycin	1.9	amoxicillin	2.1	ciprofloxacin	2.1			gentamicin	1.6	meropenem	1.6	cefepime	1.7	erythromycin	1.9
		cefazolin	1.7	ampicillin/inhib.	1.9	tobramycin	1.9			ciprofloxacin	1.5	lincomycin	1.6			meropenem	1.7
		clindamycin	1.7	trimethoprim	1.7	imipenem/inhib.	1.9			cloxacillin	1.5	amikacin	1.4
		teicoplanin	1.6	ceforanide	1.5	trimethoprim	1.6			azithromycin	1.2
				cefazolin	1.4	phenoxymethylpenicillin	1.5
				azithromycin	1.3	clindamycin	1.4
				cefepime	0.9	azithromycin	1.1
						ampicillin/inhib.	1.1

sulfa., sulfamethoxazole; inhib., β-lactamase inhibitor; comb., combination.

Grey shading indicates drug utilization 75% (DU75%) by UN region.

^aNumber of countries included from each region: Eastern Europe (3), Northern Europe (6), Southern Europe (9), Western Europe (6), Africa (4), Asia (8), Australia (1), Latin America (3), North America (1).

Antibiotic drug utilization at regional level in neonates

Figure 2 shows the most prevalently prescribed antibiotics among neonates. Aminoglycosides in combination with either ampicillin/amoxicillin (Western and Southern Europe, Asia, Latin America and North America) or benzylpenicillin (Northern Europe, Africa and Australia) were by far the most frequently used regimens.

Figure 2.

Proportion of prescribed antibiotics (ATC4 level) among neonates (<30 days) by region. The pale grey part of the stacked bars represents other antibacterial subgroups. There are no data for Eastern Europe. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Table 4 shows the antibiotics prescribed to neonates by UN region, ranked at overall DU90%. In total, 43 different systemic antibiotic substances were prescribed to neonates worldwide. Gentamicin (23.6% of total antibiotic use) ranked first in Africa (36.3%), Australia (32.8%), North America (31.3%) and Northern Europe (29.1%). The second most prescribed antibiotic was ampicillin (15.3% of total antibiotic use), ranking first in Southern Europe (29.5%), Latin America (29.0%), Asia (24.3%) and Western Europe (18.4%). The third most popular antibiotic was benzylpenicillin (12.8% of total antibiotic use), ranking second in Northern Europe (28.2%), Australia (24.6%) and Africa (21.6%), while absent in the DU90% of Southern and Western Europe, Latin America and North America. Aminoglycosides were most often prescribed in combination with ampicillin or benzylpenicillin.

Table 4.

Most prescribed antibiotics (ATC J01; 5th level) to neonates (<30 days old) (%) by UN region^a, ranked at overall drug utilization 90% (DU90%)

Northern Europe (n = 674)		Southern Europe (n = 386)		Western Europe (n = 332)		Africa (n = 245)		Asia (n = 416)		Australia (n = 122)		Latin America (n = 69)		North America (n = 48)
gentamicin	29.1	ampicillin	29.5	ampicillin	18.4	gentamicin	36.3	ampicillin	24.3	gentamicin	32.8	ampicillin	29.0	gentamicin	31.3
benzylpenicillin	28.2	gentamicin	23.6	gentamicin	12.7	benzylpenicillin	21.6	gentamicin	13.9	benzylpenicillin	24.6	gentamicin	14.5	ampicillin	29.2
cefotaxime	10.8	vancomycin	9.8	cefotaxime	10.2	ampicillin	13.5	amikacin	11.8	cefotaxime	12.3	vancomycin	13.0	amoxicillin	10.4
amoxicillin	5.0	amikacin	7.0	vancomycin	10.2	cefuroxime	5.7	cefotaxime	9.9	amoxicillin	9.0	amikacin	11.6	cefazolin	8.3
vancomycin	4.3	meropenem	5.2	amoxicillin	7.5	meropenem	4.9	meropenem	8.7	vancomycin	8.2	cefotaxime	8.7	cefotaxime	6.3
metronidazole	4.2	ceftriaxone	4.4	tobramycin	7.2	cloxacillin	4.1	piperacillin/inhib.	7.0	flucloxacillin	3.3	piperacillin/inhib.	4.3	clindamycin	4.2
flucloxacillin	2.8	cefotaxime	3.6	amikacin	6.9	cefotaxime	3.7	vancomycin	6.7			meropenem	4.3
amoxicillin/inhib.	2.4	netilmicin	2.8	meropenem	4.2			metronidazole	2.9			procaine penicillin	2.9
meropenem	2.4	teicoplanin	2.8	ampicillin/inhib.	3.9			benzylpenicillin	2.4			cloxacillin	2.9
amikacin	1.2	ceftazidime	1.6	cefuroxime	3.3			cloxacillin	1.9
				piperacillin/inhib.	3.0			ceftazidime	1.4
				amoxicillin/inhib.	1.8
				imipenem	1.8

inhib., enzyme inhibitor.

Grey shading indicates drug utilization 75% (DU75%) by UN region.

^aNumber of countries included from each region: Northern Europe (6), Southern Europe (9), Western Europe (6), Africa (4), Asia (8), Australia (1), Latin America (3), North America (1).

Antibiotic prescribing by indication

The most frequently recorded reason to treat children was a bacterial lower respiratory tract infection (18.7%) and that to treat neonates was sepsis (36.4%) (see Table 5 for the top 10 reasons for antibiotic treatment in children and neonates).

Table 5.

Top 10 most-recorded reasons to treat children (>1 month) and neonates (<30 days)

Reason to treat (children)	%	Reason to treat (neonates)	%
Bacterial lower respiratory tract infection	18.7	sepsis	36.4
Prophylaxis for medical problems	15.1	prophylaxis for maternal risk factors	12.2
Prophylaxis for surgical disease	9.9	prophylaxis for newborn risk factors	11.3
Sepsis	9.0	lower respiratory tract infection	8.7
Treatment for surgical disease	6.1	prophylaxis for surgical disease	5.4
Urinary tract infection (upper and lower)	5.6	prophylaxis for medical problems	5.1
Febrile neutropenia/fever in oncologic patients	4.8	catheter-related bloodstream infection	3.4
Upper respiratory tract infection	4.6	CNS infections	3.2
Skin/soft tissue infections	4.4	treatment for surgical disease	2.6
Viral lower respiratory tract infection	3.7	skin/soft tissue infections	2.6

Tables 6 and 7 compare therapeutic (community-acquired versus hospital-acquired; empirical versus targeted treatment) and prophylactic prescribing in children and neonates: 75.3% of antibiotics were prescribed therapeutically (75.4% in children and 74.9% in neonates) and 24.7% prophylactically (14.8% for medical and 9.8% for surgical prophylaxis in children and 19.5% for medical and 5.6% for surgical prophylaxis in neonates).

Table 6.

Antibiotic use by indication, type of treatment (empirical versus targeted) and UN region for children >1 month

Region	Therapeutic use
	community-acquired infection				hospital-acquired infection				Prophylactic use
	empirical		targeted		empirical		targeted		medical		surgical
	n	%	n	%	n	%	n	%	n	%	n	%
Europe	2073	81.8	461	18.2	676	69.9	291	30.1	768	63.3	445	36.7
Africa	343	81.7	77	18.3	54	76.1	17	23.9	27	96.4	1	3.6
Asia	411	72.4	157	27.6	138	60.8	89	39.2	124	42.9	165	57.1
Australia	201	73.1	74	26.9	54	56.8	41	43.2	79	56.4	61	43.6
Latin America	167	73.9	59	26.1	149	69.0	67	31.0	42	48.3	45	51.7
North America	149	70.6	62	29.4	66	68.0	31	32.0	119	71.7	47	28.3
Total	3344	79.0	890	21.0	1137	68.0	536	32.0	1159	60.3	764	39.7

Sixty-eight antibiotics were recorded with unknown indication.

Table 7.

Antibiotic use by indication, type of treatment (empirical versus targeted) and UN region for neonates <30 days

Region	Therapeutic use
	community-acquired infection				hospital-acquired infection				Prophylactic use
	empirical		targeted		empirical		targeted		medical		surgical
	n	%	n	%	n	%	n	%	n	%	n	%
Europe	665	91.7	60	8.3	275	77.2	81	22.8	236	76.1	74	23.9
Africa	148	72.2	57	27.8	26	76.5	8	23.5	2	28.6	5	71.4
Asia	108	92.3	9	7.7	103	73.6	37	26.4	122	80.3	30	19.7
Australia	44	100	—	—	23	71.9	9	28.1	41	89.1	5	10.9
Latin America	6	75.0	2	25.0	10	58.8	7	41.2	37	86.0	6	14.0
North America	14	93.3	1	6.7	17	94.4	1	5.6	8	53.3	7	46.7
Total	985	88.4	129	11.6	454	76.0	143	24.0	446	77.8	127	22.2

Fourteen antibiotics were recorded with unknown indication.

Therapeutic prescribing

Overall, hospital-acquired infections (HAIs) accounted for 49.9%, 34.9%, 33.7%, 28.9%, 28.5% and 14.4% of all therapeutic antibiotic prescriptions in Latin America, Asia, North America, Europe, Australia and Africa, respectively. Of all antibiotics administered for an HAI, 29.9% were recorded as being targeted based on microbiological results (range, from 23.8% in Africa to 39.4% in Australia). Overall, antibiotics were more commonly prescribed for HAIs in neonates (34.9%; range, from 14.2% in Africa to 68.0% in Latin America) compared with children (28.3%; range, from 14.5% in Africa to 48.9% in Latin America). Glycopeptides (mainly vancomycin) were most often prescribed for HAIs in Latin America, North America, Australia and Europe (26.2%, 21.7%, 17.3% and 20.4% of total antibiotic use for HAIs, respectively), followed by carbapenems (mainly meropenem) in Latin America, Africa, Asia and Europe (24.0%, 21.9%, 16.1% and 13.3%, respectively). Use of cefepime for HAIs was 15.7% in North America, but <2% elsewhere.

Of all antibiotics administered for community-acquired infections (CAIs), 19.1% were targeted (range, from 16.0% in Europe to 27.9% in North America). The most commonly prescribed antibiotics for CAIs were third-generation cephalosporins (mainly ceftriaxone) in Asia, Latin America and Europe (32.8%, 23.5% and 18.7% of total antibiotic use for treatment of CAIs, respectively) followed by aminoglycosides (mainly gentamicin) in Africa, Australia and Europe (26.6%, 17.6% and 14.6%, respectively), broad-spectrum penicillins (mainly amoxicillin) in North America, Europe and Australia (14.5%), combinations of penicillins with an enzyme inhibitor (mainly co-amoxiclav) in Europe, North America and Australia (14.2%, 11.9% and 10.0%, respectively), narrow-spectrum penicillins (mainly benzylpenicillin) in Africa (16.5%), and clindamycin in Latin and North America (12.0% and 10.2%, respectively).

Prophylactic prescribing

Medical prophylactic prescribing accounted for 64.3% of all prophylactic antibiotic prescribing, ranging from 55.8% in Asia to 82.9% in Africa. Overall, antibiotics were more commonly prescribed for medical prophylaxis in neonates (77.8%; range, from 28.6% in Africa to 89.1% in Australia) compared with children (60.3%; range, from 42.9% in Asia to 96.4% in Africa) (Tables 6 and 7). Many different antibiotics were prescribed for medical prophylactic use, with sulfamethoxazole/trimethoprim dominating in Africa, North America, Australia and Europe (79.3%, 36.2%, 34.2% and 30.5% of total medical prophylactic prescribing, respectively) and ampicillin in Latin America and Asia (21.5% and 15.9%, respectively).

Surgical prophylactic prescribing accounted for 39.7% of all prophylactic antibiotic prescribing in children, compared with 22.2% in neonates (Tables 6 and 7). Most antibiotics for surgical prophylactic use were first-generation cephalosporins, mainly cefazolin in North America and Australia (61.1% and 47.0% of total surgical prophylactic prescribing, respectively) and cefalotin in Latin America (27.5%), followed by third-generation cephalosporins (mainly ceftriaxone) in Asia and Europe (25.6% and 14.6%, respectively), second-generation cephalosporins (mainly cefuroxime) (17.5%) or combinations of penicillins with enzyme inhibitors (mainly co-amoxiclav) in Europe (16.6%). Prolonged surgical prophylaxis (>1 day) was very common in all regions, ranging from 78% in Europe to 84% in Latin America.

Discussion

This global ARPEC-PPS for the first known time explored the feasibility of producing qualitative indicators to uniformly assess antimicrobial prescribing, to identify key areas of poor practice and to propose benchmarks for improved antibiotic prescribing among hospitalized neonates and children worldwide.

Overall, we found less regional difference in antibiotic prescribing among hospitalized neonates than among children. The dominance of gentamicin used in combination with benzylpenicillin or ampicillin across all regions was very striking, and explains the high proportion of antibiotics used in combination among neonates (around 70%). Remarkably high use of amikacin was noted in neonates admitted to Western European, Southern European, Asian and Latin American hospitals, and, worryingly, meropenem was widely prescribed to Asian neonates.

In contrast, we observed striking regional differences in antibiotic prescribing among hospitalized children. A high proportion of African, Australian, Western European and Northern European children continued to receive older narrow-spectrum antibiotics, mainly benzylpenicillin, sulfamethoxazole/trimethoprim, amoxicillin and gentamicin. Africa was characterized by considerably high use of gentamicin, often in combination with benzylpenicillin. Eastern European, Southern European, Asian, North American and Latin American children received considerably more broad-spectrum antibiotics, mainly third-generation cephalosporins, cefepime and meropenem.

The prevalence of broad-spectrum agent use is an important quality indicator in children. The high use seen here may be partially explained by the remarkably high antibiotic prescribing rates for HAIs as an indication, e.g. in Latin America (49.9%), Asia (34.9%) and North America (33.7%), again with high proportions of meropenem, ceftriaxone and cefepime use. The striking finding of the high number of antibiotics prescribed for HAIs in neonates, reaching 68% of all therapeutic prescribing in Latin America, requires further investigation. In Europe we observed slightly lower prevalence rates of antimicrobials prescribed for an HAI (28.9%), comparable to the ESAC 2009-PPS (adults and children, 30.7%)⁸ and the ECDC 2010-PPS (paediatric wards, 30.3%).²⁰

High proportions of broad-spectrum antibiotic use could be explained by regionally high rates of ESBL-producing or carbapenem-resistant Gram-negative organisms. Surveillance programmes in Latin America in adults and children have demonstrated an increasing trend of resistance to extended-spectrum cephalosporins with high prevalence of ESBL-producing Escherichia coli and Klebsiella spp. and carbapenem-resistant Klebsiella spp.^21,22 Similarly, in Asia the burden of antimicrobial resistance is now very high.²³ It is unclear whether, in the surveyed children, this high level of use of broad-spectrum antibiotics was indeed justified by local resistance patterns, due to the absence of linked local neonatal and paediatric antimicrobial resistance data. However, the high level of empirical antibiotic use may indicate that at least a proportion of this prescribing may be inappropriate. Such inappropriate use was found by Levy et al.²⁴ in PICUs and paediatric wards, which the authors attributed to failure to discontinue or de-escalate therapy. Even in resource-limited settings, where de-escalation may be less commonly considered and where bacteriological cultures are much less frequently performed, successful de-escalation of carbapenems has been reported.²⁵ Therefore, to limit microbial selection pressure, de-escalation upon culture results should be implemented whenever possible.²⁶

We identified several other potential indicators to assess antimicrobial prescribing that could be used to set benchmarks for quality improvement of antibiotic use in children and neonates. The simple prevalence of antibiotic use by ward, by hospital or at national level could be used as a quality indicator if prescription rates are much higher than in other comparators as part of a benchmarking programme. The worldwide prevalence rate of antimicrobial prescribing among children's wards was 36.7%, with high rates observed in PICUs (61%) and on surgical wards (35%), which were twice as high compared with the ESAC 2009-PPS (PICUs, 30%; child surgical wards, 16%).⁹ Setting crude prevalence quantitative targets at the hospital level is very difficult because antibiotic use depends on many variables, including patient case-mix, type of hospital, proportion of HAIs and prevalence of resistance. For instance, 80% of the centres taking part in this ARPEC-PPS were tertiary care hospitals, with several specialized stand-alone paediatric hospitals. Therefore, we collected data from a high number of children receiving specialized care admitted to, e.g. haematology/oncology and transplant wards as well as very low birth weight neonates who were admitted to level 3 NICUs. This could explain the high overall antibiotic prescribing rates. There is a need to further develop the data collection with more hospitals, reducing the selection bias and stratifying for hospital type and case-mix to develop more appropriate benchmarking standards.

The early switch from parenteral to oral therapy is another quality indicator, due to its many advantages, such as decreased risk of catheter-related infections, reduced costs and the possibility of early discharge from hospital.^27,28 However, it is not known to what extent different antibiotic administration routes have an impact on antimicrobial resistance.²⁹ Parenteral administration among children was more frequent in Asia, Latin America and Europe (88%, 81% and 67%, respectively) compared with adults in Europe in the ESAC-PPS (60%).⁸ The concept of early switching from parenteral to oral therapy seems uncommon among children²⁸ and could be explained by more limited options for oral broad-spectrum antibacterial equivalents with appropriate formulations, class and potency and the challenges of oral administration of medications in young children in general.³⁰ A parenteral-to-oral switch-over programme in line with available guidance can, however, be introduced into clinical practice²⁸ and should be supported by collecting information for particular indications in children.³¹

A further key quality indicator is documentation of the reason for prescription in the medical notes of the patient, which was slightly less in European children (76%) than in European adults (80%),⁸ but considerably lower in Latin America (52%). Good documentation of the indication in the patient chart ensures communication of diagnosis and treatment among clinicians, pharmacists and other healthcare providers, and allows subsequent prescription review and interventions such as de-escalation and stopping of antimicrobial treatment. This ESAC-PPS indicator was used to set a benchmark of >95% in Scottish hospitals.¹¹

Another potential quality indicator is the administration of antibiotics for surgical prophylaxis. We proposed an international benchmark of 100% for surgical prophylaxis prescriptions being administered <24 h,⁷ because the duration of surgical prophylaxis should not exceed a 24 h perioperative period unless exceptional cases.³² Therefore, the excessively lengthy surgical prophylaxis for >24 h observed in many hospitals participating in this survey (up to 84% in Latin America hospitals) is not acceptable. Moreover, in these hospitals we observed high broad-spectrum surgical prophylactic prescribing, mainly ceftriaxone, which could be explained by the concern about increasing resistance to first- and second-generation cephalosporins among Gram-negative isolates. In other parts of the world, mostly first- and second-generation cephalosporins were administered. Equally concerning is the high proportion of antibiotics prescribed for medical prophylaxis in all regions except Africa, including high levels of neonatal medical prophylaxis. Indeed, medical prophylaxis was the second most common reason for antibiotic treatment in children, accounting for 15% of all prescriptions. Whether this is an area for quality improvement needs to be further explored with more detailed evaluation of patterns of prophylactic antibiotic use and their relation to available evidence.

Study strengths and limitations

The strength of our study lies in the uniformity of data collection, the simplicity of the protocol and data collection templates (less time consuming); and the assurance of data quality (data completeness and validation process) as described in Versporten et al.¹³ The limitations of this study are inherent to the epidemiological method of a cross-sectional survey.¹³ Additionally, as mentioned above, the recruitment of hospitals, mainly done through existing paediatric or other networks, involving a high number of highly specialized or referral hospitals, will have induced selection bias and may partially explain the high antimicrobial prevalence rates observed. This indicator is strongly associated with patient characteristics. However, it does not necessarily explain some worrying findings with respect to quality indicators, such as the low frequency of documenting the reason for treatment in the medical files or prolonged antibiotic prescribing for surgical prophylaxis. These quality indicators should be met independently of institutional characteristics. Finally, we do not suspect social desirability bias. We dealt with highly motivated infectious disease specialists or other, related specialists who often perform a supervisory function at hospital level. We also clearly communicated in different ways that they needed to complete the forms with information, which was written down in medical or other files, without discussing the appropriateness of the prescribed antimicrobial.

The 2015 WHO global action plan on antimicrobial resistance proposes collection and reporting of data on the use of antimicrobial agents so that trends can be monitored and the impact of action plans assessed, but does not specifically mention the difficulties of data collection and analysis in children.³³ Our ARPEC-PSS tool constituted a simple method to collect antimicrobial prescribing data electronically. The uniformity of data collection together with the implemented online quality assurance improves the validity of the data we collected worldwide. This is of great value to public health nationally and globally as these methods can be implemented and repeated regionally in the future.³⁴ It can also provide meaningful educational feedback to prescribers, which could have a significant effect on prescribing practices. We have now identified several measurable quality indicators that could be used to set benchmarks for antibiotic prescribing in neonates and children. The next step is to implement feasible quality targets through repeated PPSs (e.g. quarterly) on a sample of patients or wards within the same hospitals. This will allow the future prospective assessment of intervention plans aimed at improving the quality of antibiotic prescribing.

Funding

This work was supported 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)’ (http://ec.europa.eu/eahc/). The October–November 2012 Antibiotic Resistance and Prescribing in European Children Point Prevalence Survey was co-funded by the Paediatric European Network for Treatment of AIDS (PENTA) (http://www.penta-id.org/).

Transparency declarations

None to declare.

Acknowledgements

We would like to thank all colleagues who contributed to the success of this project: Graciela Maria Calle, Hospital de Pediatria Juan P. Garrahan, Buenos Aires, Argentina; Julia Clark, Royal Children's Hospital, Brisbane, Australia; Celia Cooper, Women's and Children's Hospital, North Adelaide, Australia; Christopher C. Blyth and Joshua Reginald Francis, Princess Margaret Hospital for Children, Perth, Australia; Jameela Alsalman, Salmaniya Medical Complex, Manama, Bahrain; Hilde Jansens and Ludo Mahieu, University Hospital Antwerp, Antwerp, Belgium; Paul Van Rossom, General Hospital Klina, Brasschaat, Belgium; Wouter Vandewal, AZ Sint Lucas, Brugge, Belgium; Philippe Lepage and Sophie Blumental, Hôpital Universitaire des Enfants Reine Fabiola, Brussels, Belgium; Caroline Briquet, Université Catholique de Louvain, Cliniques Universitaires Saint-Luc, Brussels, Belgium; Dirk Robbrecht, General Hospital A. Z. Damiaan, Ostend, Belgium; Pierre Maton, NICU-CHC St Vincent, Rocourt, Belgium; Patrick Gabriels, Sint-Trudo Ziekenhuis, Sint-Truiden, Belgium; Zana Rubic and Tanja Kovacevic, University Hospital Centre Split, Croatia; Jens Peter Nielsen, Pediatric Department, Regionshospital Viborg, Viborg, Denmark; Jes Reinholdt Petersen, Department of Neonatology, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark; Porntiva Poorisrisak, Department of Pediatrics, Naestved Hospital, Naestved, Denmark; Lise Heilmann Jensen, Roskilde University Hospital, Roskilde, Denmark; Mari Laan, Tallinn Children's Hospital, Tallinn, Estonia; Eda Tamm, Children's Clinic of Tartu University Hospital, Tartu, Estonia; Maire Matsinen and Maija-Liisa Rummukainen, Central Finland Health Care District, Jyväskylä, Finland; Vincent Gajdos and Romain Olivier, Hopitaux Universitaires Paris Sud, Antoine Beclere, Clamart, France; Flore Le Maréchal, Department of Pediatrics and Neonatology, University Hospital, Dijon, France; Alain Martinot, François Dubos, Marion Lagrée, CHRU Lille, Lille University Hospital, Lille, France; Sonia Prot-Labarthe and Mathie Lorrot, AP-HP Hôpital Robert-Debré, Paris, France; Daniel Orbach, Institut Curie, Paris, France; Karaman Pagava, Tbilisi State Medical University, Tbilisi, Georgia; Markus Hufnagel, Division of Pediatric Infectious Diseases and Rheumatology, Center of Pediatrics and Adolescent Medicine, University Medical Center, Freiburg, Germany; Markus Knuf, Children's Hospital, Dr Horst-Schmidt-Kliniken, Wiesbaden, Germany; Stephanie A. A. Schlag and Johannes Liese, University Children's Hospital, University of Würzburg, Würzburg, Germany; Lorna Renner, Prof. Korle Bu Teaching Hospital, Accra, Ghana; Anthony Enimil and Marah Awunyo, Komfo Anokye Teaching Hospital, Kumasi, Ghana; Garyfallia Syridou and Nikos Spyridis, Aglaia Kyriakou Children's Hospital, Athens, Greece; Elena Critselis, First University Department of Pediatrics, ‘Aghia Sophia’ Children's Hospital, Athens, Greece; Sofia Kouni and Katerina Mougkou, University of Athens Collaborative Center for Clinical Epidemiology and Outcome Research, Athens, Greece; Fani Ladomenou, University Hospital of Heraklion, Crete, Greece; Despoina Gkentzi, University General Hospital of Patras, Patras, Greece; Elias Iosifidis and Emmanuel Roilides, 3rd Department of Pediatrics, Faculty of Medicine, Aristotle University School of Health Sciences, Hippokration Hospital, Thessaloniki, Greece; Suneeta Sahu, Microbiology, Apollo Hospitals, Bhubaneswar, Odisha, India; Srinivas Murki, Fernandez Hospital, Hyderabad, India; Manoj Malviya and Durga Bhavani Kalavalapalli, Nice Hospital for Women, Newborns and Children, Hyderabad, India; Sanjeev Singh, Amrita Institute of Medical Sciences, Kochi, India; Tanu Singhal, Kokilaben Dhirubhai Ambani Hospital and Medical Research Institute, Mumbai, India; Garima Garg, Escorts Heart Institute, New Delhi, India; Pankaj Garg, Sir Ganga Ram Hospital, Rajinder Nagar, New Delhi, India; Neelam Kler, Sir Ganga Ram Hospital, Rajinder Nagar, New Delhi, India; Jafar Soltani, Besat Tertiary Hospital affiliated to Kurdistan University of Medical Sciences, Sanandaj, Kurdistan, Iran; Zahra Jafarpour and Gholamreza Pouladfar, Professor Alborzi Clinical Microbiology Research Center, Shiraz University of Medical Sciences, Shiraz, Iran; Giangiacomo Nicolini, San Martino Hospital, Pediatric Department, Belluno, Italy; Carlotta Montagnani and Luisa Galli, Department of Health Sciences, University of Florence, Meyer Children's University Hospital, Florence, Italy; Susanna Esposito, Pediatric Highly Intensive Care Unit, and Rossana Tenconi, Department of Pathophysiology and Transplantation, Università degli Studi di Milano, Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Milan, Italy; Andrea, Lo Vecchio, University of Naples Federico II, Naples, Italy; Daniele Dona' and Carlo Giaquinto, Department for Woman and Child Health, Padua, Italy; Eleonora Borgia, University Hospital of Padova, Padua, Italy; Patrizia D'Argenio and Maia De Luca, Bambino Gesù Children's Hospital, Rome, Italy; Chiara Centenari, Paediatric and Neonatal Unit, Hospital of Viareggio, Viareggio, Italy; Lul Raka, National Institute of Public Health of Kosovo and University of Prishtina, and Denis Raka, University of Prishtina, Prishtina, Kosovo; Abeer Omar and Haifaa Al-Mousa, Directorate of Infection Control, Ministry of Health, Kuwait City, Kuwait; Dzintars Mozgis, Centre for Diseases Prevention and Control, Riga, Latvia; Inese Sviestina, University Children's Hospital, Riga, Latvia; Sigita Burokiene, Children's Hospital, Affiliate of Vilnius University Hospital Santariskiu Klinikos, Vilnius, Lithuania; Vytautas Usonis, Vilnius University, Vilnius, Lithuania; Gabriela Tavchioska, General Hospital ‘Borka Taleski’, Prilep, Macedonia; Antonia Hargadon-Lowe, Queen Elizabeth Central Hospital, Blantyre, Malawi; Peter Zarb and Michael A. Borg, Mater Dei Hospital, Msida, Malta; Carlos Agustín González Lozano and Patricia Zárate Castañon, Instituto Nacional de Pediatría, México D.F., México; Martha E. Cancino, Universidad Autónoma de Nayarit, México and DURG-LA (Grupo para la Investigación de la Utilización de los Medicamentos-América Latina), Tepic, Nayarit, México; Bernadette McCullagh, South Eastern Health and Social Care Trust, Belfast, Northern Ireland; Ann McCorry, Southern Health and Social Care Trust, Craigavon, Northern Ireland; Cairine Gormley, Western Health and Social Care Trust, Derry, Northern Ireland; Zaina Al Maskari and Amina Al-Jardani, Royal Hospital, Muscat, Oman; Magdalena Pluta, Department of Children's Infectious Diseases, Medical University of Warsaw, Poland, Warsaw, Poland; Fernanda Rodrigues and Ana Brett, Hospital Pediátrico, Centro Hospitalar e Universitário de Coimbra, Coimbra, Portugal; Isabel Esteves, Hospital de Santa Maria, Pediatric Department, Pediatric Infectious Diseases Unit, Lisbon, Portugal; Laura Marques, Centro Hospitalar do Porto, Porto, Portugal; Jameela Ali AlAjmi, Hamad Medical Corporation, Doha, Qatar; Simona Claudia Cambrea, Faculty of Medicine, ‘Ovidius’ University, Constanta, Romania; Asia N. Rashed, King Abdulaziz Medical City—Jeddah, Jeddah, Saudi Arabia; Aeshah Abdu Mubarak Al Azmi, Ministry of National Guard Health Affairs, King Abdulaziz Medical City, Pharmaceutical Care Services Department, Jeddah, Saudi Arabia; Si Min Chan, Khoo Teck Puat-National University Children's Medical Institute, National University Health System, Singapore, Singapore; Mas Suhaila Isa, National University Hospital, Singapore; Peter Najdenov, General Hospital Jesenice, Jesenice, Slovenia; Milan Čižman, University Medical Centre, Ljubljana, Slovenia; Sibila Unuk, University Medical Center Maribor, Maribor, Slovenia; Heather Finlayson, Department of Paediatrics and Child Health, Stellenbosch University, Cape Town, South Africa; Angela Dramowski, Tygerberg Hospital, Cape Town, South Africa; Irene Maté-Cano, Hospital Universitario del Henares, Coslada, Madrid, Spain; Beatriz Soto, Hospital Universitario de Getafe, Getafe, Madrid, Spain; Cristina Calvo, Severo Ochoa Hospital, Leganés, Madrid, Spain; Begoña Santiago and Jesus Saavedra-Lozano and Amaya Bustinza, Gregorio Marañon Hospital, Madrid, Spain; Luis Escosa-García, Hospital Infantil Universitario La Paz, Madrid, Spain; Noelia Ureta, Elisa Lopez-Varela and Pablo Rojo, Hospital Universitario 12 de Octubre, Madrid, Spain; Alfredo Tagarro, Hospital Universitario Infanta Sofia, San Sebastian de los Reyes, Madrid, Spain; Pedro Terol Barrero, Hospital Universitario Virgen Macarena, Seville, Spain; Elena Maria Rincon-Lopez, Hospital Universitario y Politecnico La Fe, Valencia, Spain; Ismaela Abubakar, Edward Francis Small Teaching Hospital/Royal Victoria Teaching Hospital, Banjul, The Gambia; Jeff Aston, Mitul Patel and Alison Bedford Russell, Birmingham Children's Hospital NHS Foundation Trust, Birmingham, UK; Maggie Heginbothom, Public Health Wales, Cardiff, UK; Prakash Satodia, University Hospitals Coventry and Warwickshire NHS Trust, Coventry, UK; Mehdi Garbash, County Durham and Darlington NHS Foundation Trust, Durham and Darlington, UK; Alison Johnson, Wye Valley NHS Trust, Hereford, UK; David Sharpe, Alder Hey Children's NHS Foundation Trust, Liverpool, UK; Christopher Barton, Institute of Child Health, University of Liverpool, Liverpool, UK; Esse Menson and Sara Arenas-Lopez, Evelina London Children's Hospital, London, UK; Suzanne Luck, Kingston Hospital NHS Foundation Trust, London, UK; Katja Doerholt, St George's Hospital, London, UK; Paddy McMaster, North Manchester General Hospital, Manchester, UK; Neil A. Caldwell, Wirral University Teaching Hospital NHS Foundation Trust, Merseyside, UK; Andrew Lunn, Nottingham Children's Hospital, Nottingham, UK; Simon B. Drysdale, Oxford University Hospitals NHS Trust, Oxford, UK; Rachel Howe, Peterborough City Hospital, Peterborough, UK; Tim Scorrer and Florian Gahleitner, Queen Alexandra Hospital, Portsmouth, UK; Richa Gupta, Royal Preston Hospital, Lancashire Teaching Hospitals, Preston, UK; Clare Nash, Sheffield Children's NHS Foundation Trust, Sheffield, UK; John Alexander, University Hospital of North Midlands, Stoke on Trent, UK; Mala Raman, Torbay Hospital, South Devon Health Care NHS Foundation Trust Torquay, UK; Emily Bell, Royal Cornwall Hospitals NHS Trust, Truro, Cornwall, UK; Veena Rajagopal, St George's NHS Trust, London, UK; Stephan Kohlhoff, SUNY Downstate Medical Center, Brooklyn, USA; Elaine Cox, Kristen Nichols and Elaine Cox, Riley Hospital for Children at IU Health, Indianapolis, USA; Theoklis Zaoutis, The Children's Hospital of Philadelphia, Philadelphia, USA.

Author notes

†

Members are listed in the Acknowledgements section.