Excessive and inappropriate use of antibiotics is the most important contributor to the development of antimicrobial resistance in bacteria. In recent years, several strategies have been developed to confront both of these problems.1–3 In Europe, the European Surveillance of Antimicrobial Consumption Network (ESAC-Net) has been monitoring antibiotic use at the hospital and community levels in 31 European countries since 1997.2 The data evinces clear north-south and west-east gradients, with greater consumption and prevalence of antimicrobial resistance in countries in Southern and Eastern Europe.2,4 At the domestic level in Spain, in 2014 the Interterritorial Council of the National Health System approved the Plan Nacional de Resistencia a Antibióticos (National Plan on Antibiotic Resistance, PRAN) for surveillance of antibiotic use overall and of bacterial antimicrobial resistance in Spain, with its objectives including observing, limiting and improving antibiotic use in both animal and human health, and to promote and support antibiotic use and antimicrobial resistance surveillance networks.3

More than 90% of the antibiotics consumed are prescribed at the outpatient level, mainly for management of respiratory tract infections. It is estimated that nearly half of antibiotic prescriptions in primary care are inappropriate, and unnecessary prescription is the main reason antibiotic use is deemed inappropriate.2,5 In 2018, Spain was the fourth leading country in antimicrobial consumption in the community in Europe.2

Antimicrobial consumption is greater at both ends of the lifespan. It is estimated that approximately 60% of children under 4 years are exposed to antimicrobials at least once a year.6,7 Yet, there are few population-based studies in Spain describing antimicrobial consumption in the paediatric population at the community level.8–10 Antimicrobial overuse in the paediatric population has significant repercussions, as the prevalence of antimicrobial resistance is up to 30% greater in isolates in community-acquired infections compared to isolates in the hospital setting.11,12

The aim of our study was to describe the use of antibiotics at the outpatient level in the paediatric population of Asturias, Spain, and analyse quantitative and qualitative temporal trends in antimicrobial use in the 2005–2018 period.

The mean consumption of systemic antibacterials (group J01 in the ATC classification) dispensed by prescription in the community by the paediatric population of Asturias between years 2005 and 2018 was of 14 DIDs (95% CI, 13.4–14.6). The maximum difference in consumption found between time points (Fig. 1) was of 3.4 DIDs (maximum of 15.3 DIDs in 2014; minimum of 11.9 DIDs in 2018), and we found no significant differences in the comparison by periods (P = .6) (Table 1).

Figure 1.

Trends in antibiotic consumption (DID) in the paediatric population of Asturias (2005–2018).

(0.21MB).

The distribution of the different therapeutic groups by decreasing frequency was: J01C (β-lactam penicillins), 10.7 DIDs (76.7% of total consumed); J01F (macrolides, lincosamides and streptogramins), 1.8 DIDs (12.8%); J01D (other β-lactam antibacterials, which refers to cephalosporins), 1.2 DIDs (8.9%); J01A (tetracyclines), 35.3 DDD/1000/y (0.7%); J01E (sulphonamides and trimethoprim), 27.5 DDD/1000/y (0.5%); J01X (other antibacterials), 12.2 DDD/1000/y (0.2%); J01M (quinolones), 5.2 DDD/1000/y (0.1%). The most frequently consumed chemical substances were amoxicillin (J01CA04; 5.2 DID), amoxicillin-clavulanic acid (J01CR02; 5 DID), azithromycin (J01FA10; 1 DID), cefuroxime (J01DC02; 0.8 DID) and clarithromycin (J01FA09; 0.6 DID) (Table 2).

Amoxicillin and amoxicillin-clavulanic acid accounted for 94.9% of consumption in the J01C group and 72.8% of overall consumption. In the first period, the consumption of amoxicillin-clavulanic acid was greater compared to the consumption of amoxicillin, a trend that reversed from 2011 (Fig. 2). There was a significant increase in the consumption of amoxicillin between the first and last periods under study (P = .027), while the decrease in the use of e amoxicillin-clavulanic acid was not statistically significant (P = .399) (Table 2).

Figure 2.

Trends in the consumption of amoxicillin and amoxicillin-clavulanic acid (DID).

(0.26MB).

Consumption of antibacterials in the J01 F group remained stable throughout the study period (P = .739) (Table 1). Azithromycin and clarithromycin accounted for 91.6% of the consumed antibacterials belonging to group J01F. We found an increase in the use of azithromycin (P < .001) and a decrease in the use of clarithromycin (P = .001) (Table 2). From 2009, the consumption of azithromycin exceeded the consumption of clarithromycin, with the difference increasing through the end of the study period (differences of 0.1 DID in 2009 and 1.1 DID in 2016), as can be seen in Fig. 3.

Figure 3.

Trends in the consumption of J01F antibacterials, azithromycin and clarithromycin (DID).

(0.23MB).

In the J01D group, cefuroxime and cefixime accounted for 91.9% of the consumption. The frequency of consumption of J01D group antibacterials, cefuroxime and cefixime all decreased in the period under study (P < .001, P < .001 and P = .029, respectively) (Tables 1 and 2). Consumption of drugs in this therapeutic group reached the minimum in 2015 (0.5 DID) (Fig. 4). Consumption of antibacterials in all other groups amounted to 1.5% of the total. The use of trimethoprim-sulfamethoxazole remained stable through time (P = .591), while there were significant increases in the consumption of doxycycline (P = .019), fosfomycin (group J01XX; P < .001) and ciprofloxacin (P = .001). The consumption of quinolones was 3 times greater in 2018 compared to the first year under study (Fig. 5).

Figure 4.

Trends in the consumption of J01D antibacterials, cefuroxime and cefixime (DID).

(0.21MB).

Figure 5.

Trends in the consumption of J01M antibacterials and their active ingredients (DDD/1000/y).

(0.23MB).

The J01_B/N quality indicator in our study was below the ratios corresponding to the total population of Spain and also improved over time, with ratios around 10 in the last few years (Fig. 6).

Figure 6.

Trends in the J01_B/N indicator in Spain and in the paediatric population of Asturias.

(0.24MB).

The DDD is the standard recommended by the WHO to perform drug utilization studies.14 However, one of its main drawbacks is that it may underestimate consumption in the paediatric population and may not accurately reflect the doses used or recommended in everyday clinical practice. Although other parameters have been studied for the purpose of assessing antibiotic consumption in paediatric inpatients, such as the DDD adjusted for weight or days of treatment, the ideal measure of antibiotic consumption in the paediatric population has yet to be identified. Some of these indicators also require information that is usually not accessible in drug utilization studies that analyse consumption in the community and in large populations.16–18

The WHO updates DDDs regularly according to changes in dosage for the main indications for each drug. The latest update took place in 2019 and included adjustments in the DDDs of the most frequently consumed antibiotics in Spain, amoxicillin and amoxicillin-clavulanic acid. The data of this study were updated based on this latest review and can thus be compared, with the drawbacks that we already mentioned, with the data published by the ESAC-Net. The consumption in the paediatric population of Asturias increased progressively in the first time interval, from 12 DID in 2005 to 15.2 DID in 2009, and decreased from 2015, reaching levels similar to those found at the beginning of the study. Every year, consumption in Asturias was lower compared to consumption in the general population of Spain with a relatively parallel trend until 2015. In recent years, the evidence shows an excessive use of antibacterials at the outpatient level in our country that places Spain among the countries in Europe with the greatest consumption.2,3 Studies conducted in different regions in Spain have also reported a substantial exposure of the population to antibiotics and an increase in the use of broad-spectrum antibiotics through the years.6,7,19–22

Five antibiotics accounted for 90.5% of the total consumption, and amoxicillin and amoxicillin-clavulanic acid, in particular, accounted for 75% of the total. In the adult population, there has been an increase in the use of amoxicillin-clavulanic acid, which has exceeded the consumption of amoxicillin since the early 2000s, reaching 56% of total outpatient prescriptions in 2018.2,6,7,19–21,23,24 Frequent prescription of amoxicillin-clavulanic acid in the paediatric population is not justified given the infections usually managed at the outpatient level, as most are caused by bacteria that do not produce β-lactamases, such as Streptococcus pyogenes or Streptococcus pneumoniae. Vázquez et al. described a predominant use of amoxicillin over its combination with clavulanic acid since 2005 in the paediatric population of Castilla y Leon.10 We identified this shift in the paediatric population of Asturias since 2011, and the gap widened over time, which reflects an improved pattern of consumption. Although the consumption of amoxicillin-clavulanic acid in the population under study did not change significantly between periods, we ought to highlight a decrease of 26% in the use of this active ingredient in the last period under study.

Consumption of drugs in the macrolides therapeutic group remained stable in our population. However, we ought to warn of a sharp increase in the consumption of azithromycin over the study period, which is all the more remarkable given that fewer DDDs are required per course of treatment with this drug. The main consequence of the excessive use of any antibiotic is the development of antimicrobial resistance, but the potential side effects of the administration of any drug must also be considered, and in this case it is important to be aware of the recent report on the risk of severe arrhythmia associated with the use of azithromycin in children.25 In the United States, there is evidence that azithromycin is prescribed too frequently and inappropriately in the paediatric population (second most frequently prescribed antibiotic), a practice that is currently the main target of antibiotic stewardship programmes at the outpatient level in the country.26

The decrease in the consumption of cephalosporins in the paediatric population of Asturias, especially steep in 2015, could be explained by a shortage of cephalosporins at the time, a situation that prompted an alert by the Committee on Medicines of the Asociación Española de Pediatría (Spanish Association of Paediatrics) due to the short supply of oral solutions in a widely used therapeutic group with special indications in young children.6,7,27,28

Quinolones are the third most widely used therapeutic group at the outpatient level in the general Spanish population (11% of total consumption) and the most frequently prescribed group in the population over 75 years.2,6,7 Although the consumption of quinolones in our study was very small (0.1% of the total), there was a significant increase over the period under study that should be monitored in future studies.

The ESAC group proposed the J01_B/N as a quality indicator for antibacterial use at the outpatient level. This ratio allows comparison between European countries, which has evinced significant variation, with the lowest ratios found in Norway (0.16) and Sweden (0.21) and the highest in Greece (624) and Italy (226) in 2018.2 Malo-Fumanal et al. found significant quantitative and qualitative differences in antibiotic utilization between Spain and Denmark, with a greater consumption, overall and of cephalosporins and broad-spectrum penicillins in every age group and quinolones in the adult population, in Spain.29 Thus, in 2018 the J01_B/N ratio in Spain was 59.4 compared to 0.6 in Denmark.2 In our study, we found a decrease in this indicator throughout the study period, which reflects an improvement in the patterns of antibiotic use. In contrast, this same indicator was noticeably higher and was increasing over the same period in the general population of Spain.

There are limitations to our study, chiefly those common to drug utilization studies with similar methodology based on data retrieved from pharmacy records. We obtained data on the dispensation in community pharmacies of drug prescriptions funded by the SESPA, but did not include data for dispensation without prescription (self-medication) or of prescriptions issued from the private health care system or other insurers. Since 2002, a medical prescription is required in Spain for dispensation of an antibiotic agent in pharmacies. The volume of antibiotics dispensed through health insurance companies and prescriptions by private physicians is not small, although in Asturias it is below the national average; in 2019, Asturias ranked second last of all autonomous communities in Spain in the dispensation of antibiotics prescribed outside the public health service.30 In addition, the coverage of the paediatric population of Asturias by the public health system is high (98% in year 2018).15 We did not analyse consumption at the hospital level, as we know that it is very small in proportion to consumption in the community after a quantitative study in one of the central health areas found that it amounted to 1.5% of the overall consumption. We did not analyse consumption by age group, seasonal patterns in prescription or the appropriateness and suitability of antibiotherapy prescriptions. Lastly, we did not analyse prescriptions made exclusively by paediatricians, but all prescriptions made by any physician that managed paediatric patients in the period under study in primary care or after-hours urgent care centres, as we sought to assess the actual exposure of the paediatric population to antibiotic agents.

In conclusion, our study, conducted in the paediatric population of one autonomous community in Spain and spanning 14 years, found a decrease in the overall consumption of antibiotics at the outpatient level in recent years and a more rational use of these drugs. We ought to highlight and warn of the marked increase in the consumption of azithromycin and quinolones throughout the period under study. It is important that physicians be aware of local antimicrobial consumption patterns to be able to implement antimicrobial stewardship strategies and identify the determinants of antimicrobial prescription for the purpose of containing the development of antibacterial resistance.

The study received a €500 grant from the Fundación Ernesto Sánchez Villares (FESV) of the Sociedad de Pediatría de Asturias, Cantabria y Castilla y León (SCCALP).

The authors have no conflicts of interest to declare.