antibiotics

Article
Antibiotic Prescription Patterns in the Paediatric Primary Care
Setting before and after the COVID-19 Pandemic in Italy: An
Analysis Using the AWaRe Metrics
Elisa Barbieri 1, * , Cecilia Liberati 1 , Anna Cantarutti 2,3,4 , Costanza Di Chiara 1 , Angela Lupattelli 5 ,
Michael Sharland 5 , Carlo Giaquinto 1,4 , Yingfen Hsia 6,7 and Daniele Doná 1,5

1 Division of Paediatric Infectious Diseases, Department for Women’s and Children’s Health, University of
Padua, 35100 Padua, Italy; [email protected] (C.L.); [email protected] (C.D.C.);
[email protected] (C.G.); [email protected] (D.D.)
2 Unit of Biostatistics, Epidemiology and Public Health, Department of Statistics and Quantitative Methods,
University of Milano-Bicocca, 20126 Milan, Italy; [email protected]
3 National Centre for Healthcare Research and Pharmacoepidemiology, Department of Statistics and
Quantitative, Methods, University of Milano-Bicocca, 20126 Milan, Italy
4 Società Servizi Telematici—Pedianet, 35100 Padua, Italy
5 PharmacoEpidemiology and Drug Safety Research Group, Department of Pharmacy, and PharmaTox
Strategic Research Initiative, University of Oslo, 0316 Oslo, Norway; [email protected]
6 Centre for Neonatal and Paediatric Infection, St George’s University of London, London SW17 0RE, UK;
[email protected] (M.S.); [email protected] (Y.H.)
7 School of Pharmacy, Queen’s University Belfast, Belfast BT9 7BL, UK
* Correspondence: [email protected]






Citation: Barbieri, E.; Liberati, C.; Abstract: The containment measures following COVID-19 pandemic drastically reduced airway
Cantarutti, A.; Di Chiara, C.; infections, but they also limited the access of patients to healthcare services. We aimed to assess
Lupattelli, A.; Sharland, M.; the antibiotic prescription patterns in the Italian paediatric primary care setting before and after
Giaquinto, C.; Hsia, Y.; Doná, D. the containment measures implementation. For this retrospective analysis, we used a population
Antibiotic Prescription Patterns in the database, Pedianet, collecting data of patients aged 0–14 years enrolled with family paediatricians
Paediatric Primary Care Setting
(FP) from March 2019 to March 2021. Antibiotic prescriptions were classified according to WHO
before and after the COVID-19
AWaRe classification. An interrupted time series evaluating the impact of the containment measures
Pandemic in Italy: An Analysis Using
implementation on the monthly antibiotic index, on the access to watch index, and on the amoxicillin
the AWaRe Metrics. Antibiotics 2022,
to co-amoxiclav index stratified by diagnosis was performed. Overall, 121,304 antibiotic prescriptions
11, 457. https://doi.org/10.3390/
antibiotics11040457
were retrieved from 134 FP, for a total of 162,260 children. From March 2020, the antibiotic index
dropped by more than 80% for respiratory infections. The Access to Watch trend did not change after
Academic Editor: Carl Llor
the containment measures, reflecting the propensity to prescribe more broad-spectrum antibiotics for
Received: 1 March 2022 respiratory infections even during the pandemic. Similarly, co-amoxiclav was prescribed more often
Accepted: 23 March 2022 than amoxicillin alone for all the diagnoses, with a significant variation in the trend slope for upper
Published: 29 March 2022 respiratory tract infections prescriptions.
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
Keywords: antibiotics; AWaRe classification; COVID-19 pandemic
published maps and institutional affil-
iations.

1. Introduction
Antibiotics are among the most prescribed medicines worldwide, especially in pre-
Copyright: © 2022 by the authors. school-aged children [1]. Outpatient antibiotics prescriptions account for most of the total
Licensee MDPI, Basel, Switzerland.
antimicrobials use in the general paediatric population, and its trend is influenced by sea-
This article is an open access article
sons. This tendency follows the incidence of respiratory infections, especially community-
distributed under the terms and
acquired pneumonia (CAP); acute otitis media (AOM); and pharyngitis, the most common
conditions of the Creative Commons
reason for antibiotics prescription in paediatrics [2]. Over-prescribing and inappropriate
Attribution (CC BY) license (https://
prescribing increases antimicrobial resistance and results in preventable adverse drug
creativecommons.org/licenses/by/
4.0/).
reactions [3,4]. Before the coronavirus disease 2019 (COVID-19) pandemic, paediatric

Antibiotics 2022, 11, 457. https://doi.org/10.3390/antibiotics11040457 https://www.mdpi.com/journal/antibiotics

Antibiotics 2022, 11, 457 2 of 14

outpatient prescription data in Italy showed limited use of narrow-spectrum antibiotics

for CAP [5], as well as a tendency to prescribe broad-spectrum antibiotics for pharyngitis.
Moreover, early antibiotic prescription behaviour for AOM was described against the “wait
and see” recommended strategy [6,7].
Since the beginning of the COVID-19 pandemic in March 2020, the implementation
of containment measures (especially school’s closure) drastically reduced the incidence of
respiratory tract infections in the cold season globally. Consequently, differently from pre-
scriptions for chronic diseases medication [8], a reduction in overall antibiotic prescription
was observed in the first months of the pandemic, compared to the same months of the
previous season [9–11]. On the other hand, the modified procedures of access to medical
services and the implementation of remote consultations (by email or phone in place of
in-person visits) might have changed the antibiotic prescribing behaviour, possibly varying
the use of broad-spectrum antibiotics in spite of narrow-spectrum ones.
This study aimed to assess the impact of COVID-19 containment measure implemen-
tation on antibiotic prescriptions in the Italian paediatric primary care setting. To have
our results more generalisable and comparable with data from other countries, we used
the AWaRe classifications and metrics developed in 2017 by the WHO Expert Committee
on Selection and Use of Essential Medicines (and further updated) [12], as well as other
metrics specific for the paediatric population [13]. AWaRe classifies the antibiotics into
three categories on the basis of the empiric treatment: Access (first-choice and second-
choice antibiotics that are generally narrow spectrum with less resistance potential—e.g.,
amoxicillin and co-amoxiclav), Watch (antibiotic classes with broader spectrum and hence
resistance potential—e.g., fluoroquinolones, macrolides, and second- and third-generation
cephalosporins), and Reserve (last-resort option antibiotic classes).

2. Results
2.1. Overall Results
Overall, 121,304 antibiotics were prescribed from 134 family paediatricians (FPs) to
162,260 children, with 278,909 person-years of follow-up during the study period. Half
of the patients were males (51.9%), and almost 60% of the children resided in the North
of Italy (Piemonte, Lombardy, Friuli Ven.-Giulia, and Veneto regions). The mean age of
the population in the period before the COVID-19 containment measure implementation
was 6.2 years, while in the period after, it was 6.4 years (Table S1 in the Supplementary
Materials). The monthly Access prescription rates were around 60% in both pre- and post-
COVID-19 containment measure implementation periods, with co-amoxiclav accounting
for most prescriptions. Co-amoxiclav prescription rates varied from a minimum of 32%
between October and November 2019 and between December 2019 and January 2020 to a
maximum of 45% between June and July 2020 (Figure 1 and Table S2 in the Supplementary
Materials).
The antibiotic index decreased by 80% (relative rate (RR) 0.27 (95%CI: 0.15–0.46),
p < 0.001) following the containment measures implementation in March 2020 from a
maximum of 1.2 antibiotics per child in the pre-period to a maximum of 0.23 antibiotics
per child in the post-period. The increase in the Access to Watch index in the months
immediately after March 2020 (from 1.22 to 1.66 prescriptions per child) did not reach
statistical significance (Figure 1 and Table S1 in the Supplementary Materials).
The amoxicillin to co-amoxiclav index decreased in the months after the implemen-
tation of containment measures, following the increase in the co-amoxiclav prescription
rate. However, it was not significant. The peak was registered between June and July,
with 1.66 Access prescriptions per every Watch prescription (Figure 1 and Table S2 in the
Supplementary Materials).
The number of telemedicine consultations increased significantly by five times follow-
ing the containment measures implementation, with a change in the trend. On the contrary,
the number of well-child visits decreased significantly by 80% from March 2020 (Figure S1
in the Supplementary Materials).
 Antibiotics 2022, 11, x FOR PEER REVIEW
Antibiotics 2022, 11, 457 3 of 14 3

Figure 1. Monthly antibiotic prescriptions in children 0–14 years of age in Italy in the 12 mo
Figure 1. Monthly antibiotic prescriptions in children 0–14 years of age in Italy in the 12 months
before and after the COVID-19 containment measures implementation (A) and interrupted
before and after the COVID-19
series containment
of the monthly measures
index (B) implementation
along with (A) and
the 95% CI. Pedianet interrupted time
2019–2021.
series of the monthly index (B) along with the 95% CI. Pedianet 2019–2021.
Antibiotics 2022, 11, 457 4 of 14

2.2. Results by Diagnosis Class

In the study period, 20,357 (16.8%) antibiotics were prescribed for a diagnosis of lower
respiratory tract infection (LRTI), 3704 (3.1%) for skin and soft tissue infection (SSTI), 58,793
(48.5%) for upper respiratory tract infection (URTI), and 2302 (1.9%) for urinary infection
(UTI). Before the containment measures implementation, the diagnosis most frequently
associated with an antibiotic prescription was URTI, while in the post-implementation
period, especially after August 2020, most of the prescriptions were not associated with a
diagnosis (from a minimum of 33.1% in March 2020 to a maximum of 58.8% in December
2020 of prescriptions prescribed without a diagnosis) (Table S3 in the Supplementary
Materials).
The monthly Access prescriptions rate for LRTI, constituted mainly by co-amoxiclav,
was above the 35% before the COVID-19 containment measures implementation, while in
the post-period was below 30% (Figure 2 and Table S4 in the Supplementary Materials).
The monthly Access rate for SSTI and URTI was higher (around 70% and 60%, re-
spectively) than LRTI rates, and co-amoxiclav rates accounted for 70% and 60% of Access
prescriptions, respectively (Figure 2 and Table S4 in the Supplementary Materials).
Regarding UTI, the Watch rate was above 50% for most of the months, and among
Access prescriptions, co-amoxiclav was the most prescribed antibiotic, accounting for
around 90% of the Access prescriptions (Figure 2 and Table S4 in the Supplementary
Materials).
The amoxicillin index trend did not vary in level after implementing COVID-19
containment measures for any of the diagnosis classes; however, the trend slope for URTI
varied significantly (Figure 2 and Table S4 in the Supplementary Materials).
The number of antibiotics prescribed for a specific diagnosis class per child decreased
significantly after implementing the containment measures for all the diagnoses except
for UTIs (Figure 3 and Table S4 in the Supplementary Materials). The highest fall in the
trend level was observed for LRTIs (−86%), followed by URTIs (−75%) and SSTIs (−32%).
Before the COVID-19 containment measures, the antibiotic index for NA class was around
0.2, while after the implementation, the trend decreased immediately by 64% (Figure 3 and
Table S4 in the Supplementary Materials).
As regarding the Access to Watch trend, the index was above two just for SSTIs
prescriptions. The lowest index was observed for LRTIs in November 2020 (0.33 Access
antibiotics prescribed for every Watch), where it was observed the only variation in the trend
slope after the COVID-19 containment measures implementation (Figure 3 and Table S4 in
the Supplementary Materials).
Co-amoxiclav was prescribed more often than amoxicillin alone for all the diagnoses,
with no variation in the trend after the containment measures implementation, but a varia-
tion in the trend slope for URTI prescriptions (Figure 3 and Table S4 in the Supplementary
Materials).
Antibiotics 2022, 11, x FOR PEER REVIEW 5 of 14

Antibiotics 2022, 11, 457 5 of 14

variation in the trend slope for URTI prescriptions (Figure 3 and Table S4 in the Supple-
mentary Materials).

2. Monthly antibiotic prescriptions

Figure 2. prescriptions in children 0–14 years of age in Italy in the 12 months
months
before and after
afterCOVID-19
COVID-19containment
containmentmeasures
measures implementation
implementationbyby
diagnosis class
diagnosis ((A)((A)
class LRTI, (B)
LRTI,
SSTI,
(B) (C) (C)
SSTI, URTI, (D)(D)
URTI, UTI, (E) (E)
UTI, NA). Pedianet
NA). 2019–2021.
Pedianet 2019–2021.
Antibiotics
Antibiotics 2022,
2022, 11,
11, x457
FOR PEER REVIEW 6 of6 14
of 14

Figure
Figure3.3.Antibiotic
Antibioticindex
index(A),
(A),Access
AccesstotoWatch
Watchindex
index(B),
(B),and
andamoxicillin
amoxicillin to
to co-amoxiclav
co-amoxiclav index
index (C)
(C) in
in children 0–14 years of age in Italy in the 12 months before and after COVID-19 containment
children 0–14 years of age in Italy in the 12 months before and after COVID-19 containment measures
measures implementation by diagnosis class. Pedianet 2019–2021.
implementation by diagnosis class. Pedianet 2019–2021.
Antibiotics 2022, 11, 457 7 of 14

3. Discussion
Our study showed an immediate 80% decrease in the overall monthly antibiotic
prescriptions per child, but no variation was noticed in the Access to Watch index and
co-amoxiclav to amoxicillin index. In general, the decrease was noticed for all diagnoses
except for UTIs. Still, the Watch monthly antibiotic prescription rates remained higher than
30% for most of the respiratory tract infection.
To our knowledge, this is the first study assessing antibiotic prescription patterns
before and after the COVID-19 pandemic according to the AWaRe classifications in the
paediatric outpatient population in Italy. Providing information from a national electronic
Healthcare Record database allowed us to have a likely picture of antibiotic prescriptions
and infections trend (by proxy) and to explore variations in prescribing appropriateness,
using as indicators the antibiotic index and the ratio of amoxicillin on co-amoxiclav, in line
with national indicators [1]. The Access to Watch Index trend was also assessed, according
to 2017 WHO classification, which listed the antibiotics as Access, Watch, and Reserve on
the basis of recommendations for their practical use for the most common infections [14].
In the 13th General Programme of Work 2019–2023, the WHO proposed a country-level
target of at least 60% of total antibiotic consumption being Access group antibiotics.
The results of our study are consistent with previous findings in the general and
paediatric populations in other countries, including the USA, South Korea, Canada, Finland,
and the UK. The number of antibiotics prescribed per child decreased after the COVID-19
containment measures implementation in March 2020 by 80%, and the fall was more evident
for respiratory tract infections (URTIs and LRTIs), the classes for which were prescribed
most of the antibiotics. In general, the reduction in antibiotic prescriptions follows the
decline in primary care and ED visits related to infections, especially acute respiratory
infections (ARIs) [15–18]. The containment measures implementation effectively reduced
the transmission of COVID-19 and the spread of other respiratory viruses, such as influenza
virus [19,20] and respiratory syncytial virus [21,22], which used to be the leading causes for
ARI hospitalisations and ED admission every year. Likewise, other viral illnesses showed
a downward trend after the pandemic begin [23,24]. It seems that this effect lasted in the
12 months after the containment measures implementation.
In our study, overall Access and Watch antibiotics prescriptions fell just after the
beginning of the COVID-19 pandemic in March 2020. In the same way, a national report
from Portugal analysed the overall antibiotic prescriptions in the general population in
the outpatient setting after the pandemic outbreak, finding a significant reduction in the
Watch group antibiotics; this is explained by the authors as a possible result of the decrease
of infections with the containment measures and cancellation of appointments [9]. On the
opposite side, a survey in Jordan highlighted an increase in the use of Watch antibiotics by
26% from 2019 to 2020 and a reduction in amoxicillin use [25]; however, this study involved
the collection of national antimicrobial consumption data in hospitals and community
pharmacies, with a different design considering parenteral antibiotic use in hospitalised
patients that probably increased during the pandemic, as suggested by the authors.
To better describe the antibiotic prescription drivers in our country, we stratified
the data according to the diagnosis, adopting four categories. Overall, in our report,
prescriptions for RTIs dropped significantly; on the contrary, the antibiotic prescription
rate for UTIs did not change after the pandemic began, reflecting the different aetiologies
and pathogeneses, not being related to interpersonal transmission. Differently from our
findings, a study using general practice data in the Netherlands showed a reduction in
episodes treated with antibiotics both for ARIs and UTIs in adults after implementing
anti-covid measures. Authors postulated a general attitude to avoid medical visits for
uncomplicated infections during the pandemic [11].
Interestingly, oppositely to the reduction in the trend of infections and overall pre-
scriptions, antibiotic prescription appropriateness did not improve the same way. Access
antibiotics should always be the first-line therapy, with amoxicillin for URTIs and LRTIs and
co-amoxiclav for UTIs and SSTIs. In our study, the ratio of Watch on Access prescriptions,
Antibiotics 2022, 11, 457 8 of 14

as well as the ratio of amoxicillin on co-amoxiclav, did not change after the containment
measures implementation, as would be expected, assuming that the overall incidence of
severe or non-responding to first-line treatment infections is dramatically decreased [26].
As previously noted in other studies, broad-spectrum antibiotics remain widely prescribed
in primary care in Italy, and from the current study, it is evident that this did not improve
after the COVID-19 pandemic [27].
Amoxicillin to co-amoxiclav index is considered a strong indicator of antibiotic pre-
scription appropriateness in the context of antimicrobial surveillance in the outpatient
setting. The amoxicillin to co-amoxiclav index should be used by regulators as a proxy
to measure prescribing appropriateness in the general outpatient population and as a
stewardship metric. It is not indicated to measure the appropriateness of a single diagnosis
class. Amoxicillin is recommended as a first-line therapy for different respiratory tract
infections (including AOM, pharyngitis, and CAP); however, the treatment of choice can
vary on the basis of the child’s age, history of immunisation, comorbidities, and allergies.
In 2016, the European Commission indicated a value of >=4 as the target ratio to achieve an
appropriate control of broad-spectrum prescriptions [28]. Our report confirms previous
findings reporting the level of this index to be below the threshold. In our report, the index
was below 1 before and after March 2020.
Aware indicators were for the first time used in an international report analysing
the 2015 consumption data of oral antibiotic formulations for children from a global
database [13]. On average, Access to Watch Index was 6. However, in Italy, the value
was reported to be 2.5, suggesting a consumption of Watch antibiotics that is more than
double than the average. However, these values cannot be ultimately compared to the
present study for the methodological differences. While in the first case, sales were counted,
we obtained the prescriptions data that constitute a different surrogate of the effective
administration to children. Furthermore, in the 2015 report, co-amoxiclav was categorised
in the Watch group, while from 2019, it entered the Access one [12].
Amoxicillin Index and Access to Watch Index were stratified according to the diagnos-
tic categories in our report. Independently from the spread of COVID-19, in our research,
the most critical point remains the treatment of LRTI, for which Watch antibiotics (including
macrolides) are the most prevalent. From a recent WHO survey, the use of macrolides is
reported to be increased after the COVID-19 pandemic with around half of the countries
included reporting a rise in azithromycin prescription for patients with respiratory-like
symptoms in healthcare facilities [29].
Strengthening empirical therapy for CAP guidelines diffusion is urgent in our out-
patient setting in order to improve prescribing behaviour and reduce the inappropriate
broad-spectrum prescriptions (i.e., the combination of amoxicillin–macrolide or macrolides
alone as treatment strategy for CAP) [30]. Indeed, the increased incidence in Streptococcus
pneumoniae resistance to macrolides over the past two decades, up to 30% in Europe and
30–35% in USA [31], is a matter of concern. Notably, the resistance rate of S. pneumoniae
to penicillin decreased after the introduction of universal infant immunisation with the
pneumococcal conjugate vaccine and was estimated to be about 12% in 2019 in Europe [32].
Still, second-generation cephalosporin (which belong to the Watch category) prescriptions
for LRTIs, which should be considered in case of treatment failure in younger children,
contributed to the increased prescription of Watch antibiotics for both LRTI and URTI in
our study.
A study from the UK reported stratified antibiotic prescriptions for CAP in the paedi-
atric population before and after the COVID-19 pandemic: for children up to four years,
overall amoxicillin prescription rate per 1000 population decreased from March 2020 by
62.5% [33]. However, other diagnoses were not analysed, and access to watch index and
ratio of amoxicillin to co-amoxiclav were not studied.
In our report, Access prescription rates for URTI were high. However, co-amoxiclav
was still predominantly prescribed, especially just after March 2020, determining a variation
in the amoxicillin to co-amoxiclav slope. It should be addressed that most URTIs have
Antibiotics 2022, 11, 457 9 of 14

a viral aetiology, and if bacterial, S. pyogenes represents the leading cause of pharyngitis,
while S. pneumoniae is one of the leading causes of acute otitis media (AOM). Amoxicillin
alone should still be considered the first choice in most of the cases.
We can speculate that the lack of specific surveillance data on antibiotic resistance
in primary care, particularly after the COVID-19 pandemic, directly impacts prescribers’
behaviour, which might prefer antibiotics with a higher coverage, although this is not
supported by evidence and guidelines.
After the containment measures implementation, telemedicine was employed world-
wide in different fields and situations [34–36]. A reduction in face-to-face consultations
for ill children emerged in our analysis: parents and paediatricians adopted phone calls
and email more often to address health issues and respiratory symptoms with fever, which
could have led to a greater diagnostic uncertainty and a precautional prescription of broad-
spectrum antibiotics. This hypothesis can be confirmed by the drastic increase in the
number of prescriptions not associated with a formulated diagnosis after March 2020.
Indeed, even before the pandemic, children with presumed ARIs or conjunctivitis were
more likely to receive antibiotics when undergoing telemedicine visits compared to other
settings (urgency attendance, family paediatricians) [37,38]. However, evidence from a
systematic review analysing the impact of remote consultation in primary care on antibiotic
prescription before the COVID-19 pandemic were inconclusive [39].
Antimicrobial stewardship (AMS) in the outpatient setting is a well-known topic that
needs to be addressed and implemented, as family paediatricians’ prescriptions represent
the majority of antibiotics dispensed to the paediatric population. During the pandemic,
AMS programs had to cope with several problems, as reported from a WHO survey:
decreased in national and local funding for antimicrobial resistance (AMR) surveillance
activities (more frequently reported in low- and middle-income countries), decreased ability
to work in partnerships, reduction in the staff available for AMR activities, and reduction in
laboratory supplies and equipment for AMR [29]. Moreover, factors driving inappropriate
antibiotic prescribing in the outpatient setting (such as patient pressure to receive an
antibiotic, even when they are not indicated; time constraints limiting accurate treatment
plan; diagnostic uncertainty; and externalised responsibility of paediatricians outside
themselves) [40] could have been further magnified during the COVID-19 pandemic.
Considering the important impact on AMS on AMR, outpatient AMS implementation
should be continued as the pandemic continues. Continuous education and guidelines
supplied to family paediatricians may raise antibiotic prescription consciousness and
improve prescription appropriateness. The CDC in 2017 outlined the core elements of
outpatient AMS. Apart from education and expertise, the other elements include a public
commitment for AMS, action for policy and practice (e.g., requiring a written justification
for non-recommended antibiotic prescribing), and tracking and reporting systems [41].
The strengths of our study are its size, generalisability, and representative coverage of
paediatric patients in Italy. It is particularly relevant because, to the best of our knowledge,
there are no other studies assessing antibiotic prescription patterns before and after the
COVID-19 pandemic according to the AWaRe classifications in the same population.
Our study has several limitations. First, it is retrospective in nature, and there is
significant heterogeneity in how FP reported data on the electronic medical chart, not
always following the standardised list of diagnoses codes. Different diagnoses were
categorised a posteriori by the authors. Second, we cannot say with certainty that every
prescribed medicine was administered. Indeed, some paediatricians might have instructed
parents to follow a “wait and see” approach, administering antibiotics only if needed.
Third, some antibiotic indications were not considered in the analysis, such as odontoiatric
and ophthalmological prescriptions. Fourth, prescribing behaviours among the different
family paediatricians vary. However, we included only prescribers participating for the
whole study period. Moreover, significant heterogeneity is not properly a limit of our study
but a characteristic of the outpatient management, and therefore we can say it is a close
portrait of the Italian paediatric primary care situation.
Antibiotics 2022, 11, 457 10 of 14

4. Materials and Methods

This is a retrospective analysis of antibiotic prescribing from 9 March 2019 to 9 March
2021 using Pedianet, a paediatric primary case database. Pedianet captures data from
the routine patient care using JuniorBit® software. This database contains information on
patient demographic characteristics, prescriptions, vaccinations, diagnoses, and referrals to
specialists and hospitalisations. The Pedianet database contains patient-level data from
children followed by around 160 FPs distributed throughout several Italian regions, includ-
ing Friuli-Venezia Giulia, Liguria, Lombardia, Piemonte, Veneto, Lazio, Marche, Toscana,
Abruzzo, Campania, Sardegna, and Sicilia. The Pedianet database has been widely used
since 2004 to conduct pharmacoepidemiological and pharmacovigilance studies [27]. Ac-
cording to the Italian NHS, each child is assigned to a FP who is the primary referral
for health-related matters. Data are anonymised with a monthly update to a centralised
database based in So.Se.Pe. Srl. the legal owner of Pedianet, in Padova, assessing antibiotic
prescription patterns before and after the COVID-19 pandemic according to the AWaRe
classifications in the paediatric outpatient population in Italy [42]. Only family paediatri-
cians who were part of Pedianet for the whole study period were considered for this study.
Due to the study’s retrospective nature and the use of anonymised data, no ethical consent
was required to conduct the study. The study protocol and the data access were approved
by the Internal Scientific Committee of So.Se.Pe. All children aged <14 years enrolled in the
Pedianet database during the study period were included in the analysis. The person-years
of children included in the study was calculated from either March 2019 or when patients
started receiving care. All patients were following up until the last date of their care or the
end of the study period in March 2021.
Antibiotic prescriptions were classified on the basis of Anatomical Therapeutic Chemi-
cal Classification (ATC) and route of administration in Access, Watch, and Other according
to the Italian Medical competent Authority for Italy (Agenzia Italiana del Farmaco—AIFA)
AWaRe classification adapted from the WHO AWaRe classification [1].
Diagnoses were retrieved using ICD9-CM codes and classified as follows in line
with previously published studies [27] and confirmed by paediatricians specialised in
infectious diseases (C.L., D.D): lower respiratory tract infections (LRTI), upper respiratory
class infections (URTI), urinary tract infections (UTI), skin and soft tissue infection (SSTI),
and not available (NA) (Table S5). Free text algorithms were used to explore reason for
prescriptions not associated with an ICD9-CM code.
The primary outcomes were the rates of prescriptions of amoxicillin (ATC: J01CA04),
co-amoxiclav (ATC: J01CR02), Access and Watch, the antibiotic index (defined as the
number of antibiotic prescriptions per person-years), the Access to Watch index, and the
amoxicillin to co-amoxiclav index. The rate of telemedicine consultation and the rate of
well-child visits were assessed to account for variation in the use of primary healthcare
services.

Statistical Analysis
Results were summarised using numbers and percentages and stratified by diagnosis
class.
An interrupted time series analysis supposing an abrupt step change and a slope
change in monthly outcomes using quasi-Poisson regression models was performed to
determine the impact of COVID-19 containment measures implementation [43]. The total
number of monthly antibiotic prescriptions, log-transformed person-years, a variable
representing the frequency in months (i.e., from 1 to 12), a dummy variable indicating
the pre- and post-intervention periods, and a dummy variable indicating the time in
months since the intervention sand a variable with Fourier terms to account for seasonality
were all the variables considered in the first analysis. For the second analysis, the Watch
prescriptions, log-transformed Access prescriptions, and the four other variables previously
specified were used. For the third analysis, the co-amoxiclav prescriptions, log-transformed
amoxicillin prescriptions, and the four other variables previously specified were used.
Antibiotics 2022, 11, 457 11 of 14

To explore primary healthcare service variation, the models used the total telemedicine
consultations or the total well-child visit, log-transformed person-years, and the four other
previously specified variables. Interrupted time series models were performed on the
overall data, and then models stratified by diagnosis were used. Autocorrelation was
assessed, examining the plot for residuals and the partial autocorrelation function. The
corresponding relative risk and 95% confidence interval (95% CI) according to normal
approximation were calculated. Two-sided p < 0.05 was considered statistically significant.
The analysis was performed using R statistical software—v. 4.0.3 (Vienna, Austria) [44].

5. Conclusions
These data highlight the tendency in prescribing broad-spectrum antibiotics in the
Italian paediatric primary care setting during the pandemic (Watch antibiotics for LRTI and
co-amoxiclav for URTI), possibly reflecting the diagnostic uncertainty. Efforts are needed
to improve outpatient antimicrobial stewardship programmes, reinforcing the ones already
existing and implementing newly available tools for telemedicine.

Supplementary Materials: The following are available online at https://www.mdpi.com/article/10

.3390/antibiotics11040457/s1, Figure S1: Interrupted time series of the monthly index of telemedicine
visits (black) and well-child visits (gray) in children 0–14 years of age in Italy in the 12 months before
and after COVID-19 containment measures implementation along with 95% CI. Pedianet 2019–2021;
Table S1: Demographic characteristics of children included in the analysis in the 12 months before
and after COVID-19 containment measures implementation. Pedianet 2019–2021), Table S2: Monthly
antibiotic prescriptions in children 0–14 years of age in Italy in the 12 months before and after
COVID-19 containment measures implementation. Pedianet 2019–2021), Table S3: Prevalence of.
monthly antibiotic prescriptions in children 0–14 years of age in Italy in the 12 months before and
after COVID-19 containment measures implementation by diagnosis class on total monthly antibiotic
prescriptions. Pedianet 2019–2021, Table S4: Monthly antibiotic prescriptions in children 0–14 years
of age in Italy in the 12 months before and after COVID-19 containment measures implementation by
diagnosis class. Pedianet 2019–2021, Table S5: Diagnosis classes with ICD9-CM codes and descriptive
diagnosis.
Author Contributions: Conceptualisation, E.B., Y.H., D.D., C.L.; methodology, E.B., Y.H., D.D., C.L.,
C.D.C., A.C.; software, E.B.; validation, D.D., C.G.; formal analysis, E.B.; resources, C.G.; data curation,
E.B., C.L., C.D.C.; writing—original draft preparation, C.L., E.B.; writing—review and editing, D.D.,
A.C., A.L., Y.H., M.S., C.G.; visualisation, E.B., Y.H., D.D., C.L.; supervision, D.D., M.S., C.G. All
authors have read and agreed to the published version of the manuscript.
Funding: This study was funded by grants from the Research Council of Norway (“EU-COVID-19—a
multinational registry-based linkage study with focus on risk and protective factors, clinical outcomes
and mental health” project, grant number 312707).
Institutional Review Board Statement: This is an observational, retrospective, non-interventional
study. According to a by-law on the classification and implementation of observational drug-related
research, as issued by the Italian National Drug Agency (an entity belonging to the Italian Ministry
of Health), the present study does not require approval by an Ethics Committee in Italy (Italian
Drug Agency note of 3 August 2007). This study followed the principles of the Declaration of
Helsinki and was compliant with the European Network of Centres for Pharmacoepidemiology and
Pharmacovigilance Guide on Methodological Standards in Pharmacoepidemiology.
Informed Consent Statement: Not applicable.
Data Availability Statement: The data used in this study cannot be made available in the manuscript,
the supplemental files, or in a public repository due to Italian data protection laws. The anonymised
datasets generated during and/or analysed during the current study can be provided on reasonable
request from the corresponding author, after written approval by the Internal Scientific Committee
([email protected]).
Acknowledgments: The authors thank all the family paediatricians collaborating in Pedianet: Angelo
Alongi, Roberta Angelini, Giovanni Avarello, Lucia Azzoni, Maria Carolina Barbazza, Maria Barberi
Frandanisa, Patrizia Barbieri, Gabriele Belluzzi, Eleonora Benetti, Roberto Bezzi, Franca Boe, Stefano
Antibiotics 2022, 11, 457 12 of 14

Bollettini, Andrea Bruna, Ivana Brusaterra, Roberto Budassi, Massimo Caccini, Laura Cantalupi,
Luigi Cantarutti, Luigia Caprio, Massimo Castaldo, Stefano Castelli, Serenella Castronuovo, Monica
Cavedagni, Stefania Censini, Giuseppe Egidio Cera, Carla Ciscato, Mariangela Clerici Schoeller,
Giuseppe Collacciani, Fabrizio Comaita, Ugo Alfredo Conte, Nicola Costanzo, Sandra Cozzani,
Giancarlo Cuboni, Vito Francesco D’Amanti, Rita De Angelis, Roberto De Clara, Annamaria De
Marchi, Emanuele De Nicolò, Gian Piero Del Bono, Gigliola Del Ponte, Tiziana Di Giampietro,
Giuseppe Di Mauro, Giuseppe Di Santo, Piero Di Saverio, Marco Dolci, Mattia Doria, Stefano Drago,
Pietro Falco, Mario Fama, Marco Faraci, Tania Favilli, Mariagrazia Federico, Michele Felice, Enrico
Ferrara, Marta Ferrarese, Michele Ferretti, Paolo Forcina, Claudio Frattini, Ezio Frison, Fabrizio Fusco,
Giovanni Gallo, Andrea Galvagno, Alberta Gentili, Pierfrancesco Gentilucci, Giuliana Giampaolo,
Giuseppe Giancola, Silvia Girotto, Costantino Gobbi, Mauro Grelloni, Mirco Grugnetti, Urania
Lagrasta, Massimo Landi, Paola Lasalvia, M.Rosaria Letta, Giuseppe Lietti, Cinzia Lista, Ricciardo
Lucantonio, Francesco Luise, Diego Luotti, Nadia Macropodio, Francesca Marine, Lorenzo Mariniello,
Gabriele Marostica, Sergio Masotti, Stefano Meneghetti, Massimo Milani, Stella Vittoria Milone,
Angela Maria Monteleone, Pierangela Mussinu, Carmen Muzzolini, Flavia Nicoloso, Laura Olimpi,
Maria Maddalena Palma, Vittorio Pandolfini, Angela Pasinato, Andrea Passarella, Pasquale Pazzola,
Monica Perin, Danilo Perri, Silvana Pescosolido, Giovanni Petrazzuoli, Giuseppe Petrotto, Patrizia
Picco, Ambrogina Pirola, Lorena Pisanello, Daniele Pittarello, Eleonora Polidoro, Elena Porro, Adolfo
Francesco Porto, Elisabetta Profumo, Antonino Puma, Ferdinando Ragazzon, Paolo Rosas, Rino
Rosignoli, Mariagiulia Rosina, Mariella Rossitto, Bruno Ruffato, Lucia Ruggieri, Annamaria Ruscitti,
Annarita Russo, Pietro Salamone, Daniela Sambugaro, Luigi Saretta, Vittoria Sarno, Nico Maria
Sciolla, Paolo Senesi, Carla Silvan, Valter Spanevello, Francesco Speranza, Maura Sticco, Francesco
Storelli, Gianni Tamassia, Paolo Tambaro, Giacomo Toffol, Marco Tondello, Gabriele Tonelli, Angelo
Tummarello, Sergio Venditti, Concetta Volpe, Francescopaolo Volpe, Aldo Vozzi.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Ufficio Stampa-Ivano Comessatti A. Citare il Presente Rapporto Come Segue. Available online: www.aifa.gov.it (accessed on 28
September 2021).
2. Nash, D.R.; Harman, J.; Wald, E.R.; Kelleher, K.J. Antibiotic Prescribing by Primary Care Physicians for Children with Upper
Respiratory Tract Infections. Arch. Pediatr. Adolesc. Med. 2002, 156, 1114–1119. [CrossRef] [PubMed]
3. Mastrangelo, R.S.; Santesso, N.; Bognanni, A.; Darzi, A.; Karam, S.; Piggott, T.; Baldeh, T.; Schünemann, F.; Ventresca, M.; Morgano,
G.P.; et al. Consideration of antimicrobial resistance and contextual factors in infectious disease guidelines: A systematic survey.
BMJ Open 2021, 11, e046097. [CrossRef] [PubMed]
4. Principi, N.; Esposito, S. Antimicrobial stewardship in paediatrics. BMC Infect. Dis. 2016, 16, 424. [CrossRef] [PubMed]
5. Costenaro, P.; Cantarutti, A.; Barbieri, E.; Scamarcia, A.; Oletto, A.; Sacerdoti, P.; Lundin, R.; Cantarutti, L.; Giaquinto, C.; Donà, D.
Antibiotic prescriptions for children with community-acquired pneumonia: Findings from Italy. Pediatr. Infect. Dis. J. 2020, 40,
130–136. [CrossRef]
6. Barbieri, E.; Donà, D.; Cantarutti, A.; Lundin, R.; Scamarcia, A.; Corrao, G.; Giaquinto, C. Antibiotic prescriptions in acute otitis
media and pharyngitis in Italian pediatric outpatients. Ital. J. Pediatr. 2019, 45, 1–9. [CrossRef]
7. Chiappini, E.; Mazzantini, R.; Bruzzese, E.; Capuano, A.; Colombo, M.; Cricelli, C.; Di Mauro, G.; Esposito, S.; Festini, F.; Guarino,
A.; et al. Rational use of antibiotics for the management of children’s respiratory tract infections in the ambulatory setting: An
evidence-based consensus by the Italian Society of Preventive and Social Pediatrics. Paediatr. Respir. Rev. 2014, 15, 231–236.
[CrossRef]
8. Chua, K.-P.; Volerman, A.; Conti, R.M. Prescription Drug Dispensing to US Children during the COVID-19 Pandemic. Pediatrics
2021, 148, e2021049972. [CrossRef]
9. Estrela, M.; Gomes, E.R.; Piñeiro-lamas, M.; Figueiras, A.; Roque, F.; Herdeiro, M.T. The Impact of the COVID-19 Pandemic on
Antibiotic Prescribing Trends in Outpatient Care: A Nationwide, Quasi-Experimental Approach. Antibiotics 2021, 10, 1040.
10. King, L.M.; Lovegrove, M.C.; Shehab, N.; Tsay, S.; Budnitz, D.S.; Geller, A.I.; Lind, J.N.; Roberts, R.M.; Hicks, L.A.; Kabbani, S.
Trends in US Outpatient Antibiotic Prescriptions During the Coronavirus Disease 2019 Pandemic. Clin. Infect. Dis. 2020, 73,
e652–e660. [CrossRef]
11. van de Pol, A.; Boeijen, J.; Venekamp, R.; Platteel, T.; Damoiseaux, R.; Kortekaas, M.; van der Velden, A. Impact of the COVID-19
pandemic on antibiotic prescribing for common infections in The Netherlands: A primary care-based observational cohort study.
Antibiotics 2021, 10, 196. [CrossRef]
12. World Health Organization. The 2019 WHO AWaRe Classification of Antibiotics for Evaluation and Monitoring of Use. World
Health Organization. Available online: https://apps.who.int/iris/handle/10665/327957 (accessed on 23 August 2021).
Antibiotics 2022, 11, 457 13 of 14

13. Hsia, Y.; Sharland, M.; Jackson, C.; Wong, I.C.K.; Magrini, N.; Bielicki, J.A. Consumption of oral antibiotic formulations for
young children according to the WHO Access, Watch, Reserve (AWaRe) antibiotic groups: An analysis of sales data from
70 middle-income and high-income countries. Lancet Infect. Dis. 2018, 19, 67–75. [CrossRef]
14. London, A.G. WHO targets antimicrobial resistance in new essential medicines list. BMJ 2017, 356, j779. [CrossRef]
15. Santi, L.; Golinelli, D.; Tampieri, A.; Farina, G.; Greco, M.; Rosa, S.; Beleffi, M.; Biavati, B.; Campinoti, F.; Guerrini, S.; et al.
Non-COVID-19 patients in times of pandemic: Emergency department visits, hospitalizations and cause-specific mortality in
Northern Italy. PLoS ONE 2021, 16, e0248995. [CrossRef] [PubMed]
16. Isba, R.; Edge, R.; Jenner, R.; Broughton, E.; Francis, N.; Butler, J. Where have all the children gone? Decreases in paediatric
emergency department attendances at the start of the COVID-19 pandemic of 2020. Arch. Dis. Child. 2020, 105, 704. [CrossRef]
[PubMed]
17. Ferrero, F.; Ossorio, M.F.; Torres, F.A.; Debaisi, G. Impact of the COVID-19 pandemic in the paediatric emergency department
attendances in Argentina. Arch. Dis. Child. 2021, 106, e5. [CrossRef] [PubMed]
18. Lazzerini, M.; Barbi, E.; Apicella, A.; Marchetti, F.; Cardinale, F.; Trobia, G. Delayed access or provision of care in Italy resulting
from fear of COVID-19. Lancet Child Adolesc. Health 2020, 4, e10–e11. [CrossRef]
19. Ecdc. Seasonal Influenza 2020−2021 Annual Epidemiological Report. Available online: www.FlunewsEurope.org (accessed on 31
October 2021).
20. Flu Season Summary|CDC. Available online: https://www.cdc.gov/flu/season/faq-flu-season-2020-2021.htm (accessed on 31
October 2021).
21. Tang, J.W.; Bialasiewicz, S.; Dwyer, D.E.; Dilcher, M.; Tellier, R.; Taylor, J.; Hua, H.; Jennings, L.; Kok, J.; Levy, A.; et al. Where
have all the viruses gone? Disappearance of seasonal respiratory viruses during the COVID-19 pandemic. J. Med. Virol. 2021, 93,
4099–4101. [CrossRef]
22. Kim, J.-H.; Roh, Y.H.; Ahn, J.G.; Kim, M.Y.; Huh, K.; Jung, J.; Kang, J.-M. Respiratory syncytial virus and influenza epidemics
disappearance in Korea during the 2020–2021 season of COVID-19. Int. J. Infect. Dis. 2021, 110, 29–35. [CrossRef]
23. Chen, B.; Wang, M.; Huang, X.; Xie, M.; Pan, L.; Liu, H.; Liu, Z.; Zhou, P. Changes in Incidence of Notifiable Infectious Diseases in
China Under the Prevention and Control Measures of COVID-19. Front. Public Health 2021, 9, 1548. [CrossRef]
24. Zhang, D.; Jiang, Y.; Hua, J.; Liu, Q.; Wen, W.; Zhu, M.; Du, J.; Liu, H. Viral aetiology and clinical characteristics of acute respiratory
tract infections in Shenzhen during epidemic of coronavirus disease 2019. Infect. Dis. 2020, 52, 590–592. [CrossRef]
25. Al-Azzam, S.; Mhaidat, N.; Banat, H.; Alfaour, M.; Ahmad, D.; Muller, A.; Al-Nuseirat, A.; Lattyak, E.; Conway, B.; Aldeyab, M.
An Assessment of the Impact of Coronavirus Disease (COVID-19) Pandemic on National Antimicrobial Consumption in Jordan.
Antibiotics 2021, 10, 690. [CrossRef]
26. ECDC-Data from the ECDC Surveillance Atlas-Antimicrobial Resistance. Available online: https://www.ecdc.europa.eu/en/
antimicrobial-resistance/surveillance-and-disease-data/data-ecdc (accessed on 1 February 2022).
27. Barbieri, E.; di Chiara, C.; Costenaro, P.; Cantarutti, A.; Giaquinto, C.; Hsia, Y.; Doná, D. Antibiotic Prescribing Patterns in
Paediatric Primary Care in Italy: Findings from 2012–2018. Antibiotics 2021, 11, 18. [CrossRef] [PubMed]
28. Dumartin, C. Prudent Use of Antimicrobial Agents in Human Medicine: Third Report on Implementation of the Council
Recommendation, Analysis of Countries’ Reports on the Implementation of the Council Recommendation of 15 November 2001
(2002/77/EC) on the Prudent Use of. Vol 2001. 2001. Available online: https://ec.europa.eu/health/system/files/2020-06/amr_
projects_3rd-report-councilrecprudent_1_0.pdf (accessed on 1 January 2022).
29. Tomczyk, S.; Taylor, A.; Brown, A.; de Kraker, M.E.A.; El-Saed, A.; Alshamrani, M.; Hendriksen, R.S.; Jacob, M.; Löfmark, S.;
Perovic, O.; et al. Impact of the COVID-19 pandemic on the surveillance, prevention and control of antimicrobial resistance: A
global survey. J. Antimicrob. Chemother. 2021, 76, 3045–3058. [CrossRef] [PubMed]
30. Williams, D.J.; Edwards, K.M.; Self, W.; Zhu, Y.; Arnold, S.R.; McCullers, J.A.; Ampofo, K.; Pavia, A.T.; Anderson, E.J.; Hicks, L.A.;
et al. Effectiveness of β-lactam monotherapy vs macrolide combination therapy for children hospitalized with pneumonia. JAMA
Pediatr. 2017, 171, 1184–1191. [CrossRef] [PubMed]
31. Jenkins, S.G.; Farrell, D.J. Increase in Pneumococcus Macrolide Resistance, United States. Emerg. Infect. Dis. 2009, 15, 1260–1264.
[CrossRef]
32. Antimicrobial Resistance in the EU/EEA (EARS-Net)—Annual Epidemiological Report for 2019. Available online: https:
//www.ecdc.europa.eu/en/publications-data/surveillance-antimicrobial-resistance-europe-2019 (accessed on 23 August 2021).
33. Andrews, A.; Budd, E.; Hendrick, A.; Ashiru-Oredope, D.; Beech, E.; Hopkins, S.; Gerver, S.; Muller-Pebody, B.; the AMU
COVID-19 Stakeholder Group. Surveillance of antibacterial usage during the COVID-19 pandemic in England, 2020. Antibiotics
2021, 10, 841. [CrossRef]
34. Gerber, J.S.; Ross, R.K.; Bryan, M.; Localio, A.R.; Szymczak, J.E.; Wasserman, R.; Barkman, D.; Odeniyi, F.; Conaboy, K.; Bell,
L.; et al. Association of Broad-vs Narrow-Spectrum Antibiotics with Treatment Failure, Adverse Events, and Quality of Life in
Children with Acute Respiratory Tract Infections. JAMA 2017, 318, 2325–2336. [CrossRef]
35. Hincapié, M.A.; Gallego, J.C.; Gempeler, A.; Piñeros, J.A.; Nasner, D.; Escobar, M.F. Implementation and Usefulness of
Telemedicine During the COVID-19 Pandemic: A Scoping Review. J. Prim. Care Community Health 2020, 11, 2150132720980612.
[CrossRef]
36. Armitage, R.; Nellums, L.B. Antibiotic prescribing in general practice during COVID-19. Lancet Infect. Dis. 2020, 21, e144.
[CrossRef]
Antibiotics 2022, 11, 457 14 of 14

37. Hersh, A.L.; Stenehjem, E.; Daines, W. RE: Antibiotic prescribing during pediatric direct-to-consumer telemedicine visits. Pediatrics
2019, 144, e20191786B. [CrossRef]
38. Penza, K.S.; Murray, M.A.; Myers, J.F.; Maxson, J.; Furst, J.W.; Pecina, J.L. Treating pediatric conjunctivitis without an exam: An
evaluation of outcomes and antibiotic usage. J. Telemed. Telecare 2018, 26, 73–78. [CrossRef] [PubMed]
39. Han, S.M.; Greenfield, G.; Majeed, A.; Hayhoe, B. Impact of Remote Consultations on Antibiotic Prescribing in Primary Health
Care: Systematic Review. J. Med. Internet Res. 2020, 22, e23482. [CrossRef] [PubMed]
40. Zetts, R.M.; Stoesz, A.; Smith, B.A.; Hyun, D.Y. Outpatient Antibiotic Use and the Need for Increased Antibiotic Stewardship
Efforts. Pediatrics 2018, 141, e20174124. [CrossRef]
41. Sanchez, G.V.; Fleming-Dutra, K.E.; Roberts, R.M.; Hicks, L.A. Core Elements of Outpatient Antibiotic Stewardship. MMWR.
Recomm. Rep. 2016, 65, 1–12. [CrossRef] [PubMed]
42. Cantarutti, A.; Gaquinto, C. Pedianet Database. In Databases for Pharmacoepidemiological Research; Springer: Cham, Switzerland,
2021; pp. 159–164. [CrossRef]
43. Bernal, J.L.; Cummins, S.; Gasparrini, A. Interrupted time series regression for the evaluation of public health interventions: A
tutorial. Int. J. Epidemiol. 2017, 46, 348–355. [CrossRef]
44. R: A Language and Environment for Statistical Computing. Available online: https://www.gnu.org/copyleft/gpl.html (accessed
on 13 January 2022).