PDF
Citation
Print
INTRODUCTION
Antimicrobial resistance (AMR) has emerged as a global public health concern, of which the inappropriate use of antibiotics is an important factor.1 The World Health Organization (WHO) adopted a global action plan on AMR in 2015 that sought to optimize the use of antimicrobial agents. In recent years, AMR has increased in children globally.2 Kim et al.3 documented the proportion of macrolide-resistant Mycoplasma pneumoniae pneumonia worldwide; overall, it increased from 25% in 2000 to roughly 75% in 2020. Prescriptions for acute respiratory infection (ARI) account for a large proportion of antibiotic prescriptions among children, with data varying by region (from 29% in the Netherlands in 2021 to 76% in South Africa in 2020); unnecessarily prescribing antibiotics is thought to promote AMR.4567 Most cases of ARI in children are caused by viruses and are self-limiting; therefore, antibiotics are only recommended for ARI in certain circumstances, such as a suspected or confirmed bacterial infection.67 Croup, or laryngotracheobronchitis, is a common ARI in infants and children that affects the larynx and trachea and may extend to the bronchi. Croup is most often caused by a virus.8 Since the use of antibiotics is often not indicated in children with ARI, reducing antibiotic usage is an important strategy of AMR countermeasures.
The Korean health authorities consider AMR an important healthcare problem and introduced the quality evaluation program on antimicrobial prescription for upper respiratory infection (URI) in 2006. In response to the WHO’s global action plan on AMR declared in 2015, Korea initiated the Korean Action Plan on AMR in 2016. The core elements of this plan include implementation of antimicrobial stewardship programs (ASPs) in medical facilities and reduction of unnecessary antibiotic prescriptions.91011 Following implementation of healthcare policies such as assessments of antibiotic usage, the use of antibiotics in URIs has decreased in Korea (total population, 71% in 2002 to 35% in 2021), but children are still more likely to receive antibiotics than all adults or those aged ≥ 65 years old (children, 38%; adults, 35%; adults aged ≥ 65 years, 21% in 2021, respectively).12 As such, it seems like there is a gap between knowledge and clinical practice in the area of antibiotics prescribing in pediatric patients with URI.
In this study, using outpatient administrative prescription data, we aimed to describe the oral antibiotic prescribing pattern among pediatric patients with laryngotracheobronchitis in Korea. We determined the prevalence of antibiotic prescriptions and its associated factors. We also investigated the classes of antibiotics that are most commonly used. The results will be useful for improving ASPs; selecting target groups for ASP strategies such as educating target groups on appropriate use of antibiotics could contribute to reducing unnecessary use of antibiotics in pediatric patients and eventually decrease AMR in children in Korea.
METHODS
Data source
This study was conducted using data from the government agency Health Insurance Review and Assessment Service (HIRA). Korea has a universal health insurance system. Aside from 3% of the population who cannot afford the insurance premium, 97% of Korean nationals are enrolled mandatorily in the National Health Insurance Service (NHIS). For the medical expenditures supported by the NHIS, HIRA reviews the claims of reimbursement requested by medical providers. No antibiotics in Korea are approved as over-the-counter medications, and antibiotics expenses are covered by the NHIS. Only a very few cases of prescribed antibiotics not covered by the NHIS (e.g., prescribing errors by physicians) were not included in HIRA data used in this study.
Study design
We accessed HIRA’s big data analysis system. A prescription dataset was provided as we requested. The de-identified prescriptions analyzed in this study were issued as outpatient prescriptions (outpatient prescriptions issued in follow-visits after hospitalization or emergency department visits were also included) to pediatric patients ≤ 5 years old from January 2017 to December 2020 with a primary diagnosis of laryngotracheobronchitis. A primary diagnosis of laryngotracheobronchitis was defined by a prescription with either an International Classification of Disease 10th Revision, code of J050 (croup), J040 (laryngitis), or J041 (tracheitis) as the first-listed diagnosis code. Because laryngotracheobronchitis affects the larynx and trachea, both J040 and J041 were included. The prescriptions analyzed in this study did not include diagnosis codes other than first-listed diagnosis codes. Patients may have made more than one visit, and more than one prescription may have been issued during a single episode of laryngotracheobronchitis, so each prescription was analyzed in this study. In HIRA’s big data analysis system, using an information-extraction system, the diagnosis code (J050, J040, J041) was verified for each prescription. Demographic information (sex and age), information of the medical facilities visited (visit date, type of hospital, specialty of physician, location of hospital), and prescribed drugs (generic names) were also extracted.
The prescription rate was calculated as the number of prescriptions containing antibiotics divided by the total number of prescriptions in an identified group. The type of medical facility was categorized based on the hospital tier classification documented in the Health Care Law of Korea: clinic was defined as a medical facility with a bed capacity ≤ 29, a hospital was defined as a medical facility with a bed capacity of 30–99, a general hospital was defined as a medical facility with a bed capacity ≥ 100, and a tertiary referral center was defined as a hospital with a bed capacity ≥ 100 and that met the criteria of a center by the Ministry of Health and Welfare of Korea.13 Children aged ≤ 2 years with ARI are often more critically ill than older children with ARI,1415 so patient age was categorized as ≤ 2 years or 3–5 years. Assuming there are differences in knowledge in pediatric diseases and dealing with parents, physician specialty was categorized as pediatrics or non-pediatrics. Due to intercity differences in accessibility of medical facilities, the hospital location was categorized as metropolitan (the capital city Seoul, administrative city Sejong, and metropolitan cities of Busan, Daegu, Daejeon, Gwangju, Incheon, and Ulsan) and non-metropolitan areas (other than metropolitan cities, included small and medium-sized cities and rural regions). Oral antibiotics were classified as extended-spectrum penicillins (including amoxicillin, amoxicillin-clavulanate), cephalosporins (including first, second and third generation cephalosporins), and macrolides, which are the three most commonly prescribed antibiotic classes for children in outpatient clinics in Korea.16 The antibiotic prescription frequency was described according to patient age, medical facility type, physician specialty, and region.
Statistical analysis
Descriptive statistics were used to calculate frequencies. Multivariate logistic regression analysis (MLRA) was performed to determine factors associated with antibiotic prescription along with dependent variables of age, medical facility, specialty of clinician, and hospital location for total prescriptions. The variable patient age may correlate with other variables (e.g., patients aged 0–2 years are more likely to visit pediatricians than another specialty). With concern about this collinearity confounding the results, MLRA was performed for each age set (0–2 years and 3–5 years), with dependent variables of medical facility, clinician specialty, and hospital location. Adjusted odds ratios (aORs) were estimated with 95% confidence intervals (95% CIs). Statistical analysis was conducted using SAS version 9.4 (SAS Institute, Cary, NC, USA). All statistical tests were two-tailed, with a P value < 0.05 considered statistically significant.
Ethics statement
The use of data was approved by the HIRA data access committee (HIRA research data M20220330905). This study was performed according to the regulations of HIRA. This study was approved by the Institutional Review Board (IRB) of Eunpyeong St. Mary’s Hospital (PC22ZISI0124). No informed consent was required from patients due to the nature of public data from HIRA.
RESULTS
A total of 2,358,194 prescriptions was issued for 672,782 outpatients ≤ 5 years old and diagnosed with laryngotracheobronchitis from January 2017 to December 2020. Table 1 displays the demographic and hospital information related to these prescriptions. A large majority (86.1%) of these prescriptions was issued by a clinic. Among the medical specialties, 23.9% of the prescriptions were issued from pediatricians, while the rest was issued from non-pediatric specialties including otolaryngology, internal medicine, family medicine, and others.
Table 1
Baseline characteristics of the prescription for laryngotracheobronchitis between 2017 and 2020
Overall, the antibiotic prescription rate was 35.2% (829,172 of 2,358,194). Table 2 displays the results of MLRA performed to identify associated factors of antibiotic prescription. Management in a hospital demonstrated the highest aOR for antibiotic prescription (aOR, 22.33; 95% CI, 20.87–23.89; P < 0.001). Management in a clinic demonstrated the second-highest aOR for antibiotic prescription (aOR, 12.66; 95% CI, 11.83–13.54; P < 0.001). Antibiotic prescriptions were also associated with management in a general hospital (aOR, 8.96; 95% CI, 8.37–9.59; P < 0.001), management by a pediatric specialty (aOR, 1.43; 95% CI, 1.42–1.44; P < 0.001), care in a hospital located in a non-metropolitan region (aOR, 1.46; 95% CI, 1.45–1.47; P < 0.001), and patient age of 0–2 years (aOR, 1.27; 95% CI, 1.27–1.28; P < 0.001). Table 3 displays the results of MLRA performed to identify associated factors of antibiotic prescription by age subset; the age subsets demonstrated the same associated factors as the total set (Table 2). Management in a hospital demonstrated the highest aOR in the 0–2-year-old subset (aOR, 23.09; 95% CI, 21.38–24.95) and 3–5-year-old subset (aOR, 18.98; 95% CI, 16.50–21.84). Management in a clinic or general hospital, management by pediatric specialty, and care in a hospital located in a non-metropolitan region were also factors associated with antibiotic prescription in both the 0–2-year-old and 3–5-year-old subsets.
Table 2
Factors associated with antibiotics prescription (total set)
Table 3
Factors associated with antibiotics prescription by age subset
Table 4 shows the proportion of specific antibiotic types prescribed. Overall, penicillin derivatives were the most common antibiotics prescribed (18.6%), followed by cephalosporins (9.4%) and macrolides (8.5%). According to all variables (e.g., patient age 0–2 or 3–5, pediatrics or non-pediatrics), extended-spectrum penicillins were the most frequently prescribed. Fig. 1 shows the antibiotic combinations of all 2,358,194 prescriptions; among them, 31,111 prescriptions (1.3%) contained two or more antibiotics. A combination of two extended-spectrum penicillins was most common (1.1%), followed by the combination of an extended-spectrum penicillin and macrolide (0.1%) and the combination of a cephalosporin and macrolide (0.1%).
Table 4
Proportion of antibiotic types prescribed
DISCUSSION
This study described the nationwide antibiotic prescription pattern in pediatric patients with laryngotracheobronchitis, which typically is a viral disease. We found that management in a hospital was the strongest associated factor for antibiotic prescription. However, management in a clinic, general hospital and management by a pediatrician were also significantly associated with antibiotic prescriptions. Multiple factors may influence antibiotic prescriptions in pediatric patients with ARI such as a physician’s uncertainty of the infectious etiology, fear of complications, insufficient knowledge of AMR and a parent’s demand for antibiotics. Parental demand may be expressed in several ways including direct request, an emphasis on illness severity, or mention of previous use of antibiotics.17 To manage these concerns, it is helpful to rapidly diagnose the pathogen, convey the differences in viral and bacterial infections, and explain the risk of antibiotics to patient’s parents. However, a rapid diagnosis may not be accessible in small hospitals, and physicians in small hospitals may be concerned about losing their patients to follow-up. Physicians who are most aware of the consequences of antibiotic overuse are usually employed by large hospitals.18 The factors contributing to antibiotic prescriptions are more likely to occur in smaller hospitals, including clinics, hospitals and general hospitals, rather than at tertiary referral centers. It can be challenging to interact with the parents of pediatric patients, who may be anxious, ask a lot of questions, have fixed ideas about medical management, and expect their child to recover very quickly.19 To satisfy parents and reduce the time needed to persuade against certain treatment, pediatricians may have prescribed antibiotics despite recommendations in guidelines.
We also found that a younger age (of 0–2 years) was associated with antibiotic prescription. In ARI, children younger than 2 years have a higher risk of severe symptoms and progression to respiratory distress. In respiratory virus surveillance conducted in the United States among pediatric patients aged 0–18 years (2016–2021), patients younger than 2 years accounted for more than half of hospitalizations.14 Similar results were reported in Korea by Jeon et al.15; in their study of pediatric patients hospitalized with croup in a single hospital, those ≤ 2 years old accounted for more than 60% of hospitalizations. In this study, patients with 0–2 years may have truly shown more severe symptoms or younger age may have influenced physicians to prescribe antibiotics given fears of disease progression and complications. Hospitals located in non-metropolitan regions were also significantly associated with antibiotic prescriptions. In Korea, there are differences in the geographical distribution of healthcare facilities. In metropolitan areas, hospitals are more concentrated and are more accessible due to better developed public transportation systems than in non-metropolitan regions.2021 Those in non-metropolitan areas may also have difficulty scheduling follow up visits or seeking emergency care if symptoms worsen and rapid management is needed. The decreased availability of care and concerns regarding follow-up care may have influenced the increased use of antibiotics in non-metropolitan areas compared to those in metropolitan areas.
In an ARI, antibiotics are only recommended when a bacterial infection is suspected or confirmed (e.g., group A streptococcal pharyngitis, bacterial infection established sinusitis) or in certain circumstances of acute otitis media (e.g., < 6 months of age, high risk of complications). When antibiotics are indicated, amoxicillin or amoxicillin/clavulanate usually is recommended as first line therapy.2223 In this study, the overall antibiotic prescription rate was 35.2%. No nationwide data have been published regarding the use of antibiotics in children with laryngotracheobronchitis. However, several reports investigated antibiotic use among children with URI. According to a report on the appropriateness of antibiotic use by HIRA, the antibiotic prescription rate in children with URI ranged from 38–40% every year from 2017–2021,12 which was a similar prescription rate to that in this study. Shin et al.24 investigated the prevalence of antibiotics for URI among pediatric patients in Korea during 2009–2011 and ultimately determined an antibiotic prescription rate of 55–60%. Although the antibiotic use in their work was greater than that in our study, they similarly found that antibiotics were more likely to be prescribed in clinics, hospitals, and general hospitals compared to in tertiary referral centers, and extended-spectrum penicillins were most commonly used. In studies conducted in the US and Finland, extended-spectrum penicillins were also the most frequently used antibiotics in children with URI.2526 Physicians may be influenced by guidelines in choosing antibiotics, amoxicillin or amoxicillin/clavulanate is recommended as first line therapy. In this study, of all prescriptions, only 1.3% involved an antibiotic combination. A combination of antibiotics is usually indicated in the empirical treatment of life-threatening infections, treatment of polymicrobial infections, prevention of bacterial resistance, and with an expectation of drug synergism.27 Although antibiotic combinations made up a small portion of the overall prescriptions in this study, antibiotic combinations are unreasonable in pediatric patients with laryngotracheobronchitis.
The overuse of antibiotics facilitates the emergence of resistant bacterial strains, which fuel the need for a new drug and can alter microbiomes.67 AMR in children should not be overlooked in Korea. The proportion of macrolide-resistant M. pneumoniae has shown an increasing trend; Hong et al.28 reported rates of 14.7% in 2006 and 62.9% in 2011, while Kim et al.29 reported a rate of 87.2% in 2015. Park et al.30 investigated proportions of extended-spectrum β-lactamase–producing Escherichia coli and Klebsiella species among uropathogens in four hospitals located in different regions, finding that proportions ranging from 5–15% in 2011 increased to 10–25% in 2016–2017. The results of our study suggest that the ASPs need to focus on physicians in hospitals, clinics, general hospitals, and pediatric specialties. Particularly, ASPs may need to physicians in hospitals which showed strongest associations with antibiotic prescription in the total set and age subset analyses. In Korea, there are approximately 120 privately run pediatric-specialized hospitals in different regions.31 ASPs may need to these medical facilities. To operate ASP effectively, team members such as nurses and pharmacists who perform regular audit and provide personalized feedback are essential.1011 Particularly, clinics, hospitals and general hospitals may encounter financial challenge in implementing ASP. The establishment of a medical fee to support ASP is needed.11 Education program to raise awareness of AMR, appropriate use of antibiotics in children with ARI, and how to convey it to parents will be useful for physicians. ASP campaigns that use pamphlets or newsletters can be helpful in educating parents about antibiotic overuse.17
This study has provided valuable insight on antibiotic use in pediatric patients with ARI. However, it has certain limitations related to the nature of the insurance claim data. First, the diagnoses cannot be free from miscoding. Second, prescriptions only included first diagnostic codes and did not include clinical medical records; thus, co-infections requiring antibiotics (e.g., acute otitis media) and underlying diseases could not be identified. A third limitation is that the HIRA data study period was short (2017–2020). Fourth, this study was conducted based on insurance claims data from Korea, so its results cannot be generalized to other regions of the world with different healthcare settings.
In this study, the antibiotic prescription rate of pediatric patients with laryngotracheobronchitis was 35.2%. Extended-spectrum penicillins were most frequently used. Associated factor analysis demonstrated that management in a hospital was the most significant factor influencing antibiotic prescription. Management in a clinic, general hospital, by a pediatric specialist, in a hospital located in a non-metropolitan region, or of a patient aged 0–2 years were also significantly associated with antibiotic prescriptions.
XML Download