Antibiotic prescriptions for children younger than 5 years with acute upper respiratory infections in China: a retrospective nationwide claims database study

Antibiotics are one of the most common medications prescribed for outpatients [1‐3], and Upper Respiratory Infections (URIs) account for approximately 50–70% of total antibiotic prescriptions, even though most cases are of viral origin [4‐7]. A prior study in US estimating potential rate of inappropriate outpatient antibiotic use had relied on estimates of bacterial prevalence and concluded that more than 50% of antibiotics for URIs were unnecessary [8]. The latest study of China reported that [9], the proportion of inappropriate antibiotic prescribing in China is higher than that in the US and UK. Fortunately, in the past 10 years, the overall decline in the use of antibiotics was found both in US, European countries, and China [10‐14]. However, compared with European and American countries, the proportion of broad-spectrum antibiotics (second- and third generation cephalosporins and azithromycin, etc.) was still higher in China [15].

Over and inappropriate use of antibiotics, especially the broad-spectrum antibiotics, not only contributes to the financial burden on health services but also exposes patients to adverse events. More importantly, it is a major contributor to emerging antimicrobial resistance [4], which poses a substantial threat to global public health [16] and the UN General Assembly has recognized antimicrobial resistance as a global priority health issue. To response, China has taken many actions to curb the development of antimicrobial resistance. During the period of 2011–2014, the National Health and Family Planning Commission launched a 3-year Special Antimicrobial Stewardship Campaign nationwide. In 2016, the National Action Plan on Bacterial Resistance was formulated responding to WHO’s request for the “Global Action Plan on Antimicrobial Resistance”. Subsequently, in 2017, with the goal of strengthening the management of antimicrobial use in children, the China Children Action Plan for Rational Use Antimicrobials was released. However, the antimicrobial resistance in key pathogens remains high in China and there were increasing rates of resistance to carbapenems in clinical isolates of Escherichia coli and Klebsiella pneumoniae during 2014–2017 [17].

The surveillance of antibiotic use is necessary to formulate targeted strategies for prevention and containment of antimicrobial resistance. To date, in mainland China, there are mainly 2 nationwide network systems to monitor antibiotic use and antimicrobial resistance, such as China Antimicrobial Surveillance Network (CHINET) and China Antimicrobial Resistance Surveillance System (CARSS). However, the available national surveillance data samples are largely from adult patients, which might not truly reflect the situation in the pediatric settings. In addition, previous studies on pediatric antibiotic use for URIs were limited to selected health facilities or certain areas in China [18, 19], which might not reflect antibiotic use across the whole country. And thus, we used a nationwide claims database to investigate the antibiotic prescribing practice for pediatric patients with URIs in ambulatory settings.

Table 1 showed the demographics of outpatients younger than 5 years with URIs. In total, there were 92,821 sampled ambulatory care visits for URIs from 59 cities included in the analysis. 73,638 outpatient visits were from the east region of China, which accounted for 79.3% of all visits. There were 41,094 (44.3% of all visits) visits presented in metropolises, and 38,275 (41.2% of all visits) visits presented in provincial capitals.

Table 2 showed the frequency and proportion of antibiotic prescriptions in visits. Overall, antibiotics were prescribed in 27.1% of the outpatient visits for URIs (25,193 of 92,821 visits), of which 27.0% (6814 of 25,193 visits with antibiotic prescriptions) were injection formulations. Of the 25,193 visits receiving antibiotics, 5.7% (1430) received 2 or more antibiotic agents at the same visit. There was a trend towards increasing prescription rate of antibiotics with increasing age groups [(16.0% (767 of 4798 visits)] for children in the 0- to < 1-year-old age group, 20.4% (4205 of 20,625 visits) for the 1 to < 2-year-old group, 26.3% (4201 of 15,983 visits) for the 2 to < 3-year-old group, 29.9% (7951 of 26,575 visits) for the 3 to < 4-year-old group and 32.5% (8071 of 24,840 visits) for the 4 to < 5-year-old group; data were not shown]. The antibiotic prescribing practices varied by regions (p < 0.001), and the rate of prescribing antibiotics and the proportion of injection formulations were highest in central region, at 36.9 and 64.5%, respectively, followed by the west (32.8, 36.4%) and east (25.6, 23.2%) of China. Compared with other cities, visits from the fourth-tier cities received more antibiotics and injections (39.5 and 56.9%, respectively, p < 0.001). Among medical institutions, healthcare providers in primary care facilities tended to prescribe more antibiotics and injection dosage for URIs than those in tertiary hospitals (P < 0.001).

Figure 2 showed the types antibiotic prescriptions based on ATC classification. Third-generation cephalosporins (J01DD) were the most common group of antibiotics prescribed (34.9% of all visits with antibiotics prescriptions, 8790/25193), followed by macrolides and lincosamides (J01FA and J01FF; 24.3%, 6133/25193), second generation cephalosporins (J01DC; 23.3%, 5876/25193). Such a prescribing pattern was more obvious in the eastern region, economically developed cities, tertiary and secondary hospitals. Of 25,193 visits with antibiotic prescriptions, only 3937 (15.6%) were prescribed penicillins (J01CA, J01CE, J01CF, J01CG and J01CR). In Central and Western China, penicillins (J01CA, J01CE, J01CF, J01CG and J01CR) were most used, accounting for 29.5% (134/454) and 38.7% (2279/5892), respectively. The top 10 antibiotic agents appeared in Table 3. Cefixime (20.4%, 5146/25193), azithromycin (15.4%, 3881/25193), and cefaclor (12.7%, 3186/25193) were the 3 most frequently dispensed, accounting for half of all visits with antibiotic prescriptions cumulatively.

As early as 2000, the Respiratory Group of Pediatrics Society, Chinese Medical Association have published the guideline for rational use of antibiotics for acute respiratory infections [26]. Furthermore, in 2018, the Respiratory Society of Chinese Medical Association released general guidance for URIs and revised in 2020 [27], indicating that the management of the URIs should focus on relieving symptoms rather than prescribing antibiotics, unless suspected of having bacterial infection. Antibiotics recommended for treating URIs were penicillins, first-generation cephalosporins, macrolides and quinolones. Penicillins were recommended as the first-line therapy for children. At present, quinolones are not recommended as first-line therapy for children in China. In this analysis of a large, nationally representative database of outpatient prescriptions, we found that the 3 most prescribed antibacterial agents were cefixime, azithromycin, and cefaclor, which was not in accordance with the guideline recommendations in China.

Currently, the percentage of visits with an antibiotic prescribed is a main indicator for assessing and supervising antibiotic use [28]. In 2011, the European Surveillance of Antimicrobial Consumption project (ESAC) [29] recommends that the percentage of antibiotics prescribed for outpatients > 1 year with acute URIs and acute tonsillitis should be limited to 0–20%. During the 2011–2014, the nationwide special rectification campaign of China [30] stipulating that the prescription rate for outpatients in general hospitals and pediatric hospitals should not exceed 20 and 25%, respectively. In our study, we estimated that 27.1% of pediatric outpatients under 5 years with URIs received antibiotic prescriptions, which was slightly higher than the rate recommended by the ESAC [29] and the Ministry of Health of the P.R. China [30] but lower than the previous studies of China. For example, Yuan. et al. [19] reported that 90.45% pediatric outpatient visits for suppurative tonsillitis resulted in antibiotic prescriptions, 78.77% for pharyngitis, and 52.25% for other URIs in 2010–2011. In a 2017 meta-analysis in China [31], 89.2% of children received antibiotics for acute URIs in the past 16 years. The reduction in antibiotic prescription rate was largely attributed to the antibiotic stewardship campaigns of China and some evidence have confirmed its positive impact [12, 32]. Since we only had collected the data on 2015, the trend in antibiotic use and the effects of the antimicrobial stewardship program on antibiotic prescribing could not be evaluated directly in our study.

In other countries, there also showed high antibiotic prescription rates for URIs. In the United States, Fleming-Dutra. et al. [4] reported that of all the sampled ambulatory care visits for the population younger than 19 years, 84.7% were associated with antibiotic prescriptions for sinusitis, 56.2% for pharyngitis, and 21.2% for viral URIs, which was similar with the findings of Kronman. et al. in 2014 [8]. The differences between these studies in US and ours are probably resulted by the fact that we did not separate pharyngitis and sinusitis from URIs. Also, the discrepancy of results could be explained by the different regions and populations. And, Zhao. et al. [33] had reported that children aged 6–17 years had the higher antibiotic prescription rate for URIs than preschoolers. Compare with other foreign studies, 27.1% of visits resulted in antibiotic prescriptions, considerably lower than reported in the South Korea (58.7%) [34] but higher than reported in the Spain (< 20%) [35].

In our study, approximately one-third of the prescribed antibiotics were in injection formulations. According the Guiding Principles of China [36], most patients who require antimicrobial drugs for mild-to-moderate infections should be treated orally instead of intravenously or intramuscularly. Moreover, injections expose patients to unnecessary drug-related side effects, substantial financial burden, and medical risks, such as bloodborne diseases. It is gratifying that, compared with previous studies in China, our result is substantially lower than the percentages of injections reported in Yuan’s (78.22% for suppurative tonsillitis, 60.58% for pharyngitis, and 16.47% for other URIs) and Zhang’s (41%) studies [18, 19]. Furthermore, the percentage of injections used maybe overestimated in our study, because the fact that there are more visits related to this administration form, while for oral antibiotic use there is only one or maximum plus one follow-up visit.

In our study, the most common antibiotics prescribed (79%) were third generation cephalosporins, macrolides and second generation cephalosporins, which is consistent with prior studies in China [15, 19, 37] and Japan [38, 39]. In fact, penicillins were the first choice recommended for URIs by the guidelines of China [26, 27]. Recommendations from ESAC indicated that beta-lactamase-sensitive penicillins (J01CE) should account for 80–100% of antibiotics used for URIs [29]. However, this proportion for all penicillins (J01CA, J01CE, J01CF, J01CG and J01CR) was only 15.6% in China. On the contrary, a cross-national analysis on antibiotic use in children of 6 countries found that [40], in the United States, South Korea, Italy, Spain, and Norway, penicillins were the most common prescribed antibiotics. In US and Norway [40, 41], if a penicillin was prescribed for pediatrics, the narrow-spectrum penicillins were commonly used. According to the Chinese Pharmacopeia Guidelines for Clinical Drug Use, penicillins require skin tests before use in China, which largely limits the use of penicillins in our country.

Compared with the top 10 antibacterial agents in pediatrics worldwide [42], the proportion of broad-spectrum antibacterial used was significantly higher in our study, especially in the eastern region, economically developed cities, tertiary and secondary hospitals. However, compared with narrow-spectrum antibiotics, broad-spectrum antibiotics were unable to provide better clinical outcomes and were associated with higher rates of adverse events for children with acute respiratory tract infections [43]. More broad-spectrum antibiotics prescribed may be due to physicians’ uncertainty about etiology, insufficient knowledge of bacterial resistance, driven by economic interests and urging of parents [44]. Therefore, to restrict the use of broad-spectrum antibiotics, it is necessary to strengthen continuous medical education on the management of childhood common infections and supervise the compliance of antimicrobial stewardship recommendations.

Our analysis also showed that pediatric visits in midwestern China and underdeveloped cities received more antibiotics and injection dosage forms. To some extent, such differences were related to the imbalanced economic development and discrepancies in healthcare education resources. Also, studies in Japan [39] and Korea [34] suggested that health facility level are associated with antibiotic prescriptions in agreement with our results. To narrow the gaps among regions and cities and promote the rational use of antibacterial drugs, efforts are urgently needed to strengthen the supervision and intervention of antimicrobial prescribing. Establishing a nationwide pediatric surveillance network is the key to monitoring antibiotic use in children. Other studies have shown that the characteristics of physicians, such as specialty, education level and seniority, have a significant influence on the prescribing of antibiotics [45‐48], but relevant evidence is limited in China. Thus, further study should focus on the effect of provider characteristics on antibiotic prescribing.

To our knowledge, in mainland China, this study is the first to describe antibiotic prescriptions for children with URIs using a claim database covering almost all provinces of China. However, there were also some limitations. First, data from rural areas were not available in the 2015 CHIRA database, therefore, it was not representative of rural China. Second, in this study, each visit for an infection was the unit of analysis and not each infection episode. Thus, antibiotic prescriptions within the same infection episode were captured multiple times. Third, the diagnosis of URIs identified was failed to eliminate other co-infections. Therefore, we could not verify that all antibiotic prescriptions in the dataset were for URIs. Fourth, it is not possible to determine how many patients were bacterial infected in our study because testing for respiratory pathogens is not routinely performed in outpatient settings. Therefore, it is urgent to promote the development of rapid pathogen tests to guide the rational use of antibiotics. Another limitation is that the CHIRA database does not collect information on vital signs, physical exam findings, drug allergy and previous medication failure of patients, which could affect physician’s treatment decisions on antibiotic selection.

In mainland China, the overall rate of antimicrobial prescribing and the proportion of injection formulations prescribed in children under 5 years with upper respiratory infections were at a low level, but still higher in underdeveloped regions and cities. Moreover, the overuse of the third generation cephalosporins, macrolides, and the second generation cephalosporins remains a serious issue. Further efforts should be focused on reducing those non-first-line antibiotic prescribing and narrowing the gaps among regions and cities.