A cross-sectional study of antibiotic misuse among Chinese children in developed and less developed provinces

Introduction: Antimicrobial resistance is a global health crisis and primarily caused by antibiotic misuse. Antibiotic misuse among children is particularly concerning, and its prevalence may vary from region to region in China with different development levels. Methodology: Zhejiang and Shaanxi were selected to represent developed and less developed provinces in China, respectively. Data of 2924 parents in Zhejiang and 3355 parents in Shaanxi whose children were 0-13 years old were collected through multi-stage stratified cluster random sampling and a self-administrated questionnaire. Chi-square tests and logistic regression models were used for statistical analysis. Results: Compared to parents in Zhejiang, those in Shaanxi were more likely to keep antibiotics for children at home, to engage in selfmedication with antibiotics for children, and to make their children take antibiotics prophylactically. While there were no significant provincial differences between parents’ requests for antibiotics during pediatric consultations, parents in Shaanxi province were more likely to receive prescribed antibiotics. Conclusions: Children in less developed provinces face higher risks of antibiotic misuse at home as well as when attending medical practitioners. Comprehensive educational interventions are required to improve antibiotic use for children all over China but particularly in less developed provinces such as Shaanxi. Furthermore, non-prescription sales and over-prescribing of antibiotics should be reduced by targeted strategies.


Introduction
Antimicrobial resistance (AMR) is recognized as a grave global health threat [1,2]. AMR has already lengthened the course of infectious diseases and increased the mortality rates as well as treatment costs [3]. The rapid development of AMR is primarily caused by widespread misuse of antibiotics [4][5][6] which accelerates the natural process of genetic selection thereby making organisms increasingly resistant to antibiotic treatments [7,8]. Antibiotic misuse among children is particularly concerning, because children fall ill frequently and are therefore more likely to be exposed to unnecessary antibiotics. Children are also more vulnerable to the consequences of AMR than adults [9]. Furthermore, children are grouped together in nurseries and schools, catalyzing the spread of drugresistant bacteria [6]. Antibiotics are commonly used for children with viral upper respiratory tract infections, despite the fact that these diseases are self-limiting and do not respond to antibiotic therapy [10,11]. Antibiotic misuse and AMR are most prevalent in low and middle income countries due to these countries' greater burden of infectious diseases and weaker public health systems [12].
Previous research has revealed that Chinese children are becoming less likely to benefit from antibiotics specifically as a result of AMR [13]. It has also been warned that antibiotic misuse remains rampant among Chinese children, and that both the supply and demand sides are to blame [14]. On the supply side, antibiotics are frequently prescribed to children during pediatric consultations [15][16][17] and nonprescription antibiotics are widely available at community pharmacies [18] despite prohibitory regulations [14,19]. On the demand side, many parents and caregivers admit to keeping antibiotics at home [10][11][12] and giving children antibiotics without seeing a doctor, which is a form of self-medication with antibiotics (SMA) [11,12,20,21]. SMA can be wasteful and harmful, and home-stored antibiotics provide a reservoir for it [22]. Moreover, some parents expect antibiotics to be prescribed thereby influencing doctors' diagnostic and prescribing processes [23].
Antibiotic use and habits vary across China. There are appreciable socioeconomic and public health disparities between different regions which, in turn, affect the degree to which antibiotics are available and used by Chinese population [15,24]. Recent research has also proven that socioeconomic status of a population has profound impact on its antibiotic consumption and prevalence of AMR [25]. Generally, eastern coastal provinces are more prosperous than other parts of China. One study mentioned that the location of a child's residence in China might influence the rate of antibiotic misuse by his/her caregiver [20]. In accordance with that, another research in urban China found that children from an eastern city were much less likely to receive non-prescription antibiotics from caregivers than their counterparts from a western city [11]. It was also reported that regional economic statuses in China have corresponding rates of antibiotics sales in hospitals [26]. So far, little is known about the potential behavioral discrepancies on antibiotic use among different regions in China with unequal development levels especially for children's cases, which may imply intervention priorities in the future.
Therefore, this research aims to assess antibiotic use in Chinese children in two different regions in the hope of exposing regional disparities. The purpose of this study is to compare the following behaviors of parents in developed and less developed Chinese provinces: 1) requesting as well as receiving prescribed antibiotics during pediatric consultations, 2) SMA for children's illnesses, 3) keeping antibiotics at home for children, and 4) making children take antibiotics prophylactically.

Data collection
The data from this study were collected between June, 2017 and April, 2018 as a part of a larger crosssectional research project targeting Chinese parents with children under 13 years old. Based on their location and latest provincial GDP per capita [27], eastern Zhejiang province and western Shaanxi province were selected to represent developed and less developed regions of China respectively. A multi-stage stratified cluster random sampling method was adopted as this method provides adequate representation in a socioeconomically diverse population. We anticipated a study population of approximately 3000 parents per province as a sufficient sample size for subgroup analyses. In each province, a prefecture-level city was randomly selected. Within that city, an urban and a rural district were randomly chosen as sampling sites. At each site, a complete roster of kindergartens, primary schools, and vaccination clinics (in the latter, over 90% of children under 3 received basic immunizations [28]) was obtained. A certain number of them were randomly selected as clusters depending on their scales or daily flows to meet the anticipated sample size. All parents of children aged 4 to 6 years in selected kindergartens, 7 to 13 years in primary schools, and under 3 years old in vaccination clinics were sampled.
We used a self-administered questionnaire (see Supplementary File 1) on Wen Juan Xing (https://www.wjx.cn/), the largest online survey platform in China. Informative leaflets with a unique QR code linked to the questionnaire were issued to parents so that they could access and complete the questionnaire on their smartphones. In kindergartens and primary schools, class teachers helped distribute these leaflets and consent forms to students, which would be presented to their parents after school. Parents who were willing to participate signed the consent form and completed the questionnaire online. Consent forms were brought back to class the following day and collected by teachers.
At vaccination clinics, our research assistants presented the parent(s) with informative leaflets and explained about the survey while they waited for the child's vaccination. After receiving consent, parents completed the questionnaire in the observation area once the child had been vaccinated. In both the school and clinic-based contexts, participants were explicitly informed that: 1) the survey was voluntary and participants could withdraw any time they wanted; 2) anonymity and confidentiality were guaranteed; 3) if the participant had more than one child, the child in question referred to the one receiving vaccination or attending the kindergarten/primary school; 4) the questionnaire should be completed by the parent who acted as the main caregiver of the child.

Measurements
The survey contained four sections: 1) Sociodemographic characteristics, including the parent's gender, whether the parent and his/her spouse lived in a rural or urban area, their monthly household income, their level of education, whether they had ever studied medicine or embarked on a medical career, and finally the gender and age of the child in question. 2) The child's recent medical history: whether the child suffered from any illness in the past month including but not limited to a cold, fever, sore throat, diarrhea, and/or otitis media; whether the parent only selfmedicated the child, took the child to a doctor after selfmedication failed, or directly sought medical advice (including all levels of health facilities). 3) Antibiotic use: whether the parent gave antibiotics to the child during self-medication; when seeking medical advice directly or after self-medication failed, whether the parent requested antibiotics during consultation and whether the parent received a prescription of antibiotics. 4) Other behaviors, including whether the parent kept antibiotics at home for children and whether the parent made the child take antibiotics for prophylaxis in the past year. 'Prophylaxis' here refers to giving a child antibiotics who is not sick with the intention of preventing common infections (e.g. a cold or flu). This questionnaire was drafted and tailored to the Chinese context based on literature review [10,12,29,30] and through interviews with stakeholders and experts, and finalized after a pilot test on validity and reliability with 315 respondents.

Statistical analysis
Data analyses were performed with IBM Statistics 24.0 and a p-value less than 0.05 was considered statistically significant. Chi-square tests and t-tests were used to compare socio-demographic characteristics and behaviors between parents in Zhejiang and Shaanxi provinces. Logistic regression models were applied to further explore regional disparities in antibiotic-related behaviors by controlling for social-demographic variables.

Ethics approval
This study was approved by the Institutional Review Board of School of Public Health, Zhejiang University (reference number: ZGL201706-2).

Socio-demographic characteristics
A total of 2,924 questionnaires from Zhejiang and 3,355 from Shaanxi were completed and collected. Table 1 presents the socio-demographic characteristics of the children and parents. No significant differences were reported between respondents in the two provinces regarding their children's gender and age, or the parents' medical background as well as residential location (p > 0.05). Compared to parents in Zhejiang province, those in Shaanxi reported significantly lower levels of education and monthly household incomes (p < 0.001).

Parents' responses to children's illness and behaviors on antibiotic use
As Table 2 shows, more children in Shaanxi suffered from illnesses in the past month than those in Zhejiang. Among those sick children, the most commonly reported illnesses were a cold or sore throat  16.72%, p < 0.001). Table 3 presents the logistic regression analyses of the regional disparity in parents' behaviors on antibiotic use. After adjusting for socio-demographic factors,  Table 3 Logistic regression analyses of the regional disparity in parents' behaviors on antibiotic use.

Discussion
To the best of our knowledge, this is the first study investigating parental behaviors on antibiotic use for children to encompass both household and medical settings in developed and less developed Chinese provinces. Our main finding is that children in less developed provinces like Shaanxi are more likely to engage in antibiotic misuse than those in developed provinces like Zhejiang. Our study reveals that parents' reactions to children's illnesses were similar between regions at different development levels without taking antibiotics into account. However, living in a less developed province increased the likelihood of both parents and doctors misusing or unnecessarily prescribing antibiotics to treat children's illnesses. Parents in Shaanxi were also more likely to keep inhome antibiotics and administer prophylactic antibiotics to their children.
Self-medication is commonly practiced worldwide. People in both developed and developing areas selfmedicate as a supplement to, or replacement for formal medical consultations [21]. When parents temporarily feel their children's illness too minor for seeing a doctor, they may attempt to self-medicate their children at first, and turn to hospitals if their children do not get better as they expected [31]. Despite equal proportions of parents in Zhejiang and Shaanxi choosing selfmedication or going to a doctor, results were significantly different on the subject of antibiotic use. Consistent with previous research [11], our study found that parents in the less developed Shaanxi province were more likely to keep antibiotics at home and to give their children antibiotics as SMA or prophylaxis. These findings mirror a study in Europe, which showed storing in-home antibiotics and self-medicating with antimicrobial drugs were more prevalent in less developed countries [32]. The common trend can be attributed to cultural and socioeconomic factors that influence public perceptions of children and antibiotics as well as access to non-prescription antibiotics.
In China, preserving children's health is a strong priority in most Chinese families as a result of the previous national one-child policy. When children fall ill, parents strive for the best medicine available [21,33]. In less developed regions, antibiotics are often loosely regulated and erroneously seen as panaceas [34,35]. Perceived experiences of recovery after taking antibiotics for symptoms of viral infections, a phenomenon described as a self-fulfilling prophecy by some clinicians [36], make parents develop their preferences when choosing drugs [33]. More worrying is that parents tend to impose their own use of antibiotics on children [37]. In China, previous studies have found regional disparities in parental understandings of and attitudes towards antibiotic use. Fewer caregivers in the less developed city of Xi'an recognized antibiotics as prescription drugs, while many more agreed that they could self-medicate their children with antibiotics for minor illnesses compared to their counterparts in the more developed city of Shanghai [11]. Also, it has been reported in Jiangxi, a less developed province in central China, that nearly 80% of parents believed antibiotics could cure viral infections, and about 50% thought antibiotics would help shorten the duration of upper respiratory tract infections [12]. More importantly, many Chinese people used to simply regard symptoms of common infections as inflammation, and there is a widespread belief that antibiotics are the same as antiinflammatories [38]. Significantly more parents in Xi'an than in Shanghai agreed on this belief [11]. Above all, these misunderstandings potentially explain why parents in less developed provinces like Shaanxi had a greater preference for storing antibiotics at home and self-medicating children with antibiotics.
Easy access to non-prescription antibiotics in pharmacies is another contributing factor to antibiotic misuse in children [39]. Due to the lack of licensed pharmacists in China, salespeople at retail pharmacies can abet antibiotic misuse to make more profits [40]. Loose regulations on the sale of antibiotics indemnify their actions in less developed regions. In a Lebanese study, 69.2% of pharmacists in lower socioeconomic areas reported dispensing antibiotics to children compared to 27.1% in higher socioeconomic areas [41]. In China, significant regional disparities in nonprescription sales of antibiotics for the case of pediatric diarrhea were discovered, with a rate of 57.7% in Nanjing (a more developed Chinese city) versus 73.7% in Xi'an [18]. It was also found that when considering the source of non-prescription antibiotics used in SMA for children, more came from community pharmacies in Xi'an than in Shanghai, which might be attributable to better implementation of existing prohibitory regulations and better public awareness of appropriate antibiotic use in Shanghai [11]. Since Zhejiang province is also located in the Yangtze River Delta Economic Region, like Shanghai and Nanjing, it seemed reasonable to presume that parents in Zhejiang were less likely to use non-prescription antibiotics for children.
The regional disparities in receiving antibiotic prescriptions during pediatric consultations were also highlighted by our study. In contrast with a previous study which found a positive correlation between regional economic statuses and sales of antibiotics in hospitals [26], our study found a negative association between regional development levels and antibiotic prescription rates. We believe our results differ from theirs because eastern developed provinces in China have more adequate medical resources, including better availability of antibiotics than other underdeveloped sub-regions. Our study focused on the rate at which antibiotics were prescribed for children's common illnesses in outpatient settings, whereas theirs considered the aggregate antibiotic consumption of all patients. The different inclusion criteria as well as computing methods might have led to the differing results. Additionally, their findings were based on secondary data cited from earlier studies which had been conducted at different times rather than contemporaneous field research.
The Chinese government issued a National Essential Medicine List in 2009 including certain antibiotics and all governmental health institutions are required to sell the drugs in the List at cost only, which is also called 'zero markup drug policy' [42]. However, prescribing options can still be affected by personal relations with sales offers and pharmaceutical companies who may provide discounts to amenable doctors [41,43]. As a result, the antibiotic prescription rate could be influenced by doctors' financial incentives, especially in less developed regions [44]. Previous research has claimed that primary care doctors' salaries in western underdeveloped provinces are too low to live on since these doctors are usually self-employed and part-time, and they receive no government subsidies which are reserved as a bonus for medical practice [45]. Consequently, they have no choice but to rely on prescription profits to make ends meet, sometimes even prescribing multiple antibiotics for self-limited infections. Village doctors may even prescribe as a function of their patients' capacity to pay, that is, as many drugs as a patient can afford [33].
It has been asserted that parental requests for antibiotics encourage doctors to prescribe them [46]. In some less developed regions of China, doctors act as influential community members and have the freedom to tailor treatments to individual needs [33]. Our study found relatively low rates of parents requesting antibiotics in Zhejiang and Shaanxi. However, there were still high antibiotic prescribing rates in both provinces. This was particularly true in Shaanxi, where parents received many more antibiotic prescriptions than their counterparts in Zhejiang. Our research team previously conducted a survey to compare antibiotic misuse among university students in Zhejiang and the less developed province of Guizhou. We found a significantly higher proportion of students in Guizhou who reported asking for antibiotics and who received prescribed antibiotics when seeing a doctor in the past month [44].
Our results differ from the above study of university students. A possible explanation is that when parents decide to take their young children to a doctor, they may already have doubts as to the reliability of their selfdiagnosis and the safety of antibiotics, and may feel more reliant upon professional advice than university students who went to see the doctor for themselves. This kind of special concern for children appears to be universal. A study in Hong Kong showed that Chinese parents became more cautious when giving their children antibiotics because they had major concerns about children's tolerance to side effects [47]. Another study in the UK also found that mothers were more likely to accept non-antibiotic treatments for their children than for themselves [36]. As noted earlier, antibiotics are useless at treating most of the conditions reported in this study [6], thus, the prescribers of antibiotics may bear most of the responsibility for antibiotic misuse in medical contexts. A study in western China found that nearly half of prescriptions from village clinics contained antibiotics [48]. Some practitioners, especially the still practicing so-called "barefoot doctors" who were preliminarily trained to serve huge populations in less developed regions of China, might not be aware of guidelines regarding appropriate antibiotic use issued over the last few decades [49]. When antibiotic misuse is repeatedly advised by health practitioners, the practice spreads among the general public and becomes a social norm [3].
Our findings reveal the need for region-specific education programs for parents to correct their misconceptions about antibiotics and inform them of appropriate treatment strategies for children's common illnesses. New stronger regulations as well as more stringent enforcement of existing regulations are needed to eliminate non-prescription sales of antibiotics at community pharmacies. Training professional pharmacists and supervisors is particularly urgent in less developed areas. In addition, unnecessary antibiotic prescriptions should be restricted by standard guidelines and regulatory institutions. To that end, programs to train doctors on appropriate prescribing practices would be tremendously useful. Doctors could then participate in public campaigns to change social norms regarding antibiotic misuse. To decrease doctors' financial incentives, public investments in primary health care -specifically aiming to increase practitioner incomes in less developed areas -are required. If the link between doctors' and hospitals' incomes can be severed from drug sales, this will begin an unadulterated flow of prescriptions from hospitals to retail pharmacies, which may help reduce both overprescribing and non-prescription sales of antibiotics [39].
There are several limitations in this study. First, inpatient use (e.g. inpatient intravenous antibiotics) which requires no patient input was not included because our respondents were parents. Secondly, our findings were based on self-reported data, which may be distorted by report bias. We are nonetheless generally confident in our data given the short recall period of one month for questions related to children's illnesses. Finally, even though the sample sizes were large, our study only examined two Chinese provinces. More provinces at different levels of development must be investigated to further concretize the conclusions of this study.

Conclusion
There are regional disparities in antibiotic misuse among Chinese children. Children in less developed provinces face higher risks of antibiotic misuse both at home and when attending medical practitioners. To solve the problem, large-scale educational programs on appropriate antibiotic use will be required as well as more stringent enforcement of prohibitory regulations on non-prescription sales of antibiotics. Doctors and hospitals should be liberated from drug profits from unnecessary antibiotic prescriptions, and lead to change social norms of antibiotic misuse. Only when antibiotic misuse among children is addressed by considering the underlying impact of local environments and the interests of stakeholders, can the problem be effectively handled.
□Yes □No □Don't know 15. The more frequently people use antibiotics, the harder it is to cure the bacterial infections.