Use and Prescription of Antibiotics in Primary Health Care Settings in China

Importance  Appropriate antibiotic use is a key strategy to control antibacterial resistance. The first step in achieving this is to identify the major problems in antibiotic prescription in health care facilities, especially in primary health care settings, which is where most patients receive medical care.

Objective  To identify current patterns of antibiotic use and explore the reasons for inappropriate prescription in primary health care settings in China.

Design, Setting, and Participants  A total of 48 primary health care facilities in China were randomly selected from 6 provinces at various levels of economic development. Data for the years 2009 through 2011 from 39 qualifying facilities (23 city and 16 rural primary health care centers) were analyzed retrospectively. The study sample consisted of prescription records for 7311 outpatient visits and 2888 inpatient hospitalizations.

Main Outcomes and Measures  General health center information, drug usage, disease diagnoses, and antibiotic use by outpatients and inpatients were surveyed. Cases of inappropriate antibiotic prescription were identified.

Results  Most staff in the primary health care facilities had less than a college degree, and the medical staff consisted primarily of physician assistants, assistant pharmacists, nurses, and nursing assistants. The median (range) governmental contribution to each facility was 34.0% (3.6%-92.5%) of total revenue. The facilities prescribed a median (range) of 28 (8-111) types of antibiotics, including 34 (10-115) individual agents. Antibiotics were included in 52.9% of the outpatient visit prescription records: of these, only 39.4% were prescribed properly. Of the inpatients, 77.5% received antibiotic therapy: of these, only 24.6% were prescribed properly. Antibiotics were prescribed for 78.0% of colds and 93.5% of cases of acute bronchitis. Of the antibiotic prescriptions, 28.0% contained cephalosporins and 15.7% fluoroquinolones. A total of 55.0% of the antibiotic prescriptions were for antibiotic combination therapy with 2 or more agents. In nonsurgical inpatients in cities, the mean (SD) duration of antibiotic therapy was 10.1 (7.8) days. Of the surgical patients, 98.0% received antibiotics, with 63.8% of these prescriptions for prophylaxis.

Conclusions and Relevance  Antibiotics are frequently prescribed in Chinese primary health care facilities, and a large proportion of these prescriptions are inappropriate. Frequent and inappropriate use of antibiotics in primary health care settings in China is a serious problem that likely contributes to antimicrobial resistance worldwide.

Antibiotic resistance has become a worldwide challenge to public health. In 2011, a national surveillance program in Chinese tertiary hospitals found that the prevalence of methicillin-resistant Staphylococcus aureus was 50.5%; extended-spectrum β-lactamase–producing Escherichia coli, 71.2%; carbapenem-resistant Pseudomonas aeruginosa, 23.4%; and carbapenem-resistant Acinetobacter baumannii, 56.8%.¹ Antibiotic resistance results in a lack of effective drugs to fight infections, as well as increased health care costs, prolonged hospitalization, and substantially increased social and economic burden. “Combat drug resistance: no action today means no cure tomorrow”² was a theme of World Health Day 2011. The World Health Organization (WHO) recommends proper antibiotic use, drug resistance monitoring, and research and development of new methods to diagnose and treat infectious diseases as key strategies to control antibiotic resistance.²

At present, all hospitals in China are ranked as tertiary, secondary, and primary hospitals. Tertiary hospitals are large medical centers that are located in big cities, secondary hospitals are county hospitals, and primary health care facilities are community hospitals or clinics that are typically small and provide a limited range of basic medical services. All hospitals in China have to fund themselves by drug sales and service provision and thus purchase, store, and dispense or administer drugs.³ Antibiotic use in secondary and tertiary Chinese hospitals has been surveyed previously, drug management programs have been implemented to promote the proper use of antibiotics, and appropriate monitoring systems have been established.^4,5 However, information on the use and prescription of antibiotics in primary health care facilities is lacking. This is important because these facilities provide medical care to the majority of Chinese patients: the Chinese National Health and Family Planning Commission estimated that there were 5.84 billion medical visits in China in 2010 and that 3.79 billion (64.9%) of these were in primary health care facilities.⁶ Hence, an understanding of antibiotic use in primary health care facilities and promotion of proper antibiotic use, with clearly stated objectives, is needed to contain bacterial resistance.

To determine how antibiotics are used in Chinese primary health care facilities, we studied those facilities in several regions of China with differing levels of economic development between 2009 and 2011. We sought to identify frequency of use and patterns of inappropriate antibiotic prescription. We also compared antibiotic use among urban and rural health care facilities, outpatient and inpatient settings, and surgical and nonsurgical patients.

Study approval was obtained from the Research Ethics Board of Beijing Shijitan Hospital. They waived the need for patient consent because there was no contact with patients and patient anonymity was assured.

The primary health care facilities that were evaluated were city community health care service centers (CCHSCs) and rural township health care centers (RTHCs) in underdeveloped inland areas (Jilin and Henan Province), developed coastal areas (Shandong, Jiangsu, and Zhejiang Province), and a large city (Beijing). A total of 24 CCHSCs (2 from each provincial capital city, 2 from a second city in each province, and 4 from Beijing) and 24 RTHCs (4 from each province and Beijing) were included.

Outpatient and inpatient records from the selected health care institutions were randomly sampled. Health care facility characteristics, basic financial information (including the total drug sales volume), and the types and amounts of antibiotics that the hospital stocked were surveyed. The antibiotics were classified according to the WHO Anatomical Therapeutic Chemical classification system.⁷ The antibiotic survey was performed according to WHO recommendations for such studies; the major indicators that we selected were also in accordance with WHO recommendations.⁸

Patient records from 1 month in the spring and 1 month in the fall from each of 2 consecutive years from 2009 through 2011 (4 months in total) were selected by a completely random method and examined. In China, prescribers typically write prescriptions for more than 1 drug on a single prescription form; the term “prescription” hereafter refers to all drugs prescribed for 1 patient during 1 visit or 1 hospitalization. In each primary health care facility, for each month, 50 ambulatory-visit prescriptions were randomly selected. Thus, 200 ambulatory-visit prescriptions were examined in total for each facility. Inpatient admission records were also randomly selected from each facility for each month: the target number was 50 per month; if there were fewer than 50 inpatient admissions in any month, all records for the month were included. Diagnosis, proportion of prescriptions that contained antibiotics, appropriateness of antibiotic use, medical expenses, types and cost of the antibiotics that were prescribed, and other information was collected.

Two investigators were assigned to each medical facility; all investigators received the same training prior to the survey. During the survey, one investigator filled out the investigational forms and the other reviewed the data. All data were checked for completeness and logicality. Errors in the records were noted to determine the overall authenticity of the data in each institution. A facility was excluded if more than 25% of the medical records failed quality control. The data from the quality-controlled forms were entered into a database for analysis.

According to the standards of the “Principles of Clinical Use of Antibiotics” issued by the Chinese Ministry of Health following international concepts in 2004,⁹ which was modified from internationally endorsed criteria (indicated in parentheses),^10,11 an antibiotic prescription was considered to be proper (correct decision) if it contained the standard treatment regimen and duration that was indicated for the patient’s clinical infection or prophylaxis. Antibiotic use was considered to be partially proper when the clinical condition of the patient justified use of antibiotics for treatment or prophylaxis but the treatment regimen (incorrect choice) or its duration (incorrect use) was incorrect. Antibiotic use was considered improper when the clinical condition of the patient did not justify use of antibiotics for either treatment or prophylaxis (incorrect decision and data insufficient) (see eAppendix in the Supplement).

After double data entry and confirmation that the data were correct, the database was locked and could not be further edited. SSPS software, version 20.0 (IBM), was used for data analysis. Continuous variables were expressed as median and range. Categorical data were expressed as proportions. Sales and income were converted from the Chinese currency into US dollars by using an exchange rate of 6.62 yuan to 1 US dollar.

A total of 48 primary health care facilities participated in the survey. Of these, 6 with incomplete raw data and 3 with unqualified data were excluded. The data from the remaining 39 primary health care facilities were included in the analysis (8 in Jilin, 6 in Henan, 4 in Shandong, 7 in Jiangsu, 8 in Zhejiang, and 6 in Beijing). Of these 39 facilities, 22 had inpatient units. Twenty-three facilities were CCHSCs and 16 were RTHCs. Of the 7800 records of prescriptions from outpatient visits that were evaluated, 7311 met quality-control standards (3819 from CCHSCs and 3492 from RTHCs). The hospitalized patient records contained 3120 prescriptions, of which 2888 met quality-control standards. Of these, 1931 and 957 were of surgical and nonsurgical patients, respectively, and 995 were from CCHSCs and 1893 from RTHCs. Thus, in total, 10 199 quality-controlled prescriptions were evaluated in the study: 4814 from CCHSCs and 5385 from RTHCs. Of the 10 199 prescriptions, 7743 (75.9%) contained antibiotics.

The primary health care facilities varied widely in size, patient population, local government investment, and annual revenue (Table 1). Most staff had less than a college degree; the medical staff consisted primarily of physician assistants, assistant pharmacists, nurses, and nursing assistants (data not shown). The primary health care facilities mainly managed noninfectious chronic diseases. The 5 most common clinical conditions of the outpatients were colds (20.6%), hypertension (9.5%), acute bronchitis (7.2%), gastroenteritis (5.4%), and diabetes mellitus (4.6%). The 5 most common clinical conditions of the inpatients were coronary heart disease (9.2%), hypertension (8.7%), stroke (7.5%), appendicitis (6.6%), and anorectal disorders (6.4%). The most frequently performed surgical procedures were appendectomy, hemorrhoid removal, and trauma treatment.

Antibiotics accounted for a varying proportion of the total drug sales of each facility: the median was 22.8%, and the maximum was 63.3% (Table 1). Analysis of the 7743 antibiotic prescriptions indicates that all types of antibiotics except glycopeptides and linezolid were used. The most commonly used antibiotics were cephalosporins, fluoroquinolones, penicillins, and nitroimidazoles, which accounted for 70.3% of all antibiotic prescriptions (Table 2). The 5 most commonly used agents were levofloxacin, amoxicillin, azithromycin, benzylpenicillin, and cefuroxime. Of all antibiotic prescriptions, 8.0% were combinations of a β-lactamase inhibitor plus cephalosporin or penicillin, such as cefoperazone-sulbactam, amoxicillin-clavulanate, piperacillin-tazobactam, and cefotaxime-sulbactam.

Of 7311 outpatient prescriptions, 52.9% contained antibiotics, of which 39.4% were proper. Of 2888 inpatient prescriptions,77.5% contained antibiotics, of which 24.6% were proper. Of the 957 nonsurgical inpatient prescriptions, 67.4% contained antibiotics and 28.2% of these were proper (Table 3). Of nonsurgical inpatient antibiotic prescriptions, 55.0% and 35.0% contained combinations of 2 or more antibiotics in CCHSCs and RTHCs, respectively, and the mean (SD) prescribed durations of antibiotic use were 6.7 (6.1) and 10.1 (7.8) days, respectively.

Of the 1931 surgical patient prescriptions, 98.0% contained antibiotics and only 18.0% were proper (Table 3). The majority of surgical patient antibiotic prescriptions (63.8%) were for prophylaxis. In all cases, more than 2 antibiotics were prescribed. The mean prescribed duration was 5.4 days. Most of the nitroimidazole prescriptions in the study were for surgical patients.

There was a high rate of prescription of antibiotics for colds, pharyngitis, acute bronchitis, and urinary tract infection. Except for urinary tract infection, most of these prescriptions were not proper (Figure).

The antibiotic use in primary health care settings varied among the different regions. The outpatient antibiotic prescription rate was lowest in Beijing and highest in Jilin. The highest antibiotic prescription rate for inpatients was in Jiangsu, and the rate of improper use was also highest there. The different regions were generally similar in terms of their most commonly used antibiotics, but more penicillins and fluoroquinolones were used in Shandong and Jilin than in other regions (Table 4).

Antibiotics are the most frequently used medicine in Chinese medical facilities, and antibiotics account for approximately 20% of all drug sales in general hospitals.¹³ In this survey of Chinese primary health care facilities, all types of antibiotics except glycopeptides and linezolid were prescribed, and the percentage of antibiotic use was very high in both outpatient and inpatient settings. Antibiotic use without a clear indication was common, and many errors in antibiotic usage were made. Prescribers tended to favor new and broad-spectrum antibiotics, as well as treatment with combinations of multiple antibiotics, prolonged antibiotic use, and intravenous antibiotic administration. These antibiotic prescription patterns were similar across the different regions of China and in the CCHSCs and RTHCs.

Analysis of data from the Chinese Nosocomial Infection Control Surveillance Networks reveals that, compared with large secondary and tertiary hospitals, excessive and improper use of antibiotics is more severe in primary health care facilities, although the types of antibiotics that were prescribed in both settings were similar: in 2010, nearly half of the patients in tertiary hospitals received antibiotics (49.6%), and combination therapy was used in 32.0% of the hospitalized patients.¹⁴ Antibiotics were included in 14.9% to 21.8% of outpatient prescriptions, and 16% to 30% of prescriptions were for antibiotic combinations.^15,16

Compared with other countries, the rate of antibiotic use in primary health care facilities in China is relatively high, although it is comparable to that in India (78%), Indonesia (84%), South Korea, and other Asian countries.^17-18 Antibiotic use in China exceeds that in the United States (63.7%), Sweden (30%), and other developed countries.^19,20 In China and India, the most commonly used antibiotics are cephalosporins, fluoroquinolones, broad-spectrum penicillins, and macrolides. However, in the majority of European countries, the most commonly used antibiotics for outpatients are benzylpenicillin, tetracyclines, and sulfonamides.^17,20

The improper use of antibiotics in primary health care facilities in China could be associated with several factors. First, it is likely that inadequate government investment in primary health care facilities forces these facilities to rely, at least partially, on drug sales to operate: government investment accounted for 34.0% of the facility budgets in this study, whereas income from drug sales accounted for 39.3% of the total income. Antibiotic sales accounted for 22.8% of revenues from all drug sales. In some regions, personnel costs were covered by government funds and profits were not made on the drug sales: however, these changes were not associated with improved antibiotic use. This may be because the medical staff salaries are low, which motivates the personnel to seek extra income by providing extra services and selling pharmaceutical products. One of the key tasks in the ongoing health care reforms in China is to eliminate the financial incentives to facilities and their staff that are associated with drug sales.^13,21,22 Second, lack of expertise of medical staff may contribute to the improper use of antibiotics. In both the CCHSCs and RTHCs, the educational attainment and skills of the medical staff were relatively low. Most prescribers had a professional rank below that of a physician assistant. In addition, medical resources in China are concentrated in secondary and tertiary hospitals in major cities; this means that under current practices, the training and education opportunities for primary health care staff are limited.²³ Lack of training may explain why we consistently found that prescribers tended to use broad-spectrum antibiotics, recommended combination therapy with multiple antibiotics that have similar pharmacological characteristics, used excessive antibiotic prophylaxis for surgical patients, and prescribed overly long treatment durations. Third, management and supervision of primary health care facilities by health care administrative authorities could be strengthened. The Chinese National Health and Family Planning Commission recently improved the management of antibiotic use in large hospitals by setting performance targets, forming task forces, and carrying out inspections,^2,24 but similar actions have not been taken in primary health care facilities in China.

Our study had several limitations. First, we were unable to measure daily doses of antibiotics^7,8 because a reliable patient registration system has not been established in China. However, the antibiotic prescription rates are accurate. Second, although we included underdeveloped and developed areas and central cities in north and east China, we did not consider western regions because of difficulties in data acquisition and questionable data integrity. However, published data on antibiotic use in general hospitals and rural clinics in western China suggest that the results would not be different from those in the regions studied in the present study.^13,25

Substantial improper antibiotic use occurs in primary health care settings in China, probably because of the need to generate income from drug sales, inadequate educational and skill levels of professional staff, and weak oversight. To overcome these shortcomings, financial support of primary health care must be increased and drug sales should be separated from facility operation. Furthermore, medical professional training should be improved, and systems to ensure the proper use of antibiotics should be stressed.^2,5

Corresponding Author: Yonghong Xiao, MD, PhD, Collaborative Innovation Center for Diagnosis and Treatment of Infectious Diseases, State Key Laboratory for Diagnosis and Treatment of Infectious Diseases, the First Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou 310003, China (xiao-yonghong@163.com).

Accepted for Publication: June 20, 2014.

Published Online: October 6, 2014. doi:10.1001/jamainternmed.2014.5214.

Author Contributions: Drs Xiao and J. Wang had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: J. Wang, Xiao.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: J. Wang, Xiao.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: J. Wang, Zheng.

Obtained funding: Xiao.

Administrative, technical, or material support: J. Wang, P. Wang, X. Wang, Xiao.

Study supervision: Xiao.

Conflict of Interest Disclosures: None reported.

Funding/Support: Dr Xiao was supported by a grant (2013ZX10004217) from National Mega Project of Science and Technology, China.

Role of the Sponsor: The funder had no role in the design and conduct of the study; the collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: The authors are grateful to all members of the investigational team who collected the data. We also thank all investigated institutions for providing information and assistance during the study.