Change in Pattern of Antibiotic Prescription in Community Health Centers in Shenzhen, China from 2013 to 2017

Background: Antimicrobial resistance is rapidly becoming a global crisis. Approximately 700,000 people die each year due to drug-resistant diseases infections and this number could rise to upwards of 10 million deaths each year by 2050 without actions. Many resistant infections are acquired in healthcare settings, where patients are often vulnerable and the potential for large-scale spread throughout a facility is significant. Method: Descriptive analyses were conducted to characterize World Health Organization (WHO) rational prescribing indicators sampled for one summer and one winter week of July in 2013 through 2017 at 18 CHCs. This research applied the Locally Weighted Scatter-plot Smoother (LOWESS) method to obtain the overall mean trends of the probability by month and by district. Finally, the study used logistic regression to model the probability of prescribing antibiotics to each URTI patient over time, with adjustment for the fixed effects of CHCs, calendar month, and individual patient characteristics.  Result: Among the 35,746 prescriptions from 18 sites with only URTI diagnosis, the overall proportion of prescriptions with antibiotics on CHC level decreased over time in both sampled seasons. However, the trends for July indicate a plateau in the decline from 2016 to 2017. On individual level, the adjusted logistic regression model estimated that the probability of prescribing antibiotics for URTI patients dropped from 0.355, 95% CI [0.304, 0.410] in 2013 to 0.200, 95% CI [0.165, 0.239] in 2017. Discussion: There has been significant decrease in antibiotic use for URTI patients. However, efforts are still needed to continue to monitor antibiotic use and make further progress to meet governments’ requirements and to prevent drug resistance of antibiotics. Funding: None to declare. Declaration of Interest: None to declare. Ethical Approval: The protocol of this study was reviewed and approved by the Sun Yat-sen University, School of Public Health Human Subjects Division.