Prediction Methods for Antimicrobial Resistance Trends in China.

Our objective in this study was to develop robust and accurate prediction models for assessing the trends of antimicrobial resistance in China. Data for our study were derived from the China Antimicrobial Resistance Surveillance System, spanning the period from 2014 to 2021. Utilizing these data, we constructed prediction models by GM (1,1), support vector machine, polynomial fitting, and time series. Of all the antibiotics investigated in this study, the resistance rates of carbapenem-resistant Klebsiella pneumoniae and erythromycin-resistant Streptococcus pneumoniae exhibited an upward trend, while resistance rates of the remaining pathogens demonstrated a decreasing trend. The GM (1,1) model demonstrated superior robustness and accuracy among these four models. While a decline in resistance was observed in nine pathogens over time, the antimicrobial-resistant rate of erythromycin-resistant streptococcus pneumoniae and Carbapenems-resistant Pseudomonas aeruginosa was noted to increase, potentially due to the overuse of macrolides in China. These findings underscore the necessity for stricter antibiotic stewardship to counter the risk of widespread resistance. Furthermore, studies from the European Union have reported an escalation in drug resistance relative to pre-pandemic levels, underlining the pandemic's impact on the battle against bacterial resistance.

Supplementary information: The online version contains supplementary material available at 10.1007/s12088-024-01442-z.