Study on regression model of forecast for Acinetobacter baumannii resistance to anti-bacterial agents
10.3760/cma.j.issn.1008-5734.2017.02.004
- VernacularTitle:预测鲍曼不动杆菌对抗菌药物耐药率的回归模型研究
- Author:
Zhiyun YANG
1
;
Shanshan LI
1
;
Jie CHEN
1
;
Xiao CHEN
1
Author Information
1. 中山大学附属第一医院药学部, 广州,510080
- Publication Type:Journal Article
- Keywords:
Acinetobacter baumannii;
Drug resistance;
Anti-bacterial agents
- From:
Adverse Drug Reactions Journal
2017;19(2):96-102
- CountryChina
- Language:Chinese
-
Abstract:
Objective To establish a regression model for predicting the resistance rate of Acinetobacter baumannii (Ab) to anti-bacterial agents.Methods The monitoring data of resistance rate of 1 249 Ab strains to 10 frequently used anti-bacterial agents and drug consumption data [the main indicators for the frequency of medication (DDDs) and antimicrobial use intensity (AUD)] of these agents from January 2014 to December 2016 in the First Affiliated Hospital of Sun Yat-sen University were collected.Pearson correlation analysis was performed to analyze the correlation between Ab resistance and AUD.A linear regression model was established for the anti-bacterial agents with correlation between drug resistance and AUD.Then the regression model was used to predict the resistance rate of Ab in January 2017 and February 2017.Results During the 3 years from 2014 to 2016, the resistance rate of Ab to tigecycline was the lowest (<7% for the 3 years), the resistance rate of Ab to ciprofloxacin was the highest (the average resistance rate was 82.1% for 3 years), there were 3 kinds of drugs whose average resistance rates were <70% and 7 kinds of drugs whose average resistance rates were >70% during the 3 years.The AUD of cefoperazone sodium sulbactam sodium, levofloxacin, tigecycline, panipenem/betamipron, and cefepime were increased.Pearson correlation analysis showed that the resistance rates of Ab to cefoperazone sodium sulbactam sodium and levofloxacin were correlated with their AUD (r=0.681, P=0.015;r=0.694, P=0.012).The data related to Ab resistance rates and AUD of cefoperazone sodium sulbactam sodium and levofloxacin were brought into the one variable linear regression equation.The variance analysis showed that the 2 regression equations fitted well(F=8.634, P=0.015;F=9.285, P=0.012).The AUD of cefoperazone sodium sulbactam sodium and levofloxacin in January and February, 2017 were respectively brought into the regression equation and the resistance rates of Ab were predicted.The results showed that the actual resistance rates were evenly distributed around the predictive values and within 95% of the predicted range.Conclusion In our hospital, AUD of cefoperazone sodium sulbactam sodium and levofloxacin were related to resistance rates of Ab and the linear regression model could predict the resistance rate of Ab by AUD.
