Abstract:
The aim of this study was to develop a logistic modeling concept to improve understanding of the relationship between antibiotic use thresholds and the incidence of resistant pathogens. A combined approach of nonlinear modeling and logistic regression, named threshold logistic, was used to identify thresholds and risk scores in hospital-level antibiotic use associated with hospital-level incidence rates of extended-spectrum β-lactamase (ESBL)-producing Escherichia coli (E. coli). Threshold logistic models identified thresholds for fluoroquinolones (61.1 DDD/1000 occupied bed days (OBD)) and third-generation cephalosporins (9.2 DDD/1000 OBD) to control hospital ESBL-producing E. coli incidence. The 60th percentile of ESBL-producing E. coli was determined as the cutoff for defining high incidence rates. Threshold logistic analysis showed that for every one-unit increase in fluoroquinolones and third-generation cephalosporins above 61.1 and 9.2 DDD/1000 OBD levels, the average odds of the ESBL-producing E. coli incidence rate being ≥60th percentile of historical levels increased by 4.5% and 12%, respectively. Threshold logistic models estimated the risk scores of exceeding the 60th percentile of a historical ESBL-producing E. coli incidence rate. Threshold logistic models can help hospitals in defining critical levels of antibiotic use and resistant pathogen incidence and provide targets for antibiotic consumption and a near real-time performance monitoring feedback system.
摘要:
这项研究的目的是开发逻辑建模概念,以提高对抗生素使用阈值与耐药病原体发生率之间关系的理解。非线性建模和逻辑回归的组合方法,命名为阈值逻辑,用于确定医院级抗生素使用与产超广谱β-内酰胺酶(ESBL)大肠杆菌的医院级发病率相关的阈值和风险评分(E.大肠杆菌)。阈值逻辑模型确定了氟喹诺酮类药物(61.1DDD/1000卧床天数(OBD))和第三代头孢菌素(9.2DDD/1000OBD)的阈值,以控制医院产生ESBL的大肠杆菌发病率。将产生ESBL的大肠杆菌的第60百分位数确定为用于定义高发病率的截止值。阈值逻辑分析表明,氟喹诺酮类药物和第三代头孢菌素类药物在61.1和9.2DDD/1000OBD水平以上每增加一个单位,产ESBL大肠杆菌发病率≥历史水平的60百分位数的平均几率增加了4.5%和12%,分别。阈值逻辑模型估计超过历史产生ESBL的大肠杆菌发病率的60百分位数的风险评分。阈值逻辑模型可以帮助医院确定抗生素使用和耐药病原体发生率的关键水平,并提供抗生素消费目标和近乎实时的性能监测反馈系统。