Objectives

Identifying characteristics associated with the appropriateness of antibiotic prescriptions is useful to guide antibiotic stewardship interventions. Proxy indicators estimating the appropriateness of antibiotic prescriptions at the general practitioner (GP) level have recently been validated. Our objectives were to identify (a) clusters of GPs according to their appropriateness score based on these proxy indicator results, and (b) GPs', patients' and practices' characteristics associated with inappropriate prescriptions.

Methods

We conducted a cross-sectional observational study analysing antibiotics prescribed by GPs in one large French region in 2019, using the Health Insurance databases. We identified clusters of GPs according to their appropriateness score calculated from ten proxy indicators' results. We then analysed the association between the clusters with more inappropriate practices compared with the one with less inappropriate practices, and GPs', patients', and practices' characteristics. We performed bivariate and multivariable analyses using logistic polytomous regressions.

Results

We included 4819 GPs who were grouped into three clusters. GPs who belong to the clusters with more inappropriate practices were more likely to practice in certain geographical area, to be male, not to have a particular medical practice, to be practicing for longer, to have more patients and consultations, to have a higher proportion of elderly patients, and to prescribe more drugs, more antibiotics and a higher proportion of broad-spectrum antibiotics.

Conclusion

We identified clusters of practice as well as factors associated with the appropriateness of antibiotic prescriptions, using routinely collected data. This might help to guide antibiotic stewardship interventions.

中文翻译：

---

与法国一般实践中处方抗生素的适当性相关的因素：使用报销数据库的横断面研究

目标

识别与抗生素处方适当性相关的特征有助于指导抗生素管理干预。评估全科医生 (GP) 水平抗生素处方适宜性的代理指标最近已得到验证。我们的目标是（a）根据这些代理指标结果的适当性评分来识别 GP 集群，以及（b）与不适当处方相关的 GP、患者和实践特征。

方法

我们使用健康保险数据库进行了一项横断面观察研究，分析了 2019 年在法国一个大地区由全科医生开出的抗生素。我们根据从十个代理指标的结果计算出的适当性得分来确定 GP 集群。然后，我们分析了不当做法较多的集群与不当做法较少的集群之间的关联，以及全科医生、患者和实践的特征。我们使用逻辑多变量回归进行了双变量和多变量分析。

结果

我们纳入了 4819 名 GP，他们被分为三个集群。属于具有更多不适当实践的集群的全科医生更有可能在某些地理区域执业、男性、没有特定的医疗实践、执业时间更长、有更多的患者和咨询、有更高的比例老年患者，开出更多的药物、更多的抗生素和更高比例的广谱抗生素。

结论

我们使用常规收集的数据确定了实践集群以及与抗生素处方的适当性相关的因素。这可能有助于指导抗生素管理干预。