A hybrid network--deterministic model for simulation of multiresistant pathogen spread in a healthcare system is presented. The model accounts for two paths of pathogen transmission between the healthcare facilities: inter-hospital patient transfers (direct transfers) and readmission of colonized patients (indirect transfers). In the latter case, the patients may be readmitted to the same facility or to a different one. Intra-hospital pathogen transmission is governed by a SIS model expressed by a system of ordinary differential equations. Using a network model created for a Lower Saxony region (Germany), we showed that the proposed model reproduces the basic properties of healthcare-associated pathogen spread. Moreover, it shows the important contribution of the readmission of colonized patients on the prevalence of individual hospitals as well as of overall healthcare system: it can increase the overall prevalence by the factor of 4 as compared to inter-hospital transfers only. The final prevalence in individual healthcare facilities was shown to depend on average length of stay by a non-linear concave function. Finally, we demonstrated that the network parameters of the model may be derived from administrative admission/discharge records. In particular, they are sufficient to obtain inter-hospital transfer probabilities, and to express the patients' transfer as a Markov process.

中文翻译：

---

对医疗保健网络中的病原体传播进行建模：间接患者移动

提出了一种混合网络-确定性模型，用于模拟医疗保健系统中的多重耐药病原体传播。该模型考虑了医疗机构之间病原体传播的两条路径：医院间患者转移（直接转移）和定植患者的再入院（间接转移）。在后一种情况下，患者可能会重新入住同一设施或不同的设施。医院内病原体传播由常微分方程系统表示的 SIS 模型控制。使用为下萨克森州（德国）创建的网络模型，我们表明所提出的模型再现了与医疗保健相关的病原体传播的基本特性。而且，它显示了定植患者的重新入院对个别医院以及整个医疗保健系统的流行率的重要贡献：与仅医院间转移相比，它可以将总体流行率提高 4 倍。单个医疗机构的最终患病率显示取决于非线性凹函数的平均住院时间。最后，我们证明了模型的网络参数可能来自行政入院/出院记录。特别是，它们足以获得医院间转移概率，并将患者转移表示为马尔可夫过程。单个医疗机构的最终患病率显示取决于非线性凹函数的平均住院时间。最后，我们证明了模型的网络参数可能来自行政入院/出院记录。特别是，它们足以获得医院间转移概率，并将患者转移表示为马尔可夫过程。单个医疗机构的最终患病率显示取决于非线性凹函数的平均住院时间。最后，我们证明了模型的网络参数可以从行政入院/出院记录中导出。特别是，它们足以获得医院间转移概率，并将患者转移表示为马尔可夫过程。