Clinical treatment outcomes are the quality and cost targets that health-care providers aim to improve. Most existing outcome analysis focuses on a single disease or all diseases combined. Motivated by the success of molecular and phenotypic human disease networks (HDNs), this article develops a clinical treatment network that describes the interconnections among diseases in terms of inpatient length of stay (LOS) and readmission. Here one node represents one disease, and two nodes are linked with an edge if their LOS and number of readmissions are conditionally dependent. This is the very first HDN that jointly analyzes multiple clinical treatment outcomes at the pan-disease level. To accommodate the unique data characteristics, we propose a modeling approach based on two-part generalized linear models and estimation based on penalized integrative analysis. Analysis is conducted on the Medicare inpatient data of 100,000 randomly selected subjects for the period of January 2010 to December 2018. The resulted network has 1008 edges for 106 nodes. We analyze key network properties including connectivity, module/hub, and temporal variation. The findings are biomedically sensible. For example, high connectivity and hub conditions, such as disorders of lipid metabolism and essential hypertension, are identified. There are also findings that are less/not investigated in the literature. Overall, this study can provide additional insight into diseases' properties and their interconnections and assist more efficient disease management and health-care resources allocation.

中文翻译：

---

老年人类疾病临床诊疗网：医保住院住院和再入院数据分析

临床治疗结果是医疗保健提供者旨在改善的质量和成本目标。大多数现有的结果分析都侧重于单一疾病或所有疾病的结合。受到分子和表型人类疾病网络 (HDN) 成功的推动，本文开发了一个临床治疗网络，该网络描述了疾病之间在住院时间 (LOS) 和再入院方面的相互联系。这里一个节点代表一种疾病，如果两个节点的 LOS 和再入院次数有条件依赖，则两个节点与一条边相连。这是第一个在泛疾病层面联合分析多种临床治疗结果的HDN。为了适应独特的数据特征，我们提出了一种基于两部分广义线性模型的建模方法和基于惩罚综合分析的估计。对 2010 年 1 月至 2018 年 12 月期间随机选择的 100,000 名受试者的 Medicare 住院数据进行分析。结果网络具有 106 个节点的 1008 条边。我们分析关键网络属性，包括连接性、模块/集线器和时间变化。这些发现在生物医学上是明智的。例如，确定了高连通性和枢纽条件，例如脂质代谢紊乱和原发性高血压。还有一些研究结果在文献中较少/未进行调查。总的来说，这项研究可以提供对疾病特性及其相互联系的更多见解，并有助于更有效的疾病管理和医疗资源分配。