Health information systems (HIS), especially electronic health record (EHR) systems, can significantly improve efficiency and quality of healthcare because it enables the employees of Healthcare Organizations (HCOs) to coordinate and collaborate more effectively and on a large scale. At the same time, healthcare environments are composed of highly diverse and dynamic workflows and, as modern EHR systems increase in size and scope, they may exacerbate the complexity of HCOs. If organizational complexity is not managed appropriately, it could limit the benefits of EHR systems and significantly contribute to negative effects, such as longer waiting times for care, replication of diagnostics, and medical errors. We hypothesize that patterns of collaboration can assist in managing complexity and facilitating patient management and that such patterns can be discovered by data mining on the utilization of EHR systems. This work is timely because EHR adoption has grown significantly over the past several years and HCO employees are increasingly using such systems to document patient status and communicate with other providers. The quantity and detail of such data provides an opportunity for big data mining techniques to learn patterns of care that are not explicitly documented. We propose to develop methods to learn patterns of collaboration through the utilization of EHR systems to determine how the management of care providers can be optimized. This will be accomplished through three specific aims: (1) Discover effective teams of care providers tailored to specific types of disease through the analysis of utilization data. Based on such teams, HCO will be able to manage patients more efficiently by prepping team members in a more timely manner. (2) Learn dependencies between care providers, which will be critical for resource allocation and management of care teams. (3) Model disease-specific treatment workflows to assess which sequences of events lead to the most efficient and effective outcomes for patients. In doing so, this project aims to reduce the length of patients' hospital stay and, ultimately help patients (an HCOs) reduce costs.
Health information systems (HIS) facilitate collaborative environments, through which healthcare teams can improve their efficiency and patients can receive more effective treatment. Yet, at the same time, complexity in the interactions of care providers, in combination with the dynamic nature of healthcare, could limit the realization of such benefits if coordination is not appropriately handled (e.g., longer waiting times and replication in diagnostics) and care is not personalized to the patient. In this research, we aim to develop data mining strategies to enable the optimization of care providers management by learn patterns of collaboration through the utilization of HIS; specifically, the goals of this project are to 1) model disease-specific treatment communities, 2) learn dependency patterns between care providers, and 3) infer disease-specific coordinate sequence patterns.
|Chen, You; Patel, Mayur B; McNaughton, Candace D et al. (2018) Interaction patterns of trauma providers are associated with length of stay. J Am Med Inform Assoc 25:790-799|
|Li, Thomas; Gao, Cheng; Yan, Chao et al. (2018) Predicting Neonatal Encephalopathy From Maternal Data in Electronic Medical Records. AMIA Jt Summits Transl Sci Proc 2017:359-368|
|Yin, Wen; Gao, Cheng; Xu, Yaomin et al. (2018) Learning Opportunities for Drug Repositioning via GWAS and PheWAS Findings. AMIA Jt Summits Transl Sci Proc 2017:237-246|
|Chen, You; Kho, Abel N; Liebovitz, David et al. (2018) Learning bundled care opportunities from electronic medical records. J Biomed Inform 77:1-10|
|Gao, Cheng; Kho, Abel N; Ivory, Catherine et al. (2017) Predicting Length of Stay for Obstetric Patients via Electronic Medical Records. Stud Health Technol Inform 245:1019-1023|
|Chen, You; Lorenzi, Nancy M; Sandberg, Warren S et al. (2017) Identifying collaborative care teams through electronic medical record utilization patterns. J Am Med Inform Assoc 24:e111-e120|
|Zheng, Tao; Xie, Wei; Xu, Liling et al. (2017) A machine learning-based framework to identify type 2 diabetes through electronic health records. Int J Med Inform 97:120-127|
|Yan, Chao; Chen, You; Li, Bo et al. (2016) Learning Clinical Workflows to Identify Subgroups of Heart Failure Patients. AMIA Annu Symp Proc 2016:1248-1257|