Dynamic Modeling of Complex Healthcare Systems Using Big Data to Describe and Visualize Healthcare Utilization

Healthcare costs have risen drastically in the past several decades with no commensurate improvement in health outcomes. The need to improve healthcare delivery is universally recognized and requires healthcare delivery systems to understand how patients utilize care within their system. In the case of acute conditions, a patient enters a facility where they are pushed and pulled through the facility, where they receive care that improves their health state. This acute care is characterized by two primary features: 1.) relatively short care duration on the order of a patient encounter, and 2.) a siloed operational focus on organs and organ systems. The field of industrial engineering and operations research dynamically models and mathematically quantifies healthcare operations of a facility to optimize patient throughput for wait times and cost. However, healthcare systems, in recent years, have had to shift from treating acute conditions to abating the unprecedented chronic disease burden. Much more so than acute care, chronic care is characterized by: 1.) a relatively prolonged duration requiring many patient encounters and 2.) multiple interacting and often competing health factors. Applying the mathematical modeling frameworks from classic industrial engineering and operations research across many facilities quickly becomes an intractable problem. On the other hand, health services research calculates-for patients with chronic disease-aggregated measures of healthcare utilization for types of services across types of facilities. Yet, the underlying descriptive analytical methods do not support dynamic modeling of healthcare utilization across healthcare delivery system facilities. Consequently, there is a need to apply a modeling framework that allows for dynamic modeling of healthcare utilization across healthcare delivery system facilities. In this paper, we briefly describe a healthcare delivery system architecture that supports modeling healthcare delivery systems at any scale and scope. We also present two examples, including a published fictitious case. We are unable to publish a real patient-level analysis due to a data use agreement that strictly prohibits publication of patient-level data.