# Simulating feedback mechanisms in patient flow and return visits in an emergency department

## TL;DR

This paper develops a system dynamics simulation model to analyze how local protocols affect emergency department crowding and return visits, highlighting system vulnerabilities and aiding hospital management decisions.

## Contribution

It introduces a novel SD model capturing complex interactions between capacity, patient flow, and return visits in EDs, enabling virtual scenario testing for policy optimization.

## Key findings

- High leverage in bed assignment and transfer times.
- System vulnerability due to physical constraints and expedited flows.
- Protocols can balance patient flow and return rates.

## Abstract

Emergency department (ED) crowding has been an increasing problem worldwide. Prior research has identified factors that contribute to ED crowding. However, the relationships between these remain incompletely understood. This study's objective was to analyse the effects of initiating a local protocol to alleviate crowding situations at the expense of increasing returning patients through the development of a system dynamics (SD) simulation model. The SD study is from an academic care hospital in Boston, MA. Data sources include direct observations, semi-structured interviews, archival data from October 2013, and peer-reviewed literature from the domains of emergency medicine and management science. The SD model shows interrelations between inpatient capacity restraints and return visits due to potential premature discharges. The model reflects the vulnerability of the ED system when exposed to unpredicted increases in demand. Default trigger values for the protocol are tested to determine a balance between increased patient flows and the number of returning patients. Baseline simulation runs for generic variables assessment showed high leverage potential in bed assignment- and transfer times. A thorough understanding of the complex non-linear behaviour of causes and effects of ED crowding is enabled through the use of SD. The vulnerability of the system lies in the crucial interaction between the physical constraints and the expedited patient flows through protocol activation. This study is an example of how hospital managers can benefit from virtual scenario testing within a safe simulation environment to immediately visualise the impacts of policy adjustments.