Ensuring Appropriate Capacity in the Cardiac ICU

Participants:

Faculty/Clinicians/Staff: Amy Cohn, Hitinder Gurm, Julia Warner

Students: Ambika Agrawal, Emily Boltey, Imani Carson, Ryan Lentine, Luke Liu, Amanda Moreno-Hernandez, Shamit Kikani, Harini Pennathur, Camille Phaneuf, Donald Richardson, Hadi Saab, Bassel Salka, Jessica Southwell, Sheridan Tobin, Ziqi Wang, Shuhao Zhou

Project Contact: [email protected]

Project Synopsis:

Cardiac patients often need immediate care; however, access may be limited due to the lack of ICU beds. In collaboration with the Samuel & Jean Frankel Cardiovascular Center (Frankel CVC) at Michigan Medicine (MM), we were motivated to study capacity issues associated with the acceptance of transfer patients experiencing aortic dissection (AD) patients. When requesting transfer to MM, the AD patient’s approval for transfer depends on the availability of an ICU bed among other components. In this project we focus on understanding and modeling the flow of cardiac patients through the Frankel CVC at MM. The motivation for this study is to improve the capacity management strategy of the Frankel CVC.

To achieve this purpose, we have developed a discrete event simulation model of the flow of patients arriving into ICU, and then being moved into the downstream unit (Step-Down Unit, SD). The simulation also accounts for the possibility of returning to the ICU, which has a fixed bed capacity, due to deterioration of the patient’s medical condition. By collecting relevant metrics, like denial rate, utilization of ICU and SD, we evaluate the system performance with different number of ICU and SD beds and understand how sources of variability in patients can improve the outcome.

The model may be utilized to develop policies for determining whether to accept patient transfer requests to MM for AD and other high complexity cardiac conditions, and when to postpone elective surgical procedures to improve available ICU capacity for potential future urgent care needs.

Papers, Posters, & Presentations:

Posters:

Presentations: