LPCH Capacity Planning

In 2018, Lucile Packard Children’s Hospital Stanford (LPCH) opened the Bonnie Uytengsu and Family Surgery and Interventional Center, which houses six surgical suites and six interventional treatment rooms including interventional radiology (IR) and cardiac catheterization (CATH)/electrophysiology (EP) labs used for cutting-edge treatment. Accurately modeling the volume incurred by LPCH service lines is important as future projections will not only inform decisions on capital expansions, surgical service growth, and patient volume allocation within the LPCH system, but also improve patient care. To date, a capacity planning model for clinic capacity, operating room (OR) utilization, and bed demand exists, yet it does not incorporate IR (comprised of IR Pain, IR Neuro, and General IR), CATH, and EP (the core service lines). The LPCH Capacity Planning project aims to close the existing gap in the model and increase hospital leaders’ understanding of the trends and relationships between IR, CATH, and EP and other hospital service lines. To this end, a descriptive analysis of the core service lines was completed, the conversion rates between all LPCH service lines and IR, CATH, and EP were calculated, and a user-friendly volume and use time projection simulation was developed.

To fulfill the project objectives, several aspects of the core service lines were explored. First, the patient journey was illustrated to understand how patients flow through the hospital. Then, for each core service line, a historical analysis of total patient procedure volume was conducted and the fluctuation of patient class volume over time was examined. Next, the average length of stay among each inpatient patient class, for not only IR, CATH, and EP, but also each department were compared. OR utilization was calculated to inform hospital leaders on the current use of resources. Procedure and provider data were explored and reveled volume variation across service lines. Further, the conversion rates between all LPCH service lines and IR, CATH, and EP were calculated. The conversion rates will enable hospital leaders’ understanding of how IR, CATH, and EP generates demand for other hospital service lines and vice versa. Finally, a user-friendly volume and use time projection simulation was developed and the results, along with all other analyses were incorporated into an interactive RShiny dashboard. Through understanding patient volume, procedure, provider, and operational trends, hospital leaders can make informed decisions that positively influence the trajectory of LPCH and its patients. 

It is important to note that our analysis includes annualized data.

The LPCH patient journey was illustrated to understand how patients flow through the hospital. Close attention was paid to the differences between three inpatient patient classes: Medical Admit; Surgical, Elective; and Surgical, Non-Elective.  The fourth patient class, Outpatient, is not pictured as outpatients are not admitted to the hospital.

Hospital patient day (i.e., length of stay) trends are similar across each core service line, with Medical Admits and Surgical, Elective patients staying the longest and shortest, respectively.

The conversion rate graph highlights service associations. The values were calculated by locating a patient’s first core service line (the “base service”: y-axis) and summing the visits they had to each primary service line afterwards (the “subsequent service” x-axis). Specifically, CATH/IR patients tend to get repeated CATH/IR procedures as well as Neurosurgery and Cardiovascular. 

The model diagram depicts the most basic version of inputs (yellow), steps (blue) and outputs (red) of the user-friendly volume and use time projection simulation.

Valerie Przkeop

M.S. Biomedical Informatics

B.S. Management Science & Engineering

B.A. Psychology

Jenia Borisenko

M.S. Biomedical Informatics

B.S. Biomedical Computation

David Scheinker

Founder & Director, SURF Stanford Medicine