Optimization of Overtime Costs and Operating Room Utilization

Session
Student Competition Winning Paper Presentations

Authors
Ashley Wampler
North Carolina State University

Kristen Moore
North Carolina State University

Zach Adams
North Carolina State University

Description
Due to rising costs in hospital systems across the nation, it has become extremely important to examine opportunities for process efficiency improvement. The purpose of this project was to create an automated decision support system (DSS) for surgery scheduling in the thoracic unit of Mayo Clinic, in Rochester, MN. There are two primary costs to consider when scheduling operating rooms: opening costs and overtime costs. The cost of opening an OR is one flat cost incurred for each OR opened. The cost is primarily associated with staffing of the OR itself, as well as up and down stream activities such as patient intake and recovery. The cost of overtime is calculated using the average cost of overtime per minute, the expected overtime for each OR, and the probability calculated in our model. Monte-Carlo simulation was used to evaluate manually generated schedules. Monte-Carlo simulation was used to generate surgery time estimates from previously collected sample surgery data. A novel heuristic was developed to automatically produce a near optimal schedule. The heuristics is an adaptation of longest processing time (LPT) scheduling that reflects additional constraints based on surgeon and resident availability, which arise in an academic medical center. Based on selected schedules from past data, the DSS is expected to save an average of $2,500 per day. Assuming the trials based on historical data are representative of future cases, the DSS is expected to save over $900,000.00 per year by optimizing the scheduling within the thoracic unit alone.

Abstract
Due to rising costs in hospital systems across the nation, it has become extremely important to examine opportunities for process efficiency improvement. The purpose of this project was to create an automated decision support system (DSS) for surgery scheduling in the thoracic unit of Mayo Clinic, in Rochester, MN. There are two primary costs to consider when scheduling operating rooms: opening costs and overtime costs. The cost of opening an OR is one flat cost incurred for each OR opened. The cost is primarily associated with staffing of the OR itself, as well as up and down stream activities such as patient intake and recovery. The cost of overtime is calculated using the average cost of overtime per minute, the expected overtime for each OR, and the probability calculated in our model. Monte-Carlo simulation was used to evaluate manually generated schedules. Monte-Carlo simulation was used to generate surgery time estimates from previously collected sample surgery data. A novel heuristic was developed to automatically produce a near optimal schedule. The heuristics is an adaptation of longest processing time (LPT) scheduling that reflects additional constraints based on surgeon and resident availability, which arise in an academic medical center. Based on selected schedules from past data, the DSS is expected to save an average of $2,500 per day. Assuming the trials based on historical data are representative of future cases, the DSS is expected to save over $900,000.00 per year by optimizing the scheduling within the thoracic unit alone.