SEMS Says ... Simulation models help predict delays, speed up emergency department service

By the Colby Weishaar and Anna Svirsko

Using technology to reduce project schedule delays

The construction industry is complex and evolving, making the project manager’s job of planning, organizing and decision-making a challenge.

One of the most difficult decisions throughout a project is deter­mining what resources are needed to complete a task by the deadline. A study by Stephen Mulva, Daniel Oliveira, Sungmin Yun, James J. Blaschke, Miguel Contreras, Rebecca A. Cvikota and Michael Goggin for the Construction Industry Institute (“Value of CII Best Practices Report”) concluded that nearly 60 percent of the projects they reviewed either lost money or did not make a profit. Often at the heart of this lack of profitability are schedule delays resulting from numerous uncer­tainties that the project manager must negotiate.

As new technology and software developments continue to evolve, project managers are starting to understand the value of building analytical models to help predict when delays might occur, how long they will last and why they occur. By using Monte Carlo simulation to determine project delays, a project manager can quickly run tens of thousands of iterations of the model to examine how uncertainties, such as resource shortages or delays, affect the outcome of the project. The project manager can combine the simulation results with his or her experience to identify specific patterns within the data to develop a plan to mitigate potential problems. In some instances, a mitigation plan can even be created prior to the start of a project.

My research built an analytical model using Microsoft Project, Microsoft Excel and the Excel add-in Monte Carlo simulation package Palisade’s Risk and tested it on multiple construction projects. The model divided each task within the project schedule into the three largest risk categories in the construction industry, namely equipment, material and labor. The advantage of breaking the delays into categories is that it allows the project manager to understand the specific potential delay areas and which category caused the delay. Other research has been conducted analyzing the total impact of a delay, but no literature could be found that dealt with breaking the delays down further into specific categories.

After my model was completed, five individual follow-up interviews were conducted with subject matter experts to determine the usefulness of the model to the construction industry. One of these experts was the project manager who oversaw the project data used for my research. All experts indicated the model could be useful in the industry, and four said they would consider using the model on a future project.

One reason for this is because an analytical model can increase the potential for a project to finish by its due date and ensure that the correct resources are available when needed. This not only saves time, but it also will save the company money, as equipment, material and labor do not need to be rescheduled, reordered or expedited if everything is available at exactly the right moment.

Although my research was conducted in the construction industry, this analytical model can easily be applied to other industries. The project manager has the capability to add or delete resource categories as needed to align with the specific industry of interest. As projects begin to become more complex, project managers will need to embrace technology and software designed to aid them in their ability to oversee a project. Project managers who begin to combine software with their experience could begin to see a real competitive edge compared to those who have failed to keep up technologically. 

— Colby Weishaar is studying for his Master of Science in industrial engineering at the University of Arkansas.

IEs in the emergency department: How to decrease patient length of stay

Over the past three years, I have had the opportunity to work with several professors and undergraduate students in an effort to decrease patient length of stay in the emergency department of a large children’s hospital.

A key to our success was the relationships developed between our team and the physicians and nurses working in the emergency department. Through hours of observations, data analysis and in-person interviews with the hospital staff, we were able to discover the root causes of challenges in the emergency department and create action plans to address these issues.

One specific problem that we addressed was the fast track area.

The fast track is a separate area of the emergency department where “easy” patients are treated so they are not waiting for an excessive time behind sicker patients who require longer treatment times. When patients enter the emergency department, they visit a rapid triage nurse who assigns patients an acuity level indicating how sick they are. This determines each patient’s service priority. The emergency department identified fast track patients as low acuity patients and did not consider which resources patients may need.

To better understand the problem, we observed the fast track, conducted time studies and interviewed physicians. The fast track operates with two physicians and eight rooms. The physicians often mentioned that they treated patients who should be seen in the main emergency department due to patient complexity and the resources needed. Our process analysis let us create guidelines for determining which patients were truly fast track patients, meaning they could be treated in one physician engagement session that typically lasts less than 20 minutes. We then simulated the fast track process when only true fast track patients were seen. Through the simulation, we determined the fast track only required one doctor with two or three rooms while the second physician could be assigned four treatment rooms in the main emergency department.

While our fast track results were not intuitive to the physicians, we were able to show them the work we had done regarding the simulation and explain how this change would affect physician utili­zation as well as patient length of stay. Upon seeing our data, the emergency department staff agreed to a trial period for the new fast track configuration. During the trial period we had team members observe and note any problems. Analyzing the data showed a decrease in average fast track patient length of stay of approximately 30 minutes, and our recommendation became permanent.

In addition to our improvement in the fast track, we have had success implementing changes regarding the rapid triage process, equipment location, nurse scheduling, waiting time predictions and forecasting surges in patient demand. We continue to work with the staff to help improve processes within the emergency department and pilot new projects, in particular focusing on ancillary services such as pharmacy, radiology, laboratory testing and consults. Our collaboration with the staff has been instrumental in our success in implementing projects, addressing problems that arise in real time and tracking the success of the project so that the emergency department staff owns the solution.

— Anna Svirsko is a Ph.D. candidate in the industrial engineering department at the University of Pittsburgh. For her current research she has collaborated with Geisinger Health System to improve pharmacy operations and Children’s Hospital of Pittsburgh to improve emergency department operations and the ambulatory scheduling process.