"County/USC Hospital serves an enormous population in the region, with many patients going to the emergency room first, so the waits can be long," said Randolph Hall, principal investigator and professor of industrial and systems engineering in USC's Viterbi School of Engineering. "Given the hospital's budget constraints, it's really critical that they develop systems and procedures to move patients through the hospital as efficiently as possible. "
In the first study of its kind for USC's Viterbi School of Engineering, Hall and a team of engineers in the Epstein Department of Industrial and Systems Engineering partnered with Los Angeles County/USC Hospital to model the entire hospital, from top to bottom, as an integrated system. The objective was to identify the most significant bottlenecks, as well as the highest payoff strategies for improving patient flow in one of the nation's largest public hospitals.
Los Angeles County/USC Hospital is a 1,395-bed facility located in East Los Angeles. The county-run facility is a partner institution of the Keck School of Medicine of USC that treats approximately 800,000 patients annually, including at least 200,000 emergency room visits. According to hospital records, almost half of those treated are indigent or uninsured.
To keep up with the demand, Hall said, patients must be transferred without delay from emergency room triage to ancillary services, such as radiology departments, for x-rays, CT scans, MRIs and other diagnostic tests, or on to operating rooms for surgery. Those who need to be admitted to the hospital must further wait for a bed to be freed up, which depends on the speed at which patients are discharged, rooms are prepared and patients are transferred among locations.
"All of these departmental activities are interdependent as patients flow through the emergency care system," Hall said. "It's like a domino effect if there is a delay in any one of those steps. Every other department down the line will experience a delay."
Hall and his team -- co-principal investigator Maged Dessouky, an associate professor of industrial and systems engineering, and David Belson, senior research associate and lecturer in the Epstein Department of Industrial and Systems Engineering -- first identified care-giver perceptions of the most critical issues they face while serving patients.
"For instance, we talked to triage nurses to find out how patients are classified based on their level of medical need," Belson said. "Do they give gunshot victims or accident victims priority over someone having a heart attack? Then we looked at the degree to which the triage decision process ties up available resources, such as specialist physicians or equipment. Maybe the emergency room is taking away all of the hospital's resources to treat inpatients, which means they will stay longer and you'll have fewer beds for newly admitted patients."
The researchers applied engineering techniques widely used to improve the performance of manufacturing, distribution and transportation systems to design a flow chart of hospital operations.
"The engineering approach makes this study unique," said David Altman, assistant dean of the Keck School of Medicine. "We're looking at process issues that we don't normally look at. There's a lot that the industrial world has been figuring out that we don't know much about in health care, but it could be applied to us and be tremendously beneficial."
They used information from more than 100 hospital administrators and clinicians to document processes currently in use for triage, scheduling and routing.
"One of the major bottlenecks seems to be occurring in ancillary services, such as radiology," Altman said. "Ancillary services are at critical junctures of just about every hospital procedure, but these services have limited capacity and staff don't answer to the people who are responsible for the patients. So a patient could wait days before his or her surgery is scheduled, or hours for an x-ray." Discharging patients from the hospital is another time-consuming process, said Paula Packwood, chief of staff in the Los Angeles County Department of Health Services.
"We know that we have problems with some of the more complex patient discharges, because the patients are uninsured and they don't have a lot of options for care when they leave," she said. "A systems approach to the problem is really like a new way of thinking about things."
For instance, there might be a better way to coordinate interdepartmental scheduling of patients who are ready to be discharged, Packwood said, "or, with all of the new technology that's available, we might be able to develop better software tools to make these administrative processes more efficient."
In phase two of the study, Dessouky, who specializes in simulation and modeling, will develop a testbed to connect all patient flow activities together and gather data for analysis. Using this data, he and his colleagues plan to develop new patient flow models and data management tools. Processes that will be modeled include patient discharge, organization and assignment of responsibilities to departments, scheduling procedures for operating rooms and nursing staff, and forecasting patient volume for a particular day and time. Routing procedures, which involve sequencing all of the tests and examinations each patient needs, will also be examined in more detail, Dessouky said.
The emergency room, which takes in approximately 85 percent of all new inpatients, stands to benefit greatly from the hospital's ability to implement streamlined administrative processes, Hall pointed out. Staff will be able to triage and schedule patients with fewer delays and provide new patients with more accurate estimates of wait times.
In parallel with new and improved procedures to manage patient flow, engineers in the Viterbi School will also develop a collaborative educational program with County/USC Hospital, which will include student placement in the hospital and a new course in industrial engineering on patient flow improvement.
"This will be an excellent way for industrial and systems engineering students to gain hands-on experience," Hall said. "Once our testbed is fully implemented, we will be able to collect real-time data automatically.
"In the future, we'd also like to develop a capability to remotely access the data from campus, so that students and faculty working on the project would be able to monitor and implement software updates for the hospital when needed," he said.