Predicting the future’ helps cut LOS by 50%
Bed turnover time slashed from hours to minutes
ED managers may not possess a crystal ball, but the ability to predict future events is nonetheless critical to their success, notes Bonnie Coalt, RN, MS, director of nursing at Miami Valley Hospital in Dayton, OH.
"The role of ED managers is to be able to predict by hour of the day their walk-in patient arrivals, their ambulance patient arrivals — by hour of the day and by day of the week — and also to be able to predict the admissions, by hour of the day, that generate out of the ED," she asserts.
The ability to forecast such statistics, part of a comprehensive strategic improvement initiative at Miami Valley, has helped cut length of stay (LOS) nearly in half and helped slash bed turnover time from several hours to just a few minutes. Forecasting is, in fact, the key to fulfilling any number of management responsibilities, says Coalt. "No. 1, it’s important so you can do staffing appropriately," she says. "It’s also tremendously helpful in improving patient flow."
In 2002, when the initiative began, the ED diverted 2,010 hours. By 2003, it was able to reduce that to 860 hours; and in 2004, the department was able to maintain at 792 hours — with higher volume.
Coalt recognizes that many people perceive the ED as highly unpredictable, but says that when you look at trends over a long period of time, they are quite predictable — even by a statistician’s standards. Once you have studied these trends, you will know your future demand, including the demand for scheduled surgeries, elective inpatients, as well as for patients who initially come in for outpatient treatment but end up as inpatients. Subsequently, you will be able to predict peak demand times for all of those areas, she notes.
"You will then be in a better position to compete for resources, and for available, appropriate beds," Coalt explains.
This statement brings up one more key aspect of the predictive process: The recognition that the ED does not operate in a vacuum. "It is essential that the ED integrates [its own analysis] with the analysis of the whole hospital’s demand," says Diane Pleiman, CNMT, RTN, MBA, director of financial operations, patient placement, and staffing. "[Integration] is probably one of the greatest things ED managers need to know."
Learning to Excel’
Miami Valley began its own experience with predictive modeling this past fall, using a standard Microsoft Excel program.
"We entered the data in Excel, and then created pivot tables [also an Excel function] to sort the data by hour of day, day of week, and so forth," Pleiman explains.
The ED has five clinical stations in the ED, and it could pull data from those computers to identify the highest acuity station, as well as fast-track data, Coalt adds. "We pulled data out of our computer system, as well as our charge system. That’s how we knew what acuity level the patients were," she says.
"We had to collect the data in order to be able to make observations and validate that some lean process’ changes would work," Pleiman notes. (Editor’s note: "Lean process" is a Six Sigma term that basically refers to an organization becoming quicker and more agile.) "We were also able to predict takt time,’" she continues. (Editor’s note: "Takt time" is another Six Sigma term, a German word indicating the time it takes to complete an individual activity, such as triage.)
"We were able to predict the number of patients we would have to discharge out of the department in order to accommodate incoming patients, based on average length of stay, and how much we would have to reduce the LOS in order to meet that takt time," Pleiman says.
At first, the ED predicted its own demand and workload figures. Next, it integrated the numbers with those of the entire hospital.
"Without that integration, folks would think the ED does live in a vacuum," she adds. "And we would have gotten into more trouble in rerouting and diverts if we only looked at our own piece of the puzzle."
After examining the data, a number of changes were implemented. Among the most effective: changes in transportation staff. "We were able to zero in on the bottleneck point, which required more staffing," Coalt explains. "We did not change [the number of] RNs, but we did for clerical support and patient care technicians that do transport out of the ED."
A total of 21 peak-level staffing level changes were made, and in six months, the LOS (from the time it is determined to admit a patient until that person is in an inpatient bed) was cut from more than three hours to an average of 1.7 hours.
A new interactive voice response telephone system for notifying staff when a patient is transferred out of a room has helped cut bed turnaround time from five to six hours to a few minutes.
When a staff member picks up the phone, he or she hears a series of computer-generated prompts; using these prompts that person can instantly notify the entire staff of the bed availability. New standard operating procedures also were created, to decrease variation when staff personnel changed.
"Through these initiatives, we have raised awareness for the entire institution," Pleiman concludes.
For more information on predictive skills in the ED, contact:
- Bonnie Coalt, RN, MS, Director, Nursing, Miami Valley Hospital, One Wyoming St., Dayton, OH 45409-2793. Phone: (937) 208-6196.
- Diane Pleiman, CNMT, RTN, MBA, Director, Financial Operations, Patient Placement and Staffing, Miami Valley Hospital, One Wyoming St., Dayton, OH 45409-2793. Phone: (937) 208-6196.