In March 2002, Baptist Memorial Hospital-Memphis was convinced its capacity problems were intractable. Memphis community hospitals were diverting ambulances an average of 70 percent of the time, with ambulances waiting at emergency departments (EDs) for up to 90 minutes. They had tried several initiatives designed to improve flow, but with disappointing results.

 

Today, the hospital — a 736-bed tertiary care facility in East Memphis and the busiest in Tennessee, with 90 percent or better occupancy most of the time — has improved flow dramatically. Holding hours in the ED for admitted patients have been reduced by 50 percent, length of stay in the ICU has been reduced by two days, and patient satisfaction has improved from the 10th percentile to the 85th percentile, according to Gallup Organization surveys. How did they do all of this, without hiring additional staff or adding beds?

 

“This has been a very long journey,” says hospital nursing director Suzanne Horton, explaining that the hospital had been working on flow issues for four to five years, even before joining IHI’s IMPACT Collaborative in fall of 2002.

 

“Before we began working with IHI, we had tremendous holding hours in our ED, with many patients waiting for beds in the ICU,” Horton explains. The bottlenecks resulted in frequent diversions to other hospitals.

 

The focus of the hospital’s flow project was naturally the ED. One of the first changes was to develop an express admission unit, opened in fall of 2001. “When patients arrive on this unit, their admission paperwork is done, blood work and/or EKG work is done, and any urgent medications or treatment are administered,” Horton says. “This gives the physician the information needed to develop a plan of care without delay.”

 

Anticipating the winter 2002 crunch, the hospital began to develop a winter plan, deciding to keep the express admission unit open 24 hours per day, instead of 16. The results were dramatic. “We saw tremendous improvement in flow,” says Horton. “It was a tremendous high-leverage change that halved waiting times.”

 

In March 2002, the hospital agreed to stop diverting ambulances in their community. “But we knew we couldn’t just do that, without figuring out how to help the ED better manage it,” says Horton. After the hospital joined IHI’s IMPACT Collaborative, the first thing they did was to start faxing ED reports to the receiving floors. That meant ED nurses didn’t have to be on hold while waiting for the right nurse or be forced to call repeatedly.

 

They also worked on staffing, matching staff to volume without necessarily increasing it. “We rearranged staffing to place two nurses in triage during peak hours, and fewer during non-peak hours,” Horton says. “The physicians changed their coverage, with more physicians during peak hours and fewer during non-peak hours. They also changed their times to change shifts at different times from nurses. This allowed for pressure to complete care prior to the end of the shift to be applied by both groups, resulting in less lag time at shift change.”

 

One of the highest-leverage changes, Horton says, was the fast track — “sort of like having an urgent care in your ED. There's a lot of volume in an ED that can be sent to this area.”

 

“We developed the fast-track area, staffed it appropriately, and really focused on it,” she says. The hospital increased the number of patients treated in the fast track and decreased the time it took to see them, making it a “huge win” in terms of decreased waiting times.

 

While the hospital was making all of these changes in the ED, they were participating in an IHI Critical Care Collaborative as well. The initiated an intensivist program in February 2003. Intensivists conduct multidisciplinary rounds three times per week during the day and every night, making recommendations for improving care, weaning, and other issues according to IHI “bundles”  such as those for ventilators and urinary tract infections.

 

“It wasn’t rocket science,” Horton says. “They simply applied evidence-based medicine: the ventilator, central line, readiness to wean, pain and sedation, urinary tract, lines and tubes out as soon as possible, and appropriate level of care bundles. All these disciplines made sure the patient had the very best care possible. These actions prevented infection and complications, and patients were better able to stay on track.” The result was “phenomenal outcomes,” says Horton. The mortality rate in the ICU decreased by 40 percent, and ventilator-associated pneumonia was reduced by 80 percent. The average length of stay was reduced by two days, from approximately seven to five. That change has been stable over time, Horton says, and the ICU’s volume has increased by 20 percent — to about 230 patients per month. On some days, there were eight empty beds in the ICU.

 

Because the ICU’s being full had been one of the bottlenecks in the ED, the trickle-down effect of quicker flow through the ICU was felt in the ED. “We were holding a lot of ICU patients in the ED, but they now got to go to the ICU without holding,” says Horton.

 

The result of all these changes was that the ED turnaround time was reduced by 9 percent and the volume increased by 6 percent. In a community that now had no diversion, the hospital was able to decrease ambulance turnaround time (the time from arrival until the ambulance is ready to return to the field) to 7 to 10 minutes, compared to a community average of 60 minutes.

 

After a year of success, the hospital was ready for more initiatives. “We looked at that concept and said, ‘Hmmm, this really has some possibilities for patients outside of the ICU,’” Horton says. The hospital had already begun using a patient-centered care model, with an access care manager at all points of entry. The goals include early discharge planning, initiation of protocols, assurance of appropriate level of care, identification of social issues, patient education needs, flu and pneumonia vaccination, and smoking cessation.

 

The number of patients with protocol orders has increased from 199 in December 2003 to 241 per month in early 2004. The hospital has also improved its smoking cessation efforts and pneumonia vaccination rates.

 

Multidisciplinary rounds (MDRs) were another key change. “Our goal was to document and plan for daily goals and barriers to discharge, track our CMS indicators, increase our Gallup scores, and decrease the length of stay,” Horton says. “We really view length of stay as an outcome of our quality and efficiency.”

 

MDRs outside the ICU are not facilitated by physicians, but by nurse managers and their designees. Ninety-one percent of patient care areas are now performing them. “What's important about these [MDR] tools is that we’ve developed a toolkit to spread their use within our own organization and our own system,” says Horton, who is planning to create a CD of all of these tools to distribute at IHI’s December conference.

 

The hospital has identified four major areas that need improvement in MDRs: anticoagulation dosing and education, initiation of ACE inhibitors, changing medications for safety, and DVT prophylaxis.

 

The hospital uses a checklist adapted from Hackensack University Medical Center’s Checksheet of Quality Indicators to improve CMS indicators. If patients aren’t receiving something on the checklist, the case manager discusses with the physician whether he or she would like to include it in the plan of care, or to document why not.

 

There has also been significant improvement in patient satisfaction, according to Gallup scores on four measures: nurse calmed fears, safe and secure, problems resolved, education and information provided.

 

The unit that has been doing MDRs the longest has shown the most improvement. Also, the length of stay in the test units has decreased: one from 7.5 days to 6 days, and another from 7 days to about 5.7 days. “There has definitely been a trend downward in all those test units,” says Horton. And the overall length of stay is down 8.9 percent from the previous year. “I can't tell you we've met our goal,” Horton says, “but compared to last year we're significantly better.”

 

The hospital has begun including orchestrated discharge in patient planning. “It's really not meant to decrease the length of stay,” explains Horton, “but to provide an expected appointment time for discharge and do all the planning and education in advance so we can meet it.” The team makes sure they have done all of the necessary education, ordering of any kind of equipment and anything else necessary to get patients ready to go home. “We’ve successfully championed 33 of our high-admitting physicians to do this,” says Horton.

 

While medical units remain a challenge for incorporation of MDRs, surgical units proved easier because of their relative predictability. Now, with seven surgical units participating, 35 percent of discharges on surgical units are on orchestrated discharges, with 90-95 percent going home within a half hour of predicted discharge.

 

Horton heard from some team members that they couldn’t do the increased work of MDRs. To them, the team responded, “MDR is not more work; it's the work.”

 

The changes have had an effect beyond what the numbers describe. There’s been a culture change into a “’do no harm,’ safety-oriented environment,” Horton says. And staff satisfaction has improved along with improved patient care. “We’re moving from being reactive to being proactive,” says Horton. “Even though we don't round on 100 percent of patients, the staff understands that the same principles apply to their other patients, so there’s a bit of a halo effect.”

 

“Improving flow is not about building more beds or hiring more staff,” Horton wrote in the summer 2004 issue of Frontiers of Health Services Management. “It is about developing and improving processes to decrease variability and smooth flow. For BMH-M, it required a culture change of leadership and staff that moved us from ‘We’re doing the best we can’ to ‘Sorry, that is not good enough.’"