EMERGENCY PROVIDERS UNDERSTAND that patient boarding is not ideal, but they are powerless to fix the problem when there are no open beds on the upper floors. However, that doesn’t mean that improvements can’t be achieved. Indeed, researchers at Penn State Hershey Children’s Hospital in Hershey, PA, have shown that process changes in the way hospitalists typically discharge patients can make a big difference in opening up beds more quickly, thereby easing pressure on the ED.
Using Lean Six Sigma (LSS) management techniques, investigators, led by Michael Beck, MD, an associate professor of Pediatrics at Penn State College of Medicine, were able to slice as much as 90 minutes off of discharge times, enabling the hospital to discharge 45% more patients per day than control groups. Researchers report that they accomplished this without increasing length-of-stay (LOS) or re-admission rates.1
Further, since the study, many of these improvements have been sustained, at least on Beck’s general pediatrics service, which involves two hospital floors. However, even with support from the ED, Beck acknowledges that persuading other hospitalist services to implement similar changes has been a challenge. “I think it is gaining traction,” Beck says. “But culture in healthcare is a really hard thing to change. People don’t want to change. They are very resistant to it.”
Consider impact of discharge times
Like many hospitals, Penn State Hershey Children’s, which is housed within Penn State Hershey Medical Center, doesn’t have a lot of open bed space, so clinicians are dealing with a fixed capacity, high volume, and patients who are high acuity, Beck notes. “As a result of this, we suffer from a lot of ED boarding and a lot of ED crowding,” he says.
Pediatric emergency medicine is part of the general adult ED with dedicated rooms and dedicated pediatric providers, explains Subhankar Bandyopadhyay, MD, CMQ, FAAP, an associate professor of emergency medicine and pediatrics, and director of pediatric quality in the ED at Penn State Hershey Medical Center. “If the pediatric census is high, pediatric capacity in the ED becomes flexible, and children are placed in other areas in the department where rooms are available,” Bandyopadhyay says. “Also, capacity is at times constrained due to lack of available staff.”
Using LSS techniques, investigators determined that a big part of the ED crowding/boarding problem was a long-standing academic medical center practice of not issuing discharge orders until 1 p.m. to 3 p.m. in the afternoon, long after rounds have been completed. “As it turns out, when we are rounding on 12, 13, or even 15 patients, there is not enough time to do the orders during rounds,” Beck says. “You only have 180 minutes to round because you have to teach residents, so if you have 180 minutes to see 15 patients, that amounts to just 11 minutes per patient, and the cycle time for one discharge is 15 to 20 minutes.”
Given that the average number of discharges per day on Beck’s service is between four and five during busy months, that requires 100 minutes just for an inpatient provider to do discharges, so there was no way to complete the discharge orders during rounds. “I can’t teach and take care of the sickest patients in a way that won’t create errors and do all of the discharge work in the same 180 minutes,” Beck observes.
Complete discharge work during rounds
To resolve this problem while also facilitating earlier discharges, investigators considered what could be accomplished if they added an extra inpatient attending physician to the mix, at least during the high-census months — from October through April — so that instead of one provider having to round on 15 patients, he or she would round on six to eight patients. “If you still have the same 180 minutes, the math works,” Beck notes. “You have twice the amount of time to do all the things you are supposed to do, [including] the teaching on rounds and the discharges on rounds in one piece flow.”
To test what impact this staffing adjustment would have on discharges, investigators scheduled two attending physicians to staff the service during the high-census months, theorizing that the ability to complete the discharges during rounds would facilitate throughput in the ED. “If you can create an open bed earlier in the day, you can create beds for the ED, because they usually start to ramp up their admissions around 10, 11, or 12 o’clock every day, so that is also predictable,” Beck observes.
“We put in a plan so that if you have enough time on rounds, then you do the discharges on rounds. You don’t wait until the afternoon,” Beck explains. “You do the discharges when you tell the patient that [he or she] is going home. That is when you do the patient’s paperwork … and you don’t go to the next patient until you do that work.”
Compared to concurrent control groups, the investigators’ approach worked just as they had theorized. The median time for discharge orders being entered was 10:45 a.m. compared with a median time of 2:05 p.m. for the control groups. Further, the approach delivered consistent results regardless of the personnel involved, according to investigators.1
“We first tried it [as a pilot] for the month of March in 2013,” Beck says. “The nurses grumbled about it because they didn’t know who to call [with two attending physicians], and it was changing everybody’s day-to-day operations.”
The residents and faculty also complained about the changes, Beck recalls, but he notes that by the end of the month, people wanted to know why they had to switch back to the old system. “We didn’t have enough of a plan at that point to go further, but for the next year we planned on moving [the model] forward,” he says.
Examine schedule adjustments
To accomplish this, the core group of seven inpatient providers needed to agree to work an extra two or three “on service” weeks per year, Beck notes. This would enable administrators to essentially redeploy previously “off service” attending physicians during the peak months. “We actually had within our core group of faculty the amount of people we needed to staff two teams. That was all we did. It was a simple solution,” he says. “We didn’t hire anybody. We just utilized staff in a way that was more meaningful to the needs of the patient … and we have been doing this now for 15 months.”
Beck observes that while this type of scheduling works best for both the hospital and patients, it isn’t how things usually work in an academic center. “The way most hospitals run — [the providers] say that they are going to finish their rounds and then finish the paperwork at the end of the day,” he says. “That is hurting the organization because you are not doing the [discharge work], and the patient is held hostage in the hospital until you can get around to doing it.”
Further, care quality becomes an issue when patients get backed up in the ED. “You can’t have patients in the ED for eight hours or longer,” Beck says. “That is not sustainable. It is a safety issue.”
In addition to creating two rounding teams, Beck and colleagues have also devised a daily huddle to discuss and facilitate any discharges planned for the next day. “Every day at 2:30 p.m. we get together with our care coordinators and social worker as a team to identify the next day’s discharges and make sure the family is ready for the discharge, that they have any prescriptions they need filled, appointments made, and transportation [covered],” Beck explains. “Then we start with those discharges the next day.”
Bandyopadhyay notes that Beck’s work in this area has generated significant interest across the health system. “The boarding time for inpatient pediatric admissions has decreased significantly when [pre- and post-intervention times are compared],” he observes.
Bandyopadhyay adds that the intervention may be producing other benefits as well. “We have yet to analyze the data to look at the overall effect on our ED, namely the number of children left without being seen, door to [physician] times, and room turnaround times,” he says.
Further, while the two-team approach has thus far only been implemented on Beck’s service line, it is nonetheless having a positive impact on other service lines. “We were able to show that not just our patients were in the ED a shorter time, the ED boarding times for other service lines dropped as well,” Beck says.
An added bonus to the two-team approach is that it enables physicians to spend more time doing the kind of work that they prefer, offers Beck. “There is a big difference between doing a job and doing a job that you enjoy,” he says. “And you actually have more time to spend with medical students, more time to spend with families, and more time to teach on rounds, so you don’t just see a patient once a day; you actually get to see them twice a day, and sometimes more.”
Beck adds that the intervention has required no new information technology, extra beds, or new staff, but it has offered considerable value. “Two teams working in parallel are going to be much more efficient and more beneficial to education and the teaching mission of the organization than one large team,” he says. “You are much more nimble, and you start your day differently.”
-
Beck M, Gosik K. Redesigning an inpatient pediatric service using Lean to improve throughput efficiency. J Hosp Med 2015;10:220-227.
-
Subhankar Bandyopadhyay, MD, CMQ, FAAP, Associate Professor, Emergency Medicine and Pediatrics, and Director, Pediatric Quality, Emergency Department, Penn State Hershey Medical Center, Hershey, PA. E-mail: [email protected].
-
Michael Beck, MD, Associate Professor, Pediatrics, Penn State College of Medicine, Hershey, PA. E-mail: [email protected].
