Published in the August 2017 issue of Today’s Hospitalist
FOR SEVERAL YEARS, hospitalists at the University of Virginia Medical Center in Charlottesville worked to meet an early-discharge target: Physicians were expected to place 50% of their discharge orders by 9 a.m., and patients were supposed to be out of their beds by noon.
But the hospital ran into the same problems that many groups struggle with when it comes to early discharges. Length of stay went up, and “there was clear evidence of gaming in some instances,” with doctors keeping patients who could have been discharged the day before overnight to meet the incentive, says hospitalist director, George Hoke, MD. The UVA initiative was abandoned by 2015 and never resurrected.
“It defeated the whole purpose of improving throughput,” Dr. Hoke points out. “The concept of discharge by noon as a means to improve bed availability and capacity problems is flawed.” Why? “Because most hospitals are operating at too high a capacity for optimal patient flow. The right answer is to either reduce length of stay by improving efficiency or add more beds.”
“You have to know your hospital, your hospitalists and your patient population, then set the right target.”
Not so, says hospitalist Katherine Hochman, MD, the former hospitalist director and now associate chair of quality at New York’s NYU Langone Medical Center. For years, NYU Langone has sustained a better-than-40% discharge-before-noon rate—with patients out of their beds by noon—as described in separate studies published in the April 2014 and October 2015 issues of the Journal of Hospital Medicine.
The center’s observed-to-expected length of stay has decreased, even though patient acuity and surgery volume have risen. “The discharge-before-noon metric,” Dr. Hochman contends, “is a measure of teamwork and coordination on your service.”
So who’s right? The conflicting opinions underscore the fact that hospitalist groups implementing early-discharge initiatives report a wide range of experiences, from train wreck to high-functioning discharge machine.
The 2016 Society of Hospital Medicine survey notes that just over half— 54%—of nonacademic hospitalist groups that treat adults have an incentive tied to early-morning discharge orders or times, meaning that just under half don’t. It’s a controversial topic that generates a lot of debate.
The right answer is that it depends, says Jose Fernandez-Duarte, MD, the system medical director for five TeamHealth hospitalist programs within Memorial Healthcare System in southern Florida. He and his colleagues launched an early-discharge initiative in one hospital last year.
“You have to know your hospital, your hospitalists and your patient population, then set the right target,” Dr. Fernandez-Duarte says. “It’s certainly not right for every hospital.”
The need for comprehensive coordination
The 227-bed Hamilton Medical Center in Dalton, Ga., launched an early-discharge initiative earlier this year. But as Laura Conger, MD, the hospitalist director, explains, it’s already crashed and burned.
“The financial incentive got doctors’ attention and changed their behavior.”
~ Matthew George, MD
Mississippi Baptist Medical Center
The hospitalists did meet their target, filing at least 40% of discharge orders by 11 a.m. But “doctors got very creative at finding ways to file orders tagged as dependent on something else, like a certain specialist clearing the patient first,” Dr. Conger notes. “The order was on time, but the patient didn’t actually leave the building.”
But what really killed the initiative, she explains, was the fact that all the other hospital services involved in discharges weren’t retooled to follow the hospitalists’ lead. Housekeeping was particularly hard-hit by the early orders and ended up being over-staffed at some times, under-staffed at others. As a result, patients ended up spending many more hours in the hospital after discharge orders were placed than they had before.
“It totally screwed up throughput because the services involved weren’t coordinated,” Dr. Conger points out. “There was such a domino effect that our CEO said, ‘Don’t do this any more.’ ”
At the other end of the spectrum, NYU Langone’s Dr. Hochman credits comprehensive coordination for her center’s discharge-before-noon (DBN) success. Far from being just a hospitalist project, the initiative became “a really big part of hospital culture. We’re very metric-driven,” an approach that includes ancillary and administrative services.
The original 2012 initiative involved only one general medicine floor but plenty of preparation and fanfare: a kickoff event and unit champions, an interdisciplinary checklist and twice-daily DBN automated e-mails, real-time progress posters, and a late-morning meeting with medical directors and nurse managers to review that day’s “failed” DBNs.
“The concept of discharge by noon as a means to improve bed availability and capacity problems is flawed.”
Many “fails” were due to transportation and communication problems, Dr. Hochman says. As one fix, case management began targeting patient transportation the day before discharge.
And “dialysis patients weren’t leaving by noon because their four-hour dialysis session began at 8 a.m,” she notes. “One dialysis nurse stepped up and started working at 7 a.m. instead.” That speaks, she adds, to the “innovation, creativity and grit this takes to get it done.”
Should it be incentivized?
After five years, says Dr. Hochman, the DBN metric is so well-established that it’s now part of physician incentives.
But initially, it wasn’t incentivized. Instead, Dr. Hochman made the case for patient safety as to why patients should be discharged as soon as appropriate, and success with the metric was celebrated with pizza parties and cupcakes, not bonuses.
At St. Peter’s Hospital in Albany, N.Y., an early-discharge initiative likewise isn’t incentivized. Instead, “it’s all about the pride,” says Thea Dalfino, MD, hospitalist department chief. Launched in 2015, the initiative’s goal is to have the group’s 13 rounding physicians and five advanced practitioners submit discharge orders for seven patients before 10 a.m.
“We were at 15% when we began, and we’re up to 29% without any increase in length of stay,” says Dr. Dalfino. To motivate doctors, she posts “everyone’s names, how many orders they did that day and that week” in the hospitalist office every day. While doctors have been meeting (or beating) their target for more than a year, that daily list is still posted.
A push to be more proactive
Dr. Fernandez-Duarte in southern Florida also chose to not tie the metric to any compensation. “Discharge before noon is controversial, and our group was the first among our programs to launch an initiative,” he says. “I wanted to make sure we weren’t running before we learned how to walk.”
“The discharge-before-noon metric is a measure of teamwork and coordination on your service.”
~ Katherine Hochman, MD
NYU Langone Medical Center
He does, however, track individual physicians’ percentage of before-noon orders. “Most are good performers, but you need to have a conversation with perhaps 20% to help them reach their goal,” he points out. If you rely on only a group metric, “it’s hard for individual physicians to feel accountable.”
But at Mississippi Baptist Medical Center in Jackson, Miss., hospitalist medical director Matthew George, MD, says his doctors have received financial incentives since their early-discharge initiative (50% of discharge orders before noon) began in January 2016.
Having the metric count toward their quarterly bonus wasn’t a huge problem, Dr. George notes, because many physicians were within striking distance of the target anyway. But for those who weren’t, “the financial incentive got doctors’ attention and changed their behavior.”
One change: The group has had a longstanding policy that doctors should see emergent patients first in the morning, then potential discharges. (See ““Which patients should you see first?“.) But “some people still wouldn’t make discharges a priority,” he says, until the incentive was put in place.
Creating the incentive also encouraged physicians to arrive on time, or even early. But most importantly, it made hospitalists “much more proactive in hunting down consultants and actively looking for test results, anything to expedite the process,” says Dr. George. “When timing of discharge orders didn’t make much difference to their paycheck, there wasn’t a push to go that extra mile.”
What’s the right target?
One reason why UVA’s Dr. Hoke remains skeptical of early discharges is that he believes the targets being set aren’t reasonable, at least not for general medicine patients.
High discharge targets for “protocolized services,” like the elective joint-replacement unit his hospitalists comanage, have been helpful. “They forced case managers, social workers and nurses to change their schedules to make discharges happen earlier,” he says. With such a predictable patient population, hospitalists can routinely file 50% of discharge orders by 9 a.m.
But with general medicine patients, “we were never at 25%, and often much lower.” As Dr. Hoke sees it, only a few general medicine patients are even appropriate for early-morning discharge, such as those needing another day of IV antibiotics or renal-failure patients for whom doctors want to check a morning creatinine.
Otherwise, he doesn’t see why many early-morning discharges aren’t being processed out the afternoon or evening before. Most patients do not undergo additional diagnostic testing overnight nor have new treatments initiated, he argues. And “almost none have unstable vital signs at that 7 p.m. check.”
Dr. Hoke admits, however, that it’s tough to process discharges after 6 p.m., particularly because peak admissions occur between 3 p.m. and 11 p.m. Starting this September, his center is adding a swing shift with an additional doctor doing admissions during those hours. That should help day rounders more efficiently process late-afternoon discharges.
But in southern Florida, Dr. Fernandez-Duarte says that a target of 50%-55% of discharge orders before noon (not before 9 a.m., which was the now-defunct threshold at UVA) works for his group. For one, the hospitalists were already filing between 35% and 40% of discharge orders by noon, so the target was achievable.
Currently, the hospitalists are part of a non-teaching service, which frees their time to concentrate on early discharges. In addition, his group tends to have what Dr. Fernandez-Duarte calls “the right patient population” for early-morning discharges: mostly insured patients with available follow-up.
“Some safety-net hospitals may have more complicated discharge planning,” he points out. “Ours wouldn’t be a good target for every hospital.”
Safeguards in place
At the same time, says Dr. Fernandez-Duarte, hospitals chasing early-discharge targets have to build in safeguards to prevent inappropriate utilization of the metric. Along with the percentage of discharge orders filed before noon, he tracks the hospital’s raw length of stay as well as CMI-adjusted length of stay.
In Albany, Dr. Dalfino notes that in addition to length of stay, each service that helps make discharges happen has a target that’s likewise tracked:
Transportation’s goal, for instance, is a 30-minute turnaround, while the target for bed turnover with a clean bed is 60 minutes.
“We look at all those metrics every other week at our leadership rounds,” she points out. “We also discuss any staffing issues that may get in the way of meeting targets to try to help those services improve.”
At Mississippi Baptist, Dr. George says hospitalists easily meet their 50% target of discharge orders by noon with “crazy high” percentages that sometimes top 80%. Data show, for instance, that so far in 2017, more than 70% of discharge orders were filed before 11 a.m.
At the same time, the hospitalists have seen their patient satisfaction scores rise and length of stay go down. But Dr. George chalks that reduction to several other initiatives that have been put in place.
First and foremost is unit-based rounding with multidisciplinary team rounds. (Both Dr. Dalfino and Dr. Hochman also credit regionalization with successful a.m. discharges.)
Also tamping down length of stay, says Dr. George: low physician turnover and eliminating the use of locums. In addition, putting a dedicated admitter in place has meant that rounders have more time to focus on timely discharges.
“And we’ve put nurse practitioners on the floors with high turnover, like the cardiac floors,” says Dr. George. “They help doctors get discharge orders ready.”
At Good Samaritan Hospital in San Jose, Calif., Mariam Hasan, MD, the medical director of the hospitalist group affiliated with the staffing and management company CEP America, is on the fence about the early-discharge initiative she and her colleagues put in place early this year.
That initiative set a target of 60% of discharge orders filed by 11 a.m., backed by a quarterly bonus. While the doctors have been hitting the target, Dr. Hasan also watched length of stay creep up to be a half-day longer.
But then the May data arrived, “and we brought length of stay down a whole day from 5.6 in April to 4.7,” she says. So does that mean the initiative is a success? Dr. Hasan isn’t sure yet.
For one, May also brought a lower case load and a better mix of insured patients. A new swing shift implemented that month helped physicians in the afternoon focus more on next-day discharges.
And Dr. Hasan thinks another initiative has had a big impact. “We’ve spent a lot of time re-educating physicians about how to improve their documentation.” That has, she adds, made a “big difference” in terms of improving their observed-over-expected length of stay.
“We’ve had success, but only for one month, with a lot of contributing factors,” she says. “I’ll continue to check the data and I’m keeping my fingers crossed, but my eyes open.” If it turns out that a lower length of stay can’t be sustained, it may be time, she notes, to revisit that 60% target.
Phyllis Maguire is Executive Editor of Today’s Hospitalist.
At the University of Virginia Medical Center in Charlottesville, for instance, hospitalist director George Hoke, MD, likes to see unstable patients first, then overnight admissions and then patients who may be discharged that day.
But Thea Dalfino, MD, hospitalist department chief at St. Peter’s Hospital in Albany, N.Y., visits potential discharges first. “Patients admitted overnight already have orders in for consults, and the consultants here round early,” she points out. “It makes no sense for me to round on those patients before the consultants see them or before test results come back.”
Dr. Dalfino’s group is also rescheduling rounds to better accommodate early-morning discharge orders. Previously held at 9 a.m., rounds are being moved to 10 a.m. or 10:15 a.m., depending on the unit. The target goal was for each hospitalist to try and file a discharge order for one patient before 10 a.m.
“With another hour before you round,” she says, “hopefully, you can get two done.”
At New York’s NYU Langone Medical Center, hospitalist and associate chair of quality Katherine Hochman, MD, describes the schedule that’s helped the medical center sustain a betterthan-40% discharge-before-noon rate, with patients actually out of their beds by noon.
First thing in the morning, hospitalists round on potential discharges, then confirm that those patients can leave at a 9 a.m. interdisciplinary safety huddle. Multidisciplinary rounds with patients aren’t held until 1:30 in the afternoon.
Having rounds in the afternoon, she explains, helps clinicians have a better idea of who can be discharged the next morning. “We then spend the afternoon teeing up those discharges,” she says, “writing next-day discharge summaries, equipment orders and prescriptions.”
MANY HOSPITALIST GROUPS are succeeding with early-discharge initiatives; others, not so much. Regardless of their success with initiatives, sources shared innovations they use to produce more timely, efficient discharges and to tamp down length of stay:
• Tee up discharge summaries and orders the day before. Many groups use afternoon huddles with nurses and other providers to identify next-day discharges. They also rely on interdisciplinary rounds the day before discharge to make sure insurance authorizations and home equipment orders are in place. “I’m going to dictate pretty much the whole discharge summary today,” says Thea Dalfino, MD, hospitalist department chief at St. Peter’s Hospital in Albany, N.Y. “Tomorrow, I just do a brief addendum.”
• Try piloting super-early discharges. Even though the rounders in her group each try to file discharge orders for one patient before 10 a.m., Dr. Dalfino notes that rehab facilities complained that patients being transferred weren’t arriving early enough to receive rehab that day.
In response, hospitalists have piloted what they call “super-early discharges” on two floors (out of 13). Instead of having the day hospitalists, who begin at 8 a.m., file discharge orders for patients in the morning, day doctors and ancillary services get all the paperwork in place the day before.
The nocturnists visit those patients at 5 a.m. the morning of discharge and file the order, allowing patients to be out the door by 7 or 8 a.m. The pilot has proceeded slowly, says Dr. Dalfino, because “it has a lot of pieces to it. Both overnight nurses and hospitalists had to learn to do discharges.” She has also assured nocturnists that if the volume of super-early discharges becomes onerous, she’ll hire a nurse practitioner to fill that role.
• Go easy on the whiteboard—at least in terms of posting patients’ anticipated day and time of discharge. “We used to post a sign in patients’ room that said, ‘Discharge time is by 11 a.m.,’ but we took it down,” says Dr. Dalfino. “If we ended up wanting to discharge patients the afternoon before, they didn’t want to leave. They’d point to the sign and say, ‘I’m supposed to be discharged tomorrow before 11.’ ”
Hospitalist director George Hoke, MD, of the University of Virginia Medical Center in Charlottesville, agrees. The whiteboards in patients’ rooms at his medical center have a spot for estimated discharge date. But the hospitalists don’t fill that in until the day before they believe patients will leave. Even then, Dr. Hoke notes, “we have a decent failure rate at predicting.”
• Make discharges run more smoothly. One program in place in Dr. Hoke’s center is a meds-to-beds program: A pharmacy tech takes the prescription list of patients being discharged, fills those scripts at the hospital’s outpatient pharmacy, then delivers the drugs to patients before they leave.
And to help accommodate late-afternoon discharges, the pharmacy extended its hours. “If the tech wasn’t there by 3 p.m., the pharmacy wouldn’t guarantee that scripts would be filled that afternoon,” Dr. Hoke points out. Expanded pharmacy hours have helped promote more late-afternoon and evening discharges.”
Another innovation his group put in place: An administrative assistant the afternoon before discharge sets up outpatient follow-up appointments for patients. Knowing that patients have timely follow-up, Dr. Hoke notes, reassures some doctors who may be reluctant to discharge patients that they can leave sooner rather than later.
• Make use of a discharge pharmacist. That’s been a big help at Hamilton Medical Center in Dalton, Ga., says hospitalist director Laura Conger, MD. When the group was trying to get an early-discharge initiative off the ground, the discharge pharmacist changed her workflow “to get in front of potential discharge patients earlier in the day,” Dr. Conger explains.
Even though the hospital has since scrapped the initiative, the discharge pharmacist still troubleshoots medications for patients the day before discharge, targeting prescriptions that may be too expensive for patients to buy or difficult to use, like inhalers. “She works out alternatives or gets prior authorization paperwork it the day before.”
• Standardize discharge criteria with other services. Dr. Dalfino worked with her GI department to set expectations on the diet patients will be discharged on. “They really need to tolerate only clears, not be eating full diets,” she says. “That’s supported by the literature and by our GIs and surgeons.” That’s also shaved two days off many GI patients’ length of stay.
• Employ hospitalists in the ED during capacity surges. Even as her group has maintained a better-than-40% discharge-before-noon rate, her hospital has seen a sharp rise in its number of surgical patients, says hospitalist Katherine Hochman, MD, of NYU Langone Medical Center in New York. During surges, “we deploy a hospitalist team to the ED. Even though patients are still in the ED, they’re on our service, and we start to care for them there.”
• Create a discharge lounge. When Memorial Hospital West in Pembroke Pines, Fla., instituted a discharge-before-noon initiative last year, it also created a discharge lounge. According to Jose Fernandez-Duarte, MD, the system medical director for the TeamHealth hospitalist groups within Memorial Healthcare System, that gives patients being discharged a place to wait for transportation once their discharge education and medication reconciliation is completed.
“It cuts down on the time it takes,” Dr. Fernandez-Duarte says, “between discharge orders being filed and beds becoming available.”
Just say “no” to conditional discharges. Even before Dr. Fernandez-Duarte’s group put an early-discharge initiative in place, it adopted a policy that “you cannot write discharge orders with a bunch of ‘ifs.’ You can’t meet a discharge metric with orders that say, ‘Patient can be discharged if an X-ray comes back with a certain result or if cardiology clears the patient.’ ” Otherwise, he says, it may lead to inappropriate use of the metric.
Dr. Dalfino also put that policy in place. “The only thing the doctors look is the time the order goes in, but that skews the nursing metric,” she explains. “That nursing metric is based on the time between when the order goes in and when the nurse actually gets the patient out of the room.”