A recent study published in the Journal of Hospital Medicine examined patient discharges from hospitals in Ontario, Canada, to determine if morning discharges were associated with positive outcomes. Some hospitalist programs have embraced discharge before noon (DBN) initiatives like those studied in the article.1 Unfortunately, the researchers concluded that the Canadian DBNs did not positively impact hospital length of stay, readmissions, or mortality rates.
DBN has been a quality improvement target for hospitals hoping to improve throughput and free up scarce beds, while promoting patient safety by encouraging discharge as soon as patients are ready to leave. Yet other researchers have questioned its actual impact on quality metrics. One author called DBN’s purported impact an “urban legend,”2 while a JHM editorial accompanying the Ontario study noted, “Hospitals are delicate organisms; a singular focus on one metric will undoubtedly impact others.”3
Might DBN be an artificial target that doesn’t actually enhance throughput, but leads instead to unintended consequences, such as patients being held over for an additional night in the hospital, rather than being discharged when they are ready to go on the afternoon before, in order to boost DBN rates? A perennial debate in hospital medicine is likely to be reignited by the new findings.
‘No significant overall association’
Quality improvement initiatives targeting morning discharges have included stakeholder meetings, incentives programs, discharge-centered breakfast programs, and creation of deadlines for discharge orders, the new study’s authors noted. Although these initiatives have gained support, critics have suggested that their supporting evidence is not robust.
The Canadian researchers retrospectively reviewed all patient admissions to general internal medicine services (GIMs) – largely similar to hospital medicine services in the United States – at seven hospitals in Toronto and Mississauga over a 7-year period ending Oct. 31, 2017, counting all of these patients who were discharged alive between 8 a.m. and noon. DBN averaged 19% of total live discharges across the diverse hospitals, with their diverse discharge practices.
But they found no significant overall association between morning discharge and hospital or emergency department length of stay. “Our findings suggest that increasing the number of morning discharges alone is unlikely to substantially improve patient throughput in GIM, but further research is needed to determine the effectiveness of specific interventions,” they concluded.
“We used a very narrow lens, looking specifically at throughput for the hospitals and emergency departments and whether DBN makes it more efficient,” said corresponding author Amol Verma, MD, MPhil, FRCPC, clinician-scientist at St. Michael’s Hospital, University of Toronto, in a recent interview. “What we found was that, on days when more patients are discharged in the morning, patients do not flow more quickly through the hospital. That suggests that increasing morning discharges is unlikely to make a difference.”
What does DBN really mean?
The semantics of DBN deserve further exploration. Is DBN about the actual hour of discharge, or the time when the hospitalist signs a discharge order – which may be well before the patient actually gets a wheelchair ride down to the hospital’s front doors? And if DBN is an organized program promoting morning discharges, how is it incentivized or otherwise rewarded?
Other factors, such as arrival of medications from the pharmacy or results from clinical tests, access to an ambulance if needed, transport to the front door, and bed cleaning will impact how quickly a doctor’s discharge orders get acted upon – and how quickly the newly emptied bed is available for the next occupant.
The clinician’s views on discharge practices may diverge from hospital administrator or health system perspectives, with its imperatives for efficient throughput in order to bring in more patients, Dr. Verma said. The hospitalist is also concerned about whether the patient feels ready to go home. “We can all agree that patients should leave the hospital as soon as they are medically able to do so,” he said. Longer hospital stays are associated with increased rates of hospital-acquired infections and other iatrogenic complications.
But there is not agreement on the components of a safe discharge – or on the other dimensions of effective patient flow and transitions of care. How do we optimize treatments initiated in the hospital? Does the patient need one more CAT scan? And what about the concerns of patient-centered care? Does the patient have a caregiver able to help them when they get home? There is a lot of uncertainty, Dr. Verma said. “These kinds of decisions have to get made many times every day by hospitalists,” he noted.
“We find ourselves trying to mirror the ebbs and flows of the emergency department with what’s happening in the hospital,” said Venkat Gundareddy, MBBS, MPH, associate director of the division of hospital medicine at Johns Hopkins Medicine in Baltimore. “The majority of hospital discharges happen during business hours, but the emergency department doesn’t stop admitting overnight, thus creating a throughput challenge.” Discharges are also based on clinical outcomes and on patients transferring to other facilities that prefer patients to arrive earlier in the day.
“Hospitalists may not fully appreciate these dynamics, because we’re siloed on our units,” Dr. Gundareddy said. “There is a subset of patients who would fit the bill for early discharge, but other patients come into the hospital with greater complexities, and a need for more coordination. Their discharges are harder to predict, although it gets clearer as their care progresses.”
The hospitals included in the Ontario study are at 90% -100% capacity, so their flexibility is constrained and throughput is a critical issue, Dr. Verma said. “But if you start with the target of more efficient throughput, there is no logical or practical reason to assume that discharge before noon would help. If we believe someone is ready for discharge based on physiologic changes, their response to treatment, and the conclusion of medical investigations, none of these conform to the clock. It’s equally likely the patient achieves them in the afternoon or evening.”
Other views on morning discharge
An alternative perspective comes from New York University’s Langone Medical Center, which has published positive results, including earlier subsequent arrivals to the inpatient unit from the emergency department, from increasing its hospital’s DBN rate.4
The hospital has continued to encourage morning discharges, which have consistently run 35%-40% or more of total discharges on two acute inpatient units at Langone’s Tisch Hospital. A previous study described the multidisciplinary intervention that resulted in a statistically significant increase in DBN – from 11% to 38% in the first 13 months – while significantly reducing high-frequency admission peaks.5
“We’ve been doing DBN for a number of years,” said Benjamin Wertheimer, MD, a hospitalist at Langone Medical Center and one of the studies’ authors. It is an achievable – and sustainable – goal. “Many hospitals around the country have problems with the flow of patients. Many hospitals are full – even before accounting for the COVID pandemic.” There is good evidence that, for a patient who no longer requires hospitalization, getting them out as early as possible, with a safe plan for their discharge, is a good thing, he said. “We see DBN as an important operational metric.”
If the necessary work is done correctly on the afternoon before the discharge, then a DBN approach can push communication, coordination, and advance planning, Dr Wertheimer said. Otherwise, essential discharge tasks may lag until the last minute. “We try to put the pieces in place the day before through a better planned process. But it should never be that DBN takes precedence over when the patient is safely ready to go,” he said.
“Our true measure of success would be how well we are preparing, communicating, putting safe plans into place,” he added. “DBN does not in and of itself answer all the safety and quality concerns. We set priorities around specific quality targets. DBN is just one of our operational and safety measures.”
The DBN intervention at Langone started with a multidisciplinary kickoff event in which all team members received education on its importance, a clear description of roles in the DBN process, and a corresponding checklist of daily responsibilities. The checklist was utilized at newly implemented afternoon interdisciplinary rounds, scripted to identify next-day DBNs, and make sure everything is in place for them, he explained.
“We provide daily feedback to floor staff on the DBN percentage, celebrate success, and offer real-time opportunities for case review,” Dr. Wertheimer said. “We have been careful about how we message this goal. Quality and safety come first, and we want to be prepared for discharge in advance of when the patient is ready.”
A boost for discharges
Mark Williams, MD, MHM, recently appointed chief of hospital medicine at Washington University School of Medicine in St. Louis, and a principal investigator for Project BOOST (Better Outcomes by Optimizing Safe Transitions), SHM’s quality improvement mentoring initiative aimed at helping hospitals improve care transitions, said that debates about DBN have gone on for a long time in hospital medicine.
“Around 2002, consultants told the CEO of a community hospital affiliated with Emory Healthcare that if our hospitalists could discharge patients before noon it would improve throughput,” he recalled. The consultants came from the hospitality industry, where DBN is easier to achieve.
But in hospital medicine, he said, “We use the whole day of the discharge in delivering care. I said to the CEO, ‘I can get you 100% discharge before noon – I’ll just hold the patients overnight,’” he explained. “In our initial experience, we pushed DBN up to about 10% -15%, and it opened up a few beds, which rapidly filled.”
Project BOOST encouraged the goal of getting patients ready to go out as soon as they were clinically ready, but did not advocate specifically for DBN, Dr. Williams said. “The problem is that hospital throughput starts to gum up when occupancy goes over 80% or 90%, and many academic medical centers regularly reach occupancy rates greater than 100%, particularly in the afternoon.” The deluge of patients includes transfers from other hospitals, postsurgical patients, and admissions from the emergency department.
“Boarding in the ED is a real issue,” he said. “Right now, it’s a crisis of overoccupancy, and the problem is that the pipeline is pouring patients into the system faster than they can be discharged.”
Dr. Williams believes there needs to be bigger thinking about these issues. Could hospitals, health systems, and hospitalists practice more preventive medicine so that some of these patients don’t need to come to the hospital? “Can you better address high blood pressure to prevent strokes and make sure patients with heart disease risk factors are enrolled in exercise and nutrition programs? What about access to healthy foods and the other social determinants of health? What if we provided adequate, consistent housing and transportation to medical visits?” he wondered.
Hospital at home programs may also offer some relief, he said. “If suddenly there weren’t so many emergency room visits by patients who need to get admitted, we’d have enough beds in the hospital.”