Setting a target time of day for discharging patients, like noon, has become commonplace in hospital medicine.
But this goal turned out to have no effect on efficiency and raised concerns and complaints from hospitalists, according to a recent randomized controlled trial at three U.S. hospitals published by the Journal of Hospital Medicine on Feb. 16 and summarized in the March 1 ACP Hospitalist.
Over six months in 2021, the researchers randomized 30 attendings to prioritize seeing patients who were expected to be discharged that day before seeing other patients. On average, clinicians entered discharge orders at 1:03 p.m. and patients were actually discharged at 3:22 p.m., times that were very similar to the times of 1:11 p.m. and 3:21 p.m., respectively, found in a control group of 31 physicians. Length of stay didn't differ either, and in surveys, hospitalists in the intervention group reported negative effects on their workflow and concerns about patient harm.
ACP Hospitalist recently talked to the co-lead authors of the study about what this means for current and future throughput initiatives. Marisha Burden, MD, FACP, is a professor of medicine and division head of hospital medicine at the University of Colorado School of Medicine in Aurora. Angela Keniston, MSPH, is an assistant professor and the director of data and analytics for the division of hospital medicine at the University of Colorado School of Medicine.
Q: What motivated this study?
A: Dr. Burden: It's a career's worth of experience trying to improve hospital throughput. We previously conducted a study, published in 2018, looking at barriers to discharge. We found, not surprisingly, that clinicians are busy doing many things. As a clinician, you have to try to get everything done—seeing the patient so they are prepared for what the day has in store for them, meeting with care management, social work, getting discharges in—there is this perfect storm of trying to get everything done, ideally before noon. That study did show that there were several factors that may contribute to earlier discharges. One of those was workload, one was whether you prioritize discharging patients first, and the last was whether or not you were on a teaching team. The next step logically was a randomized controlled trial to actually start to build evidence-based practices.
Q: Were you surprised by the trial's results?
A: Dr. Burden: As a clinician, I was not surprised at all. But the beauty of a randomized controlled trial is that even a negative finding is important. We looked at it in so many different directions, I think we can safely say that prioritizing discharging patients first did not improve throughput and caused a lot of unintended consequences. … From the response on Twitter, I think most clinicians are not surprised by the results. Some of the initiatives that we do, and this is not unique to one institution or another, sometimes feel low yield. The intent is good. Everybody wants to get patients out as quickly as possible. But sometimes we get so caught up [in that goal] that any initiative is better than nothing. In reality, what we see in the study is that maybe it even causes harm.
Q: What were some of the important takeaways from study physicians' comments?
A: Dr. Burden: Participants said, “It made me do rework.” “It frustrated me.” There's the travel—you lose some efficiency. But it wasn't all negative. I do think the takeaway from the comments, as well as the results, is clinicians are very good at prioritizing what they need to do. There are competing priorities. If you have a really sick patient, or you have to travel to another floor and there are two patients there, maybe you don't need to go see the next discharging patient first and then walk back over to the other unit. You have to allow the clinician, who is the expert, and the team of experts with them, to decide what makes the most sense.
Q: Do the results indicate that hospitals should give up targeting a certain discharge time?
A: Ms. Keniston: We like to say, “Discharge when the patient is ready to leave.” And make sure that the infrastructure in place actually supports clinicians such that they might be successful in doing so.
Dr. Burden: We need to foster systems that allow for real-time care. There's a lot of energy needed to get someone out of the hospital, in particular in the evening hours. … There's all these barriers to real-time care. Next steps are to figure out what are those work structures needed to facilitate real-time care and that allow clinicians to do their best work.
Q: Hospitals are likely to continue to focus on increasing efficiency and shortening length of stay. Do you think there are better ways than prioritizing early discharges to do that?
A: Ms. Keniston: We do—100%.
Dr. Burden: What the evidence suggests to date is that we have to look at work structure, and a big part of that is workload. There's an interesting conundrum. In the business world, we sometimes think if you lower a clinician's workload, it's more expensive. But what we see in the hospitalist world increasingly is that there is a workload and work structure sweet spot. There's a point of too much work where you lose efficiency. There's also a point of too little. I don't think clinicians want too little work. They want to be utilized for their skill set. So how do you get to that sweet spot where you optimize efficiency, worker well-being, and patient throughput? That is not what we're doing today, by any stretch of the imagination. So I think efforts need to focus on specifically looking at work structure and how that drives core outcomes in three main categories—worker, patient, and institution—and applying rigorous methods to do so, not just making changes to the way that we structure work or implementing a given initiative without being really rigorous in our evaluation of that.
Q: Has this study led to future research questions for you?
A: Dr. Burden: How do we build those optimal structures, and how do you do it in a way that it's financially viable? And then how do you make sure that it's not harming either patients or workers? Our dream is to do more and more randomized controlled trials and build the best practices around clinical operations, because it's such a key component to patient outcomes, worker outcomes, and institutional outcomes. So yes, there'll be more to come.
Q: Is there anything else you want hospitalists to take away from your research?
A: Ms. Keniston: Don't be afraid to evaluate those initiatives that seem like foregone conclusions. The study in 2018 looking at the barriers to discharge wasn't funded. We just had this question we wanted to answer, and we pulled together a multicenter team of people who were also interested in answering that question. From that jumping-off point, we have lots of other work that we're doing and grant funding that we're getting. So I would say don't be afraid to challenge the status quo and make friends with a methodologist or statistician and figure out how to evaluate things rigorously, so that we know we're putting our time and our resources and our energy towards things that are truly going to make a difference for patients and for clinicians and institutions.