Published in the October 2014 issue of Today’s Hospitalist
Editor’s note: This is the first in a series on how to improve patient flow. The full series covers three themes:
PATIENT FLOW used to be the emergency department’s problem. When the ED went on bypass or the CEO received an irate patient letter about wait times, the administration leaned on the ED chair to do something about throughput. Even today, most patient flow committees are probably chaired by ED physicians.
But all that is changing. With Medicare funding cuts and the rebirth of capitation, hospitals have to transition from revenue centers to cost centers, shifting their focus from generating revenue to cutting costs. Executives and hospitalists are finally realizing what fixing patient flow should have been all along: a comprehensive, strategic initiative that hospitals need to align all their clinical departments, departmental budgets and administrative processes around to achieve.
I believe this changing paradigm creates incredible opportunities for hospital medicine. The political power within hospitals will gradually shift from the revenue generators of high-margin cases such as cardiology, surgery and orthopedics to the patient-flow process owners. And that is us.
Or it should be. But we hospitalists do not have a great track record in being part of an integrated process. Here’s hospital medicine’s dirty little secret: Besides being physically in the hospital all day, the typical hospitalist practices no differently than a traditional internist 20 years ago. The typical hospitalist rounds as a lone wolf, roaming different medical and surgical floors, interacting with nursing, care coordination and physical therapy on a random, sporadic basis.
Another problem: The medical community has ignored almost a century of evidence-based management practices. While we buy into evidence-based medicine, we have an almost complete disregard for a mountain of evidence-based management. Instead, we insist that medicine is an “art” and that there is no way we could apply business principles to medicine or treatment.
But in our group “the Adult Inpatient Medicine Service (AIMS) within Presbyterian Medical Group in Albuquerque “we are strong proponents of Lean Six Sigma and the work of W. Edwards Deming. We believe that there are no bad workers, only bad work processes and management.
We see the hospital as a factory and our hospitalist group as an assembly line that is in the business of manufacturing perfect discharges. And we have the data to back up those beliefs: In 2013, our program at Presbyterian Hospital discharged 16,500 patients, while our ED downstairs saw 68,000 patients. Since we redesigned our hospitalist program in 2009 to optimize patient flow, the percentage of patients who leave without being seen in our ED has dropped from 10% to 2%, which is respectable for an inner-city hospital.
Even more importantly, we cut our group’s average length of stay between 2009 and 2013 from 5.06 days to 4.25. That created a huge amount of capacity, added to the contribution margin, cut controllable costs per discharge “and unclogged our hospital.
We never tell a doctor how to diagnose or treat a patient. But for anything not disease-related, we try to reduce variation in the work process, then continually improve that process.
Unclogging the ED
While every department in the hospital has to be designed to optimize patient flow, unclogging the ED is a good place to begin.
For starters, you have to move your doctors to the work. That’s why we staff to match ED workload, not to make our own schedules easier.
From 8 a.m. to 4:30 p.m., we have 17 or 18 hospitalists (out of a group of more than 70) available for admissions. In addition, we have seven hospitalists performing admissions from 10 p.m. to 2:30 a.m., and our swing shift “and we have four of them “lasts from 4:30 or 5 p.m. until 2:30 or 3 in the morning. That is an odd schedule and the hospitalists in the group fought me over it, but that’s where the work is.
But putting physicians on the ground is just the first step. We view the ED as our customer, a very different relationship than most hospitalists have with their ED colleagues. We also have put several innovations in place to reduce variations in how we accept admissions and work up patients in the ED.
For one, the hospitalist group has placed ICU-trained triage nurses in the ED 24/7. When ED physicians have an admission, they call the triage nurse, who in turn uploads all the relevant patient information into the electronic health record and then sends an alpha-numeric text to the triage hospitalist.
We created two triage-physician shifts designed specifically to interface with the ED and to accept regional transfers. Because the vast majority of admissions are straightforward, the triage hospitalist accepts them electronically and gets the process started.
Or the triage hospitalist calls the ED doctor directly for more information or studies. Once the admission is in the works, the triage nurse pages the hospitalist rounding team that’s up next to accept an admission.
This allows the ED physician to make only one call for an admission, instead of dealing with any one of 18 doctors, each of whom may want to triage or work up a patient differently. Streamlining this process has removed many of the problems that crop up in ED-hospitalist handoffs.
Something else to streamline admissions: We don’t bicker or put an ED doctor in the middle while we fight with some surgical or medical subspecialty. Instead, we have clear service-line agreements with each surgical subspecialty about who will admit and who will consult.
Cutting census to reduce length of stay
To dramatically reduce our average length of stay, we likewise standardized how we communicate and interact with supporting disciplines on the floor.
We launched our first geographic unit in 2010 and now all of our hospitalists across seven of our medicine floors are geographically isolated and teamed with care coordinators. Every morning at 9:40 a.m., each hospitalist meets with all the nursing staff, therapists and care coordinators in the unit to discuss every patient during whiteboard rounds.
Moving to multidisciplinary rounds was key to improving quality and throughput. But to drive down our average length of stay, we made another major change: We cut each hospitalist’s daily census and hired more rounding teams. That’s because we were convinced that the more patients that doctors have to see, the harder it is for them to discharge patients and the slower patient flow will be.
Before unit-based rounds, each hospitalist had about 20 patient encounters. Now, with 17 or 18 rounding teams, we maintain between 13 and 16 patient encounters per day per physician, including discharges and admissions. That reduced our average length of stay by almost 20%.
We also had to make many other changes in our work processes and culture. One major change was killing off many core beliefs, such as the need to “level-load” hospitalists’ census and the drive to discharge early in the day, that most groups still subscribe to. Those changes will be the subject of my next article.
David J. Yu, MD, MBA, is the medical director of Adult Inpatient Medicine Services with Presbyterian Medical Group at Presbyterian Hospital in Albuquerque, N.M. You can send him questions or comments to firstname.lastname@example.org.