Readmission rates, LOS decrease with BOOST
Teamwork, accountability key
He believes it's the right thing to do for the patient and the right thing to do for the hospital. And with staff happier, patients happier, and length of stay and readmission rates decreasing, it seems as if it's working. In 2008, Piedmont Hospital in Atlanta chose to implement one of the myriad transitional care models out there today. Their choice: Project BOOST (Better Outcomes for Older Adults Through Safe Transitions) from the Society of Hospital Medicine.
Matthew Schreiber, MD, chief medical officer for Piedmont Hospital, says before adopting BOOST, the hospital was working on process improvement and lean techniques. At a meeting, he half-jokingly asked for his very own unit one "where I could bring some new ideas and just see what happens." To his surprise, the chief nursing officer took him up on it. He was given a 20-bed general medicine hospitalist unit, 6 North.
Before adding the discharge model, the unit had asked and answered some very basic questions: "How do we restructure the patient's experience from door to discharge in a patient-centered way, in a way that makes sense and is most efficient, and how can we get rid of waste and all of that stuff?" he says.
Staff knew that discharge was going to be an integral component of improving quality care, he says. About three weeks after opening the unit and working on care processes, an administrative director told Schreiber about the BOOST project, saying, "This is something we need to do." The hospital was selected as one of the first six sites to pilot the project, which just recently went statewide in Michigan.
What distinguished 6 North was that it was primarily overseen by hospitalists dedicated to that unit, which Schreiber says removed the time doctors spend logistically moving around the hospital. "Almost an hour of their day was just spent going from one place to another, which is not value added for anybody." But for him the unit wasn't just about hospitalists; it was about the entire care team and creating an environment where teamwork melds with structuring responsibilities so that the team knows who is responsible for what and everybody is responsible for something.
"We got all of the participants, all of the ancillary services, nursing, respiratory therapy, physical therapy, all around the table, and we said, 'OK, if you were king for the day and you wanted to make the process better, what would you do? What could you offer? What jobs do you find yourself doing that you feel like you're not the expert in? What jobs do you see other people doing that you feel you are the expert in? How could we make this a better experience for the patient? And the only rule here is everybody has to contribute something.'"
The team looked at care processes and streamlining those to decrease inefficiencies. One process improvement project included "specialized testing triage." For example, Schreiber says, let's say you have three patients who need MRIs. One needs an MRI for spinal cord compression, a medical emergency. The second had a transient ischemic attack, in which case, he says, a normal MRI would mean he could go home. The third has osteomyeletis, which is not a critical need. The radiology department schedules the orders and puts the osteomyeletis patient first, the spinal cord patient second, and the brain patient third. But someone from the unit could call radiology and change the schedule to meet the most urgent need. That person would call the radiology department and say, "Move number two to number one. Number one to number three. And number three to number two.
"So the time slots have been assigned already. We're just retriaging the testing for the most appropriate orders. That also takes into consideration a patient's discharge readiness," he says. The team would also be able to work with radiology on reducing cases if the department is overbooked. For instance, the osteomyeletis patient could be booked for the next day, but the other patients need to be seen that day.
"So it really cuts down on the waste for everybody when there's good two-way communication; everybody knows what needs to be done. It gives you the opportunity to begin to negotiate. 'I want you to do this for me. I want you to do the acute cord study right now because that's what the patient needs, and what I'm going to do for you is I'm going to identify the guy you could bump until tomorrow without any pain and suffering,'" Schreiber says.
"We look for those win-win opportunities so that we can create a better environment, more efficient for the patients and quite honestly get more work done in less time. And we were very successful in that," he says.
Assigning clear responsibilities was a major focus of improvement. And Schreiber says it all starts with communication. For instance, the team noticed that patients were being delayed because sputum samples were not being processed in a timely fashion. "Then we start digging. 'OK, why aren't sputums done in a timely fashion?'" he says. The team found they weren't being collected. And when asked why, nurses looked at respiratory therapists and RTs looked at nurses and both said: "I thought you were doing it."
"So we said OK, from here on in RT is now responsible for doing all sputums." RT notifies nursing every time a sputum sample is being collected. If RT is unsuccessful in collecting the sample after three attempts, they place a flag in the chart notifying whoever is looking that a sputum sample was not collected.
The unit also has a whiteboard in the nursing station with patients' names, their main diagnosis, and specific barriers to discharge. For instance, instead of writing that patient Joe Smith needs three sets of cardiac enzymes and a stress test, it would note that patient Joe Smith's next set of cardiac enzymes is due at noon and if the results are negative, he needs a stress test. "So if at 10 past noon we don't have the results of the cardiac enzymes, we can be on the phone saying, 'Where are those test results?' And if the results are normal, then we're making sure that the stress test is done in a timely fashion. That's good for patients. It's good for the staff. It's good for everybody," Schreiber says.
At first, the whole team would huddle at the whiteboard. "It became abundantly clear very quickly that PT would tune in when we were talking about PT needs, and RT would tune in when we're talking about RT needs, and case management would tune in [when we're talking about case management], but otherwise they were just kind of politely standing around for 10 minutes." So they did away with the team huddle and instructed staff "whenever you set foot on the floor, just come in and check in with the board and the board manager. So this person, our nurse, is aware of the status of all patients at all times. And everyone can go to the board and see what's important to them to help them prioritize their work. And so people check in with the board 20, 30, 40 times a day to know where are we at with their patients, what's happening next, etc."
Implementing transitional care model
Piedmont's success with Project BOOST started before the model began because, as Schreiber says, the unit had already looked at process improvement, efficiency, and accountability. Those underlying processes should be in place, he says, before you implement the tools. "[W]e did the workflow process issues first and then we took these tools when we were primed and ready.
"For me, personally, it was like somebody just did the homework for me. I knew that we needed a new discharge form. I knew we needed a risk assessment tool for patients that are at risk for failed discharge. I knew we needed all of those elements and here comes this wonderful BOOST toolkit," Schreiber recounts.
Included in the toolkit is a risk assessment form called 7P that looks at issues such as polypharmacy, how many medications the patient is on, the patient's degree of sophistication, and his or her family support. There are also checklists for logistical issues. "Does that patient have their keys? Do they have steps in their house that they can get up and navigate? There are a lot of questions that are really important for patients having a successful discharge but don't really get up into the front of doctors' minds as a medical thing. We think about their blood pressure medicine not how many steps they have at home," Schreiber says.
There's a discharge checklist similar to a preflight checklist, he says, with reminders: Before the patients walks out of the door, have you done all these things? And there's the patient Pass form a discharge form written in patient-friendly language.
A big part of the BOOST intervention is communication using the teach-back method. Christopher Kim, MD, MBA, a hospitalist at the University of Michigan and director of the statewide collaboration using BOOST, says when patients are given information they may say they understand. "But if you really probe them further, they might not have as good of an understanding. Using the teach-back method, the concept is to explain something to the patient and then ask them how well we may have done at explaining to them and ask them to repeat it back to us. And if they could teach it back to us then it helps us to be more reassured that we did a good job and that the patient has a good understanding. If they do not understand and they can't teach you back what you just said to them, then it identifies some gaps and perhaps it's another opportunity to intervene again."
At Piedmont, when a patient is ready for discharge, the clinician visits the patient with a blank discharge form and goes over elements with the patient before the form is filled out. So a conversation might be:
Clinician: "Ms. Smith, why were you in the hospital?"
Patient: "I don't know. The doctors were concerned about me."
Clinician: "Ms. Smith, remember how your legs were swelling and you were short of breath?"
Patient: "Oh yeah, that's right. That's why I came to the emergency department."
Clinician: "We call that congestive heart failure. Ms. Smith, can you tell me why you were in the hospital?"
Patient: "Oh yes, I was in the hospital for congestive heart failure."
Clinician: "Great. I'm going to write this down so you remember."
There is a section on the form for red flag symptoms. So a clinician may say: "Ms. Smith, remember how we were concerned about your legs swelling on you? If your legs swell on you then I want you to call your primary care doctor because they're going to need to make an adjustment in your water pills. I'm going to write that down for you. If you're short of breath, then you should call 911. And I'm going to write that down for you. Ms. Smith I'm going to write down all the care providers that you've seen here: your hospital doctor, your primary care doctor. I'm going to write down your pharmacy number. We've made all of your follow-up appointments, and in this section is your homework. You need to go over your echocardiogram with your primary care doctor. That's where you need to adjust your water pills and check your potassium again."
Schreiber says the form was created as a "memoir" for patients, something they would stick on their refrigerators at home. But he found that patients became so attached to the form that they took it with them to their visit with their primary care doctors, which is scheduled for the patient before he or she leaves the hospital.
"It's a big win for so many reasons. Even though the discharge summary may have been dictated, the primary care doctor may not have it in front of them when the patient shows up. So now they've got something that says, 'I was in the hospital because I had congestive heart failure. These were the major things that happened to me when I was in the hospital. These are the things that I'm supposed to follow up for today.' So now they have a much more productive, engaged visit with their primary care doctor, which is part of the essential theme to keeping people out of the hospital."
Results speak for themselves
On the unit are patients seen by the hospitalist group and others not in the group. Thirty-day all-cause readmissions for patients 69 and younger, Schreiber says, for the non-hospitalist patients was 25.5%, and their case weight was 1.16. For the hospitalist patients, the rate was 8.52%, and their case weight was 1.15. "So very similar populations, similar diagnoses, and dramatically different results," he says. The length of stay for the non-hospitalist group was 4.96 while the hospitalist group's length of stay on the unit is 4.09
Along with participating in the BOOST project, Piedmont had to provide certain data including patient satisfaction, length of stay, and readmission rates. "Readmissions were a big challenge for us as an institute, and I should tell you that our readmission rates are probably a lot better than even the numbers I've suggested because our readmission rate is anybody who is put in a hospital bed, whether they were on observation status or inpatient status. And if they even return to the ED even though they weren't readmitted, it counts in the 30-day readmission rate. And that's because while the government may be deciding how to pay you based on your readmission rate, you haven't won the game unless you've done right by the patient, unless you get them out of the hospital and you keep them out. If they to come back to the ED every three days but they never got readmitted, have you really done the right thing for the patient? So for us we took the broadest definition possible because that's the goal we want to achieve."
The team now does a root-cause analysis of patients readmitted within 72 hours. Everyone is called together at least once a week with the hospitalist nurse and home health to discuss possible interventions that could have helped. What were the primary issues? Was this something that was preventable? Not preventable?
"Anyone who comes back within 72 hours gets a very focused review on 'OK, something happened for that to happen. Was it they were discharged too early? Was is that their home health agent didn't get out to them? Did they not understand their instructions? Did we prescribe them something they can't afford or any of those elements so that we can learn those things and incorporate them so that it won't happen again?'"
The clincher in success? Schreiber says, "We created a team environment that's nonconfrontational, that's totally focused on the outcome. We've decided that nothing matters, not our egos, not anything, except for the patient outcome. So when they see that dot starting to move, it really revs them up."