Clinician leaders at the Spectrum Health System, an eight-hospital, integrated regional health system based on Grand Rapids, Mich., have been succeeding in reducing avoidable hospital readmissions. In fact, the health system has successfully reduced readmissions for patients in their cardiac care unit in less than two years.

At the core of this accomplishment has been work that Michael Dickinson, M.D., medical director for heart failure and for heart transplants for Spectrum Health, and his multidisciplinary team have led in this area. Dr. Dickinson and his team have designed a cardiology-specific unit dedicated to the care of discharged patients with congestive heart failure [CHF] and cardiovascular diseases. Among other elements in the program, a cardiologist visits one day a week to see new and existing patients, while a heart failure-trained physician’s assistant visits on another day and provides additional input concerning each patient’s progress.

The methodology involved has attracted strong interest on the part of the health system’s cardiac surgeons, who have made it their referral center of choice for their post-open heart surgery patients. Dr. Dickinson spoke recently with HCI Editor-in-Chief Mark Hagland about this innovative program. Below are excerpts from that interview.

What are the main elements of your organization’s readmissions reduction program?

Spectrum Health is, like many [multi-hospital] systems, an amalgam; it was created by a merger and by acquisitions. Because of that, we have a very high volume; we average over 100 CHF admissions per month just at the main facility. And everyone’s concerned about readmissions requirements. To me, that’s an appropriate trigger for looking at what we do in terms of readmissions, and about quality. But it’s not easy to make progress; the reasons a patient can be readmitted can be complicated and multi-factorial; and the readmission might be caused by different things. So you have to look at different things. So we created something called “H2O”—Hospital To Office. We decided that transitions to the physician office were a key area. Our hope was that managing that area would reduce the odds of readmission; and that was based not only on our own personal impressions, but also on IHI’s recommendations on seeing a patient within 5 days of discharge [recommendations from the Cambridge, Mass.-based Institute for Healthcare Improvement]. And we found that 50 percent of our patients readmitted had not been seen by an outpatient provider.

We didn’t know whether these readmissions were just a failure of transition, or whether or not the patients involved were really sick enough that the readmissions couldn’t have been averted anyway. We didn’t know, because we didn’t have the data. And we discovered that we had a very chaotic discharge process, with patients seen anywhere from a week to a month after discharge, and that was unacceptable to us. We had had some experience with Lean and Six Sigma, and so we brought in a Six Sigma expert who created a Kaizen event for us, using Lean and Six Sigma principles, and we took this on as a project, with the goal of rapid-cycle improvement.

Michael Dickinson, M.D.

What was the timeframe around this activity?

This started in November 2009, when the Six Sigma expert came in. And we decided this needed to be a partnership between the hospital and the physician group, and we also decided to bring in Priority Health, a large insurer, and one of the parts of our system—it’s the second-largest third-party payer in Michigan after Blue Cross. At that time, the medical group was an independent private-practice group; in January 2010, Spectrum acquired the group.

So we discovered that our discharge process was chaotic; we also realized that with traditional office scheduling, we couldn’t work this out. Our physicians were booked out three to six months in advance with appointments, and even our mid-level providers were booked out for two to four weeks. And how do you coordinate that? Because you may decide tomorrow afternoon that a patient may need to go home, but you don’t even have the system set up to arrange that follow-up appointment. The process was inefficient, and involved too many steps.

So we took one of our experienced schedulers from the office and moved her into the hospital, so now, whenever we have a patient in the hospital, we send a message to her to arrange a follow-up. We also took hospital-based physician assistants, and partly shifted them over to the office. Three of them spend a day a week or part of a day a week in the office, and that shift created enough timeslots, as well as improving satisfaction for the PAs in terms of the variety of their workload. Another factor here is that office-based  PAs aren’t usually used to seeing patients who have just been discharged.

So this improved the available expertise, correct, since the formerly hospital-based physician assistants are now office-based?

Exactly. And also, a lot of this involves just making sure the patients understand what they need to understand at discharge, so they’re compliant with their medications, and so forth. What’s more, a lot of times, if a patient doesn’t have a family member to help them, like an adult son or daughter to pick up a new prescription, they’ll just go back to what they were doing before.

And we now have an experienced scheduler in the hospital, and she has her scheduling system set up on her own screen, and on the other screen is the Cerner discharge system. And she will look at the two screens and make sure that she schedules Mr. Smith as he’s getting discharged.

So on one level, it’s simple, but processes and workflows are so complex in healthcare and have to be worked out, right?

Exactly. That’s why we needed a systems engineer to help us work all this out. So instead of having six RNs, who are higher-paid, to stand on hold in the hallway trying to get appointments, now we have a system where people can send messages for appointments, and we’re paying a scheduler, not clinicians, to do this. And we’re happy, because we really don’t want to stand around in hallways on hold.

So that was the H2O initiative. It clearly reduced variation, and we have a very high rate of seeing our patients within two to five days. The H2O initiative did reduce variability, and also the us-to-us readmission. It doesn’t cover every single practitioner in the hospital. The next phase is to reinforce the standard of care where every patient goes home with a full arsenal of support.

Also, we learned when we studied the situation that we were reducing readmissions, but that we were losing in the area of patients being discharged to subacute rehabilitation. They go to a rehab hospital, where they get up to two hours a day of therapy, and they’re too infirm or fragile to go home. And in the subacute rehab facilities, no one is seeing them [for follow-up]. Now, you could insist that those patients make [primary care physician appointments], but we tried that, and it was a fiasco. Suddenly, you were forcing them to get into a Heuer lift and an ambulance, and creating great disruption to those patients. So we opted for a phone-based follow-up with them instead, so a nurse is calling up that kind of patient and going over their medications with them in that way.

The second step was actually taking on [the processes in] subacute rehabilitation. We actually own a subacute rehab hospital; so we called them up and met with them. And we discovered that their stays were more like nursing home stays than like rehospitalizations. Here’s one important example, with regard to CHF patients. Now, here’s the thing: in a hospital, we can enforce certain health requirements, such as dietary, but not in rehab. By regulation, they’re not allowed to impinge on patients’ rights, and they can order anything off the menu in the rehab hospital. So it became clear that they needed to shift their model to become educators and advocates, and that ended up being at the nurse level. So we educated their nurses about heart failure, and brought in our team, and created a series of classes on heart failure and why sodium restriction was important. We also asked them to be educators for the patient, modeling it after telehealth programs. In telehealth, you model behaviors for patients and explain things to them. And we developed a flowsheet with a questionnaire, and encouraged the nurses to do blood pressure and weight checks with the patients every day and to rank them low, medium or high; and we also encourage them to help the patients make dietary choice as well, and that’s been very effective as well. That program has been live for about a year now.

What should CIOs and CMIOs know about all this?

I think they should understand that what I call transitional care, or how the patient transitions from inpatient to outpatient, can have a significant impact on readmissions. And looking at processes is really valuable. And I would recommend that they bring clinicians and administrators together around a table to look at this. And if they’re able to find mechanisms within the information system to support this, they’ll find clinician champions.