Taking on Readmissions

In our July Intelligence Report, healthcare leaders identified "all-cause hospitalwide readmissions" as the outcome-of-care measure that presents the biggest challenge. HealthLeaders Media Council members discuss how they are addressing this critical metric.

This article first appeared in the January/February 2015 issue of HealthLeaders magazine.

Michael Murphy, MD
Chief Medical Officer
Sharp Grossmont Hospital
La Mesa, California

We concentrate on congestive heart failure, heart attacks, and strokes, and what we have found and what the literature supports is that among the biggest issues are medication management and knowledge of medications. We have dedicated nurses who contact the patients and go to their homes and make sure their medications are correct, that they're taking them and they understand what they're doing.

We also deal with a lot of social factors. One is a lack of food. We connect with the food bank and social support systems.

The skilled nursing facility project is kind of the same thing. If we can go out to these SNFs and teach them when it's appropriate to have heart failure patients return to the hospital, then we can improve the care and decrease the costs for the whole system. Because when we teach them with our patients, it is going to work for other patients, too.

Our readmissions have been decreasing over three or four years, and we see a bit of a percentage decrease each year. We are putting more resources into this effort than we lose from the readmission from Medicare reductions. For readmissions, we lose $150,000, but we are putting far more dollars into it to save $150,000. If you talk to our CFO, there is no ROI and it is the right thing to do.

Stephen Moore, MD
Senior Vice President and Chief Medical Officer
Catholic Health Initiatives
Englewood, Colorado

The majority of the patients who come back to the hospitals within 30 days for all-cause readmissions are a key subset of folks with multiple comorbid diseases. They have issues with access to physicians' offices, follow-up, transportation issues, and a number of things, and this can be predicted.

At CHI we use the LACE tool—Length of stay, Acuity of admission, Comorbid conditions, and the number of ER visits in the last six months. Any patients admitted who trigger a score with this tool are focused on in a different way than we would with a low-risk patient population. We ensure that care management, pharmacy, and the physician team caring for the patients are aware of their high risk for readmissions. We also ensure a more rigorous medication reconciliation process with the patient wherever they may go afterward, whether it is a nursing home, home, long-term acute care facility, or other postdischarge disposition area.

Of all patients, those who trigger LACE are between 10% and 12% of our patient population. These are the folks who are responsible for about 60%–70% of the readmissions. We've seen our readmission rates for all-cause over the past three years drop by about 15%–20%. We expect we will see this drop even more as we get the tool fully implemented over the organization.

Jeffrey DiLisi, MD, MBA
Senior Vice President and Chief Medical Officer
Virginia Hospital Center
Arlington, Virginia

In the latest data dump from Hospital Compare, we are going to be listed as better-than-expected for all-cause hospitalwide readmissions. I attribute that to two things: rounding and case management.

We have care coordination rounds on all of our units. Every morning at about 9 a.m. on each of our nursing units there is a meeting involving physicians, nurses taking care of the patients, and case managers where we talk about every patient on every unit. One of the questions we ask is "What day will the patient go home?"

It's important because if we know that someone is going to go home the next day and we predict that correctly over half the time, then we are able to properly get all of the right things set up for the patient for that transition of care: to make sure that the transportation is set up, get the discharge instructions together, make sure the family is aware.

The second big element is we have a fantastic case management department. This comes down to making sure you have the right director of case management who hires the right case managers. Our case managers do utilization review, and they really understand the patients they are taking care of. We have great relationships with the free clinics, the subacute nursing facilities, the nursing homes in our area.

William Cors, MD
Vice President and Chief Medical Officer
Pocono Medical Center
East Stroudsburg, Pennsylvania

On getting your feet wet. We became involved in a CMS demonstration project for bundled payments. We wanted to get our feet wet to see what is actually involved in the total financial picture of taking care of an episode of care for three days prior to admission and 90 days postadmission. As a result we realized we needed a mechanism to identify high-risk chronic disease patients. We formed what we call a community care network.

On developing a community care network. We hired an internal medicine physician with a specialty in geriatrics and a nurse who assists in running a course at East Stroudsburg University, where we take premed, pre–physician assistant, and pharmacy students and we teach them to be health coaches who go into the homes of patients we have identified as high-risk.

We are running that almost as outpatient care management. As time goes on, we are probably going to have to add a social worker to the network. But we're finding that so far, at least with the COPD patients, we probably are now slightly lowering readmissions, maybe in the 5%–10% range.

On patient-centered medical homes. We're also investing in establishing patient-centered medical homes at key strategic areas of the county. Coincident with that is the establishment and expansion of a primary care health network. Without a primary care network supporting the home, you basically have a house of cards. I would love to tell you we get reimbursed for this. We don't.