The Camden Coalition of Healthcare Providers in New Jersey works to increase the quality and decrease the cost of care for high utilizers of health care, known as “superusers.” In a recent conversation with hfm, Jeffrey Brenner, MD, who serves as the group’s executive director and medical director, describes how the Camden Coalition uses data to improve the city’s healthcare delivery system
Q. Tell us about the database that drives the Coalition’s work.
A. There are two inpatient hospitals and three emergency departments [EDs] in the city of Camden, and we managed to get billing data from all of them back in 2003. This effort began as a medical student project to collect all of the billing data from these three hospitals, including name, address, date of birth, date of admission, charges, insurance information, and diagnosis codes for every Camden resident who had visited an ED or hospital in the city. We now have 10 years of that data. This is called an all-payer database.
Q. What do the data tell you?
A. We quickly learned that half of the population uses an ED or hospital in any given year. One person had actually visited an ED or a hospital 324 times in five years—and another was seen 113 times in one year. The total revenue for the Camden hospitals, for hospital and ED care alone for Camden residents, is $100 million per year. That doesn’t count things like medications provided to Camden residents and outpatient and radiology visits. I look at that number and I think of all the great things we could be doing with that money. We are not buying better lives; what we’re doing is providing sick care, not health care. A fraction of that money could buy much better primary care and much better patient education and could really do a lot for patients at the community level.
Q. What does a healthcare superuser look like?
A. The single most expensive patient—one person over five years—had accumulated $3.5 million in receipts for medical care. We found that 30 percent of the costs go to 1 percent of the patients; 80 percent of the costs go to 13 percent of the patients; and 90 percent of the costs go to 20 percent of the patients. That basic rule—30 percent of costs spent on 1 percent of patients—is true in whichever bucket you look. It’s true for poor populations; it’s true for union populations; it’s true for Medicare recipients; it’s true for the entire population. So a small sliver of patients are driving much of the cost.
The problem is that our healthcare system is designed to ignore those patients unless it can cut, scan, zap, or hospitalize them. We don’t spend a lot of time paying attention to these patients—talking to them, educating them, and making sure their needs are met.
Q. How do you use data to improve care for people with complex medical needs?
A. We run a health information exchange. We get real-time data feeds from all three hospitals for every Camden resident who goes to an ED or hospital here. Every morning, we export the raw data out, chew them up, and spit them out into a report. The report includes the name, address, primary care provider, insurance, and number of times the patient has been to an ED or hospital—not just at one institution, but in the entire city.
That is an amazing joint care management/care coordination/readmission list. We prioritize patients into a higher-risk group and an intermediate risk group, primarily based on the number of inpatient visits. So we are not doing predictive modeling; we are not guessing who will go to the hospital in the next year. We are targeting people who are in a hospital right now, and we show up at their bedsides.
Q. What services do you provide to those high utilizers?
A. If they have had four or more admissions in the past six months, they end up on our high-risk team. If they have had two to four admissions in the past six months, they end up on our intermediate-risk team. The high-risk team hangs onto them for six to nine months, because these patients have off-the-wall, complex stuff going on. The intermediate-risk team is more like a care transition team, where they work with the patients for 30 to 90 days. That’s much more like a classic care transition model.
At the front lines of our intervention are AmeriCorps volunteers—young people who have a college degree and are pre-med, pre-nursing, or pre-public health. They spend a year with us and work very closely with a team of nurses and social workers, who delegate work to them. The volunteers go with patients to their appointments, help them fill out paperwork, help them get their medication, and help them sort things out. So care navigation is a key part of this effort. Support is a key part. The clinical knowledge that social workers and nurses bring is also of vital importance.
Q. What effect has this initiative had on healthcare costs?
A. Let me give you an example. One of our patients is a 55-year-old male who was admitted with a stomach bleed and shortness of breath. He is a dual-eligible—he has both Medicaid and Medicare—and he lives alone in a high-rise apartment. This is a patient with end-stage renal disease, kidney failure, kidney cancer, hepatitis B, high blood pressure, high cholesterol, artery blockages, asthma, glaucoma, sleep apnea, and severe back pain. He takes 12 medications a day. In the six-month period before we cared for him, he had nine ED visits, six inpatient visits, and an average time between hospital visits of about 45 days.
So, in November 2011, he was enrolled in the coalition program. We went to the hospital he had been admitted to, picked him up there, and followed his care out through subacute rehab. Then he had a repeated hospitalization because of another medical issue. We coordinated his home care, transport, meals, crutches, wheelchair, and dialysis. We arranged for him to see a nephrologist, got him on the transplant list, and arranged numerous other services.
In the year before we picked him up, he had accumulated $312,000 in hospital charges and had $59,000 in payments. In the six months after we picked him up, he had no hospitalizations
Q. What is the key to success in working with high utilizers?
A. Most of this is about relationship building. I believe deeply in establishing protocols and standards. But I think that hospitals rush too quickly to implement the protocols and standards. The crux of the intervention is establishing a meaningful relationship between a patient and provider, nurse, or social worker. Also, single-disease approaches are turning out to be a mistake, because patients with complex conditions have comorbidities. If you segment your data by disease from the outset, and then build separate programs for heart failure, chronic obstructive pulmonary disease (COPD), diabetes, and so forth, you end up with very linear, narrow strategies. Sooner or later, a patient in the heart failure program is going to show up in the COPD program and the diabetes program. Nurses have ample talent, as do social workers and health coaches, to learn how to manage people with comorbidities more effectively.
We are rapidly creating standards and protocols for our work so we can train all of our nurses and health coaches to optimize management of all of these common diseases. But our first priority has been to hire the right staff for the initiative—to put the right people in the right place with the right relationship, with all the logistics and the data working to support optimal care management.
Too often, I see hospitals doing it the other way around. They build optimal disease management programs for heart failure—but in too many instances, those programs are designed just to drive more service and revenue into the cath lab. Cardiology is not going to be the place that figures out how to reduce the need for cardiology. It just wouldn’t make any sense for the department to take on this role. It’s a backward strategy and a simplistic one.
It’s not that protocols around single diseases aren’t important. But such protocols constitute the second or third phase of this work. The first phase involves finding the patients and engaging them, hiring the right staff, manipulating data in the right way, and engaging the key partners. Once you have accomplished all of these things, then you say, “Gee, the particular patients we’ve managed to pick up and identify have these sets of problems. How can we do a better job of managing those problems?
Q. After data analysis identified a high-rise housing complex as a healthcare hotspot, the coalition helped open a clinic in the building. What is your strategy there?
A. Two buildings in Camden are among the most expensive buildings in the city. Poor, elderly, and disabled residents over the age of 50 live in one building; 600 patients generated $12 million in payments over a five-year period just for their hospital and ED care. The other building, a subacute rehab and nursing home, had 300 patients with$15 million in receipts over a five-year period. And these buildings are about six blocks away from a major academic medical center.
Several months ago, one of our board members opened a two-exam-room private practice in the housing complex—with no grant money or federal funds. Based on our billing data, we were able to determine that there was a business model to be had here. About 115 patients who live in the building have switched over to this practice for their care.
Q. Does every community have a hotspot? And should clinics be opened in every hotspot?
A. So far, every community we’ve looked at—Trenton, Newark, Camden, and three counties in Maine—have all had geographic patterns of patients with high-cost, complex conditions being collected into specific buildings. That is true even in a rural state like Maine. So we think that it’s going to be a theme all over the country.
The clinic is just one of many strategies. Why move around people whose conditions are complex and expensive? It would be better just to bring care to them. But the clinic is not going to solve the whole problem.
High utilizers present a difficult and complex challenge. You need to create an entity with people who have access to data and can solve problems in real time. We have a problem-solving methodology. We tinker with data. We go out and meet people and talk to them about the data and say, “Does this make any sense? Is this what your life is like?” We meet the people who run the healthcare facilities. And then we tinker with projects. We say, “Let’s build a project with one, two, or three people, and let’s listen to the stories that come out of it,” and then scale it up from there. That basic problem-solving methodology works for complex problems that we slowly chip our way through.
Q. As other communities try to replicate what you are doing in Camden, what challenges are they encountering?
A. Trying to scale and standardize too quickly are common challenges. Very large marketplaces are not conducive to our approach; you have to divide the marketplace into smaller pieces. The city of Camden is only nine square miles and has only 79,000 people. You can’t do this kind of work in a city like Philadelphia with 2 million people. If I were in a big city, I would just bite off one area, not the whole city.
About the Camden Coalition of Healthcare Providers
The Camden Coalition of Healthcare Providers is a not-for-profit organization that works to improve the care and coordination of health care for patients with complex medical needs in Camden, N.J. The coalition is a virtual organization that is supported by all three health systems—Cooper University Hospital, Lourdes Health System, and
Virtua—serving Camden. The coalition is governed by a board of directors that includes health system representatives, physicians, nurses, social workers, and community members. Staff members provide short-term care management for high utilizers, but most of the coalition’s initiatives to improve access and capacity are in partnership with other healthcare providers
Publication Date: Tuesday, January 01, 2013