MB: The genesis came from a recognition that the historical way of approaching the management of members within Gateway Health Plan and in much of managed care focused on people who were already in the hospital – trying to manage the length of stay, trying to manage the discharge to the needed lesser level of care, and doing that in an efficient manner. That did achieve a lot of savings and efficiencies over the years, but we felt it just did not get at the underlining morbidity of the member population because it did not address the reasons nor the etiology of the morbidity before it was actually manifested. So, we believed that we had to look at all the different aspects of that person’s life, including their medical history, to get an assessment of their risk factors, try to stratify the risk, and then act accordingly with our case management, disease management and high risk management programs so that we could prevent the need for acute care admissions or intervene and change the trajectory of their disease process and lower it, and thereby reduce their medical problems and reduce their utilization accordingly.

MB: We started to implement it in May 2005, having worked on it diligently within our company for over three years to build a model, train our staff and coordinate with our providers. We have contracts with over 7,000 doctors and about 105 hospitals in the state, and have about 240,000 Medicaid members in 39 Pa. counties, concentrated in western and central Pa. When members come into the plan we have devised a complex assessment tool which we administer to those members, with their consent. It was derived from asking our own staff of over 100 clinicians, nurses and social workers what kind of problems repeated, as well as a worldwide search of the best predictive questions one can ask to get at all the different risk factors that one may have in their life, including but not limited to their medical issues. That assessment involves six different dimensions of daily life: behavioral issues, environmental issues, economic, medical, social and any spiritual issues that the person would like to tell us about in their lives.

Once we get that tool administered and get the answers, it is run through a proprietary algorithm, which we created internally, to try to stratify the risk factors, rank order the involvements that are needed for that member, and then create the treatment plan to surround the member with services and support. We then track it with our case management staff: we make sure the treatment plan is actually carried out, it’s documented so that we can measure improvements and outcomes and give feedback to all the people who are involved in managing their care, including their physicians. We have approximately 75 registered nurses and about 12 Masters Degree social workers, who all work in concert with four board-certified physicians at Gateway Health Plan, and our chief medical officer, who is a board-certified family practitioner. Whenever we identify that a person should be case-managed, we get in touch with the primary care physician and we coordinate that treatment plan with them and get them to validate it, add anything that they wish to it, or identify members who should go under case management. What we’re trying to do is take the burden off the physician to take care of those elements that surround the member but aren’t strictly medical.

MB: We started out first in the pharmacy area, and some of the early results are very encouraging. We were able to reduce the medical cost trend in pharmacy from a little over 13 percent per year down to five percent per year between 2004 and 2005. At that time we were running about a $300 million pharmacy bill, so that’s a $24 million savings. We’re following evidence-based medicine in the use of our pharmaceutical formulary, and member satisfaction has remained extremely high. Now, we’re deeply involved in the medical/surgical part of the implementation. The most important thing we’re looking at is the admission rate to acute care hospitals – that’s the final measure to see if our interventions made any difference. When we started the program the admission rate was around 127 admissions per thousand members, and our goal is to see if we can lower that by about five percent. We’re not far enough along yet to know if that’s going to be achieved, but we think it is achievable. We’re closely tracking it to make sure that we have a solidly defined baseline, and then we can hopefully measure any improvements that occur as a result of our interventions.

MB: We will stay with what works and dispose anything that doesn’t. Something that has already occurred is that our case management staff felt the need to have a community repository, which gives us access to key service providers like community food banks, housing, clothing, anything that has to do with emergency shelters – the kinds of things that are frequently needed in order to support our members beyond the purely medical. We have built a database covering 39 counties in Pa. and put it online so that our case managers have access to those resources whenever they are identified as major risk factors for our members, coming at this with the belief that we have to meet the member where they are, and not simply follow our own agenda. We have to know what’s very important in their lives in order to make sure that the medical care that they and their children need is a high priority, and that the barriers to that care are eliminated whenever we can break those down. If they can’t keep their houses warm, it’s very hard to talk to them about trying to get a mammogram. We have a complete listing of all the food banks that are located in a member’s area that have been helpful to us in the past. Our members many times are connected to churches or church-like organizations, which are very involved in delivering services – transportation, clothing, food, socialization. If it’s important to our members, then it becomes important to us, in terms of organizing services that may help out with their risk factors. For instance, if they need to get two children to pediatric visits and they don’t have a car, perhaps the church van can get them where they’re going and get them back so that the children can get care. We try to plug in to the resources that they, themselves have already validated as sources of care that they would prefer. The community repository is putting the tool in the hands of skilled clinicians, so that when a member tells us they don’t have food or they don’t have clothing for their children, we have a reservoir of capability in each county to help make sure they get what they need.

MB: It’s moving away from a purely medical model that just addresses the medical needs of our members. In a Medicaid population the over-riding factor in their lives is poverty, and they have barriers that people who are of middle class or higher incomes typically do not have. When we look at all the different aspects of their life, what we’re trying to do is find risk factors and problems that inhibit their ability to get the health care that they need, separate from the actual delivery system that they might access on the medical side. So, if they’ve got a behavioral health problem, if they’ve got an economic problem that’s particularly acute, if they’ve got a very unsafe neighborhood, if they can’t get to the doctor because of issues around child care – the assessment tool is meant to get at those issues so we can sequentially eliminate those problems and at the same time facilitate their access to direct medical resources. We’re tapping into community resources that are already there, have their own separate funding streams and see it as their mission to take care of the poor. We have made it our business to investigate all of those community resources, rate their capability to provide services to people in poverty, and then we build relationships with those organizations to help our members obtain the services that they provide.

MB: We either try to call them up and get them on phone, or we write them a letter and ask them to fill it out. A vast majority of the interviews that occur are done by phone. We move heaven and earth to try to find out what their phone number is, if they have one, because many times it changes. They’re a very mobile group. But we’ve had pretty good success and the higher percentage of the people we reach, the better we like it because it gives us a chance to intervene before disease manifests itself, or to catch it at an earlier stage. It’s all about preventing problems rather than simply reacting to them.

MB: We’re looking for a complex array of problems that can indicate that they’re at high risk for further deterioration or admission to an acute care hospital. For example, someone who is a diabetic who has a very elevated Hemoglobin A1c count and was recently prescribed an oral steroid. If a person is found to have those particular things happening in a fairly short period of time, they’re at extremely high risk for further medical deterioration and acute care admission if you don’t do something to change the trajectory of their disease, which obviously involves their physicians but also many of our case managers and others involved in their care. We try to find out why their diabetes is going out of control. Is it an education problem? Is it access to medications? Is it their own behaviors? What can we do to change the course of their disease now? That’s where our case managers come in and bring all the resources to bear that might prevent further problems in their lives. Another example would be if someone is socially isolated: they’re on the third floor of an apartment, they have limited mobility, they don’t get out and socialize with people their own age, they don’t go down to get the mail. That is a very bad situation and it’s very predictive of admissions to emergency rooms, independent of whatever medical problems they have.

If someone tells you that they have smokers in the house and they also have small children who have asthma – that’s considered to be a very high risk factor. Smoking in the house can be an asthma trigger for children, particularly if we notice a lot of trips to the emergency room for asthma. We’ll check to see their pharmaceutical profile: are those children identified as asthmatic? Are they getting both their maintenance meds and the rescue meds that they need for their asthma, if that’s called for? Does the family understand what the asthma triggers are? If we identify the risk factor as being smoking, it can have many downstream effects relative to the children.

MB: Yes. First of all, we’re the ones taking the initiative to try to assess all of the members’ other issues beyond the medical, and of course we draw directly upon the medical information that we and the physician have. But the physicians’ role is to help us further understand the members’ needs, from their perspective, and to help us validate the holistic treatment plan as being proper, and also to be a source of referrals for any members that they believe are at high risk. We’re trying to break down all the issues that may have to do with compliance, such as pharmaceutical regimens and compliance with going to see the specialist that the primary care physician referred the member to. The results coming back to the primary care physician should be fuller and richer because the typical barriers that are faced would have been addressed and hopefully reduced. So, they’re getting a return visit from a member hopefully who has actually been able to carry out the treatment plan, including the medical treatment plan he or she helped devise.