Implementing the Oncology Care Model Improves Value and Reduces Costs

An Interview with J Russell Hoverman, MD, PhD, the vice president of Quality Programs at Texas Oncology and the medical director of Managed Care for The US Oncology Network.

Given only three months between notification, approval, and launch of the oncology care model (OCM), what were some of the difficulties in creating the internal pilot program and anticipating that the needs of the model were met?

Well, first was selecting sites.  They had to be small, they had to be fairly well-organized, sympathetic to the program, willing to be guinea pigs.  So, we couldn’t do it at the huge sites so we, for the most part, took medium-sized sites, anywhere from 5-8 clinical oncologists, usually were the radiation oncologists.

And then the next big part of it was to determine what we had as part of the requirements that what we had in place already that may need some enhancements and what we didn’t have at all.  And if we didn’t have it, could we start building it, or do we start looking for places where we could get it.  It was pretty clear that the Institute of Medicine 13 requirements for their care plan was something that we hadn’t, did not have in an organized fashion, although almost all of us thought that we did those pieces.  And the same is true for the navigation.  The questions for us were, how do we break up the Institute of Medicine care plan into a piece that we could start documenting and the same was true for the navigation.  We felt that we were doing navigation to a great degree, but certainly had it in a fashion that was not as organized as the OCM was asking for, and we had to begin to see how we would change that.

The other piece that we didn’t have was to figure out who belonged in the program and this turned out to be problematic.  We can talk more about that one later.  But just determining who belonged in the program, especially with oral medications turned out to be a difficult issue.  

How did data reporting in particular, and documentation, feed into the challenges surrounding navigation, quality improvement, and incorporation of the Institute of Medicine’s Care Management Plan?

Well, as I said there was no organized data recording or reporting for either the IOM care plan or navigation. So, we developed a treatment plan document that could have its components checked off into our electronic medical record, and we also developed a navigation drop down. We had to then identify who would be responsible for those, so the physician would be responsible for a number of components for the treatment plan, but other folks could be responsible as well including advanced practice providers, nurse managers, but primarily physicians, and we expected most of the treatment plan to be completed by the physician at the time that chemotherapy was ordered.

For the navigation drop down, that again could be a number of folks who would be responsible for that, but somebody had to own it.  The nurse manager could own it, the physician, nurse could own it, but somebody had to own it, to be responsible for completing it.  And again, those are work.  These are more than check boxes and it doesn’t flow naturally to have those check boxes.  And then we have to have a way to retrieve them to be sure that we’re doing it.  And that was not available at the time either, so we had to work to develop that.  

What was the impact of the administrative burden?

Sure.  The big one, the big administrative burden, first of all, was getting people enrolled and getting MEOS (monthly enhanced oncology service) payments billed.  So that’s identification, billing, and collection.  And there was no structure for that in our business office so it allowed us to, the pilot allowed us to understand that we would need those kinds of staff and we started recruiting before the program started to have, to beef up the business office to be able to do that.  It turns out that’s, probably in terms of new personnel, that’s probably the biggest addition that we’ve had.

In terms of the other administrative burden for the navigation and the island of care plan, some of that has been shifted to the advanced practice providers, but some of that takes away from the nurse managers’ or the nurses’ usual business, so it’s an add-on.  It’s not to take away, but it adds on to the workload for the RNs and the nurse managers.  It’s still unclear about how much that is, whether we had to hire nurses to be sure that people responsible had enough time to be able to do that.  We haven’t had “navigators” per se.  We have added other personnel to assume some of those responsibilities like counselors, and particularly social workers.  

Do you anticipate the increased emphasis on clinical pathways to be positive, problematic, or both for your programs?

We were at the start, a pathways-supportive organization, medical oncology even before all of this was 80% on pathways or so.  It was probably even a little bit more.  And we’ve always been supportive of pathways, so pathways is key to patient education, to cost savings, to safety.  But it’s also key to building care paths that include risk assessment, a specific focus on high-risk patients to keep them comfortable and out of the hospital, to trigger palliative care consults, for instance, to begin to develop some guidelines about, and when, hospice would be appropriate.  So, there are a number of ways in which the pathways end up being the backbone for a more inclusive, extensive program which we feel we had to build.  At some point we have to build imaging around pathways.  Radiation pathways already exist and we have to be able to build those to complement the medical oncology pathways.  There’s every reason to think that we have to bring the best evidence to bear as to what’s right, and then we have to bring the best evidence to determine what’s the most valuable to the patients and for their families and for us.  

As claims data and outcomes data become available from your participating practices, what do you anticipate will be the biggest issue arising from the OCM?

The first thing is we started seeing, so, we’re a private medical oncology practice.  We’re not an integrated network.  We never have seen hospital data, except in bits and pieces.  We started doing some pilots and especially in the Medicare population and we started seeing hospital data.  It was a revelation to see how sick our patients get, or see how sick they are to start with and how often they end up in the hospital, and that’s particularly true obviously in older folks, but in patients with lung cancer, pancreatic cancer, and more recently, we’ve seen this more in myeloma.  We’ve never had those data before and we’ve never had how imaging labs, even skilled nursing facilities that we have fit into the total cost of care for this group of patients.  At some point, we should be able to piece those, piece that information together so that we can see better what we’re doing.  Right now, the big emphasis is still on hospitalization and we have, I think, made great strides in addressing that.  But the next big one is going to be end-of-life care, specifically hospice.  We have seen that in some situations we have just as many people admitted to hospice for less than four days as admitted for over four days and just a difference of a few days hospice admission would push more of those folks into having hospice be more useful for patients and their families.  And again, physicians have never had this kind of feedback.  Medicare has the best and the right data out there.  Even in commercial populations we’ve not seen, or we found it difficult to get good end-of-life care data, good dates of death, good sites of death, whether they died in the hospital or not, but Medicare is by far the best.  Clearly, the government does death and taxes, so they should be good at that.  They have to have a date of death.  The Medicare claims you will get whether they were on hospice or not and whether they were in the hospital or not.  

Is there anything else you would like to add about implementing the OCM in a large-scale practice?

One thing, one way to think about this whole program, the oncology care model, is that it’s a process improvement model and that means that you will make a lot of changes or we are making a lot of changes.  When you do that, it’s very difficult to tell what the value of each change is.  What we hope is that the sum total of all those changes will be improvement of care and reduction in cost.  But part of the risk for that is that we’re overburdened with metrics, or administrative detail.  Hopefully, this will be a program that will be successful initially, and then it will be looked at and refined so that it’s efficient for both the government and for us.

The bottom line is that we could not make improvements without getting these data, but we will likely be far from having the most efficient system by the end of this program and it will be important for us to look at everything critically.  Probably many things that we’re doing, they all vary in the same direction, we don’t need to report all of them and there are some that may not be helpful at all.  And it’s really important for us to weed out what is critical and what is unnecessary so that we don’t all die from too many clicks.

For articles by First Report Managed Care, click here

To view the First Report Managed Care print issue, click here