Provider Perspective: Stephen M. Schleicher, MD, MBA, Addresses Accountability Versus Control in Oncology Care First

The interview is edited slightly for clarity.

On October 31, 2019, authors from Tennessee Oncology, OneOncology, and Tuple Health published an article in JAMA Oncology that offered a stunning conclusion: Even if a practice in the Oncology Care Model (OCM) complied precisely with current guidelines from the National Comprehensive Cancer Network (NCCN) for metastatic non—small cell lung cancer and did everything right, drug costs would make it impossible to meet the financial targets spelled out in the model.¹ The authors offered a solution: Instead of being judged against unattainable benchmarks, OCM practices should be measured based on their adherence to

clinical pathways or ability to document why a pathway didn’t make sense for a particular patient.

Almost on cue, the next day, the Center for Medicare & Medicaid Innovation (CMMI) released its proposal for Oncology Care First (OCF), the would-be successor to the OCM.² One of the paper’s authors, Tennessee Oncology’s Stephen M. Schleicher, MD, MBA, was pleased to see that OCF reflected some of the issues he had his colleagues had discussed. Schleicher, who also serves as medical director for value-based care for OneOncology, a network that includes 242 providers, discussed the findings and CMMI’s proposal with Evidence-Based Oncology™ (EBO):

EBO: Please describe your practice’s overall experience with the OCM. What changes have had the biggest impact on clinical outcomes?

SCHLEICHER: I’m a medical oncologist at Tennessee Oncology; we are part of OneOncology, where I am chair of the Quality and Value Committee. I’ve been very passionate about value-based payment for the past 5 to 6 years since the Affordable Care Act came out and, specifically, as an oncologist, when the OCM came to fruition. Our experience with the OCM, most importantly, has been a learning experience. I think that’s one of the biggest benefits of the model and why it has been a voluntary pilot. We’re getting lots of information about how we’re doing in keeping patients out of the emergency room when we can and how we’re using hospice.

We’ve had the opportunity to get rewarded through the care-management fees to provide nurse navigators, to help be the bridge between when the patient is in clinic and when the patient is at home. We’ve learned how to continue that communication and help direct patients to the clinic for fluids instead of to the emergency [department], to help assess a patient’s pain and act quickly before the pain gets out of control—similarly with nausea and constipation. So, we’ve been able to, through the resources provided from Medicare as part of the OCM, hire some additional staff to help with that. Staff is one thing; secondly, with the analytic capabilities we’ve been encouraged to develop through our partners with Flatiron Health and now with OneOncology, as I mentioned, an understanding of our utilization patterns and how they compare with our peers’, both within the Oncology Care Model and without. Those are some huge benefits of the model, and one of the most exciting aspects of the model involves all we have learned to improve the care we’re delivering to our patients.

EBO: In your recent paper in JAMA Oncology,1 one of the central points you and your co-authors make is that practices should be held accountable for adhering to clinical pathways or documenting good reasons the pathway does not make sense. How did you arrive at that conclusion?

SCHLEICHER: So, it’s a great question. The impetus for putting this paper together was the whole issue of accountability versus control. As oncologists, we want to be accountable for things we do incorrectly, of course—even if that’s not on purpose. But delivering a more expensive chemotherapy to a patient when there’s a better option, using an emergency room instead of a clinic for things that could be handled outpatient—those are things we can be accountable for, that we can control, and we should be accountable for. But there also are aspects of cancer care that we have no control over, and the main thing there—the elephant in the room—is the price of drugs.

New drugs have tremendous benefit for patients. We want to be able to give new drugs to our patients. Immunotherapy is really the big thing that has boomed since the start of the OCM. But those drugs that have good benefits also have high price tags, and we want to make sure we can give those drugs to patients without being penalized, because in the current OCM, the total cost of drug use is something we are accountable for and could be penalized for. What we wanted to do with the paper and this research is understand how well the OCM is incorporating the new prices of drugs into their expected costs of treating cancer patients. We looked at lung cancer.…The field has [advanced] even since the time of the data that we looked at for this paper. But since the start of the OCM, immunotherapy made its way into the second-line treatment of metastatic non—small cell lung cancer with both Keytruda [pembrolizumab] and Opdivo [nivolumab], which both have survival benefits. We thought that was a good opportunity to look and see if, when we used these drugs correctly—which we can control—we are still over target on what we should be spending, and if it’s due to the price alone, we cannot control that. And that’s a problem with a value-based care model—trying to improve the quality of care and the cost of care if we can’t control that.

So, that’s why we looked at lung cancer. We looked at all our cases where Medicare told us we had spent more than expected, and within that, where we had prevented everything that was the goal of the OCM. [We looked at cases where] we prevented emergency [department] visits, we prevented hospitalizations, postacute care, inappropriate chemotherapy at end of life—all these aspects that we can control and one should be able to prevent. We saw that in many of our cases—over half—we had done all that, but we were still over target because we had used either Keytruda or Opdivo in the second-line setting for metastatic non—small cell lung cancer. That was NCCN compliant at the time and is still standard of care, although we now often use immunotherapy first line. We found in that group of patients that we were over target for using the right drug at the right time for patients that deserved to have that drug, yet the model, despite Medicare’s best attempts, couldn’t account for that, and that is what we are trying to communicate with the paper. Our conclusion was that we were pathway or guideline concordant, which is what we can control and that’s what we should be doing; the cost was higher because of the price, which we can’t control. So maybe the pathway concordance—following pathways we can all agree on—is a better measure of us using the right drug than the cost of care itself, from a drug perspective.

EBO: And the data you present were consistent across the practices that you surveyed?

SCHLEICHER: Exactly. We have 30 clinics in Tennessee Oncology right now—OneOncology has over 100—but we were looking at Tennessee Oncology, and this was not clinic specific; this was across all our providers at the different clinical sites.

EBO: The day after your paper appeared, CMMI released a request for information on a model that is being called Oncology Care First, or OCF, which will be the successor to OCM when it expires in 2021. What are your observations so far about OCF?

SCHLEICHER: OCF, I think, is an example of Medicare being very thoughtful—of recognizing that we need to change the way care is delivered but also recognizing that cancer is a very difficult disease because it is very heterogeneous—every patient with cancer is different, every cancer is different. We have high costs of care both from utilization and drugs. So, I think they tried to come up with a thoughtful model to do this.

There were 2 components of the OCF model that I thought are definite improvements from OCM that relate to our paper in terms of the accountability versus control. The first is how they calculate the expected cost of treating cancer care and how drug costs are incorporated into that; there are 2 mechanisms for that through the trend factor that accounts for overall implementation of healthcare costs, and the novel therapy adjustment, which is how a practice is using new drugs that have been FDA approved compared with their peers.

In the OCM, they applied this across all cancer types and all patients in conglomerate; OCF, they will look at the specific cancer itself. When a new drug or a new indication comes to that disease, they will try to account for that change at the [level of the] disease itself—which, hopefully, improves the accuracy of predicting the cost of care as healthcare standards of care change in oncology. That’s a benefit and shows that Medicare was listening to us as we all provided feedback during the OCM, and it is clearly very relevant to the paper we just published in JAMA Oncology.

A second improvement is how providers should be accountable for some of the low-risk cancers, such as low-risk breast, low-risk bladder, and low-risk prostate. Let’s take low-risk breast cancer—that could be a patient who had surgery, chemotherapy, and radiation therapy 4 to 5 years ago but still requires antiestrogen therapy—a hormonal pill. In that case, the provider might be seeing the woman every 6 months to prescribe that pill; standard of care might be a mammogram once a year for a patient who still has her breasts intact, with an exam. In the current setting, if we are treating a patient like that but they go to the emergency room because they have a stroke or because they need a hip replacement or they have a heart attack—something completely unrelated to their disease—we are still accountable for that care under the current OCM. They have corrected this in the current OCF by removing these patients from the accountability portion of the total cost of care, such that we are more accountable for patients for whom we are providing intense chemotherapy, seeing them regularly and really are responsible for their care, versus patients we are seeing every 6 months who have other unrelated health problems that could throw off our ability to succeed in the value-based care model.

EBO: That would seem to make sense. Do you have any other observations as you move forward with alternative payment models?

SCHLEICHER: It’s important for groups to be in these models, even if they are voluntary, just to learn about how they are doing compared with their peers and find opportunities for improvement. And as long as groups can publish on these matters and make it known to their other colleagues across the country interested in improving the value of care of cancer—if we can communicate that through publications such as ours [in JAMA Oncology], through conferences, through Medicare, to help these models get better over time, that’s really how we succeed: all of us coming together versus working independently. REFERENCES

1. Lyss AJ, Supalla SN, Schleicher SM. The Oncology Care Model-why it

works and why it could work better: accounting for novel therapies in

value-based payment [published online October 31, 2019]. JAMA Oncol.

doi: 10.1001/jamaoncol.2019.4385.

2. Oncology Care Model. CMS website. innovation.cms.gov/initiatives/

oncology-care/. Updated November 13, 2019. Accessed November 13, 2019.

Download PDF: Provider Perspective: Stephen M. Schleicher, MD, MBA, Addresses Accountability Versus Control in Oncology Care First