December 03, 2019
Charles Saunders, MD, chief executive office, Integra Connect, breaks down what the Oncology Care Model is, highlights the various components of it, and reviews recent performance data.
My name is Dr. Charles Saunders. I am the chief executive officer for Integra Connect. Integra Connect is a five-year-old company that is dedicated to enabling oncologists to manage cancer populations under new alternative payment models, in particular, the Oncology Care Model.
I have been with the company for about five years. Before that, I was CEO of Healthagen, which is a large division of Aetna that focuses on population health management, ACOs, patient‑centered medical homes, value‑based contracting, clinical analytics, care management, and so on.
The Oncology Care Model is an alternative payment model created by CMMI, the Center for Medicare & Medicaid Innovation. What it does is it provides aligned financial incentives to oncologists to manage the totality of an episode of cancer care, where they are rewarded for delivering value rather than volume.
The components of it are: the providers are paid, in addition to their typical fee for service, a Monthly Enhanced Oncology management fee or MEOS payment, which is $160 per member per month for the duration of an episode of treatment, which is defined as six months.
It can be continued with another episode of a patient who is still getting chemotherapy. The episode is triggered by the administration of chemotherapy, either by infusion or orally, once the prescription is filled.
Providers are accountable for the entire cost of care – not just their cost, but hospitalizations and every other service that occurs during that six‑month period of time. A number of quality measures are tracked by CMS, including hospital admissions, ER visits, and end-of-life and utilization of hospice care, as well as some quality measures that relate to processes of care.
Providers also receive a performance‑based bonus every six months. If the total cost of care falls below the CMS-established target for that period of time, there is a share of savings that are shared with the provider. That is, of course, adjusted based on the quality measures. That is the one‑sided model.
There is also a two‑sided risk model in which the provider is responsible for reimbursing CMS for some percentage of the cost if they are over the target price. Those are the two models.
In exchange for the payment, the providers must use an ONC‑certified EHR for care navigation. They must provide a care plan for each patient, do a certain amount of quality reporting, and use evidence‑based treatment guidelines for care.
Those are the basic components of it. There are a number of commercial carriers that have signed up – about a dozen or so signed up for that same program. But there are other commercial carriers who are coming out with their own novel alternative payment programs. Some are similar and some are different.
The performance period for data continues a trend with some metrics that are improving. At Integra, we support about 1,000 oncologists in more than a dozen practices across the United States.
We have seen a steady improvement from PP1 to PP4 in a number of the important metrics, so a steady decline in the percentage of patients with a hospital admission, a steady decline in ER utilization, and a steady improvement in all of the quality measures across the board.
The delta, or the difference between the target price that CMS established and the actual price was, in the aggregate, negative – in other words, a net negative or a loss for the practices in Performance Period 1, but then that steadily declined through Performance Period 2 and through Performance Period 3.
We saw an aggregate leveling off in Performance Period 2. In other words, the amount was small. The aggregate delta was relatively small on a per-episode basis between the two. But it did not improve much, if at all, in aggregate, between 3 and 4.
However, we did find that in our cohort, 60% of the practices individually actually did improve from Performance Period 3 to 4. 40% did not or got slightly worse.
Interestingly, the drivers of improvement in the first three performance periods were largely influenced by reductions in acute care visits. The practices that did the best or had the performance improvements in Period 4 were primarily driven by control of drug costs.
That is important because drug costs were 56% of the entire episode costs in Performance Period 1 and now they are 60% of the entire episode cost in Performance Period 4. Those practices that were able to utilize lower-cost alternatives actually saw an improvement from Performance Period 3 to 4. Many of those, if not most, received a performance‑based bonus.
Why this is important to payers is because first of all, CMS has really only reported the results from Performance Period 1 so far. We are now reporting the results from periods 1 through 4. With our cohort of approximately 1,000 oncologists and about 50,000 active episodes at any one time, we are showing that the program works.
We are showing a steady reduction in acute care days, a steady reduction in ER visits, a steady improvement in the quality measures, and a narrowing of the gap between the target price and the actual results, with about half the practices achieving savings.
From a payer's perspective, this is a mechanism to address the dramatically rising cost of cancer care in a way that not only reduces costs, but also improves quality at the same time.
That is not to say that there are not flaws with the OCM program that we would like to see fixed or addressed. You will also hear practices say that it is complicated and expensive for them to administer. Then there are some difficulties, like the data acquisition reporting that is required, but the concept and direction are valid.
There still remains enormous opportunity. For example, a large percentage of cost and opportunity is the consumption of resources in the last 30 days of life. That has not changed over the performance period. There are a lot of inroads that can be made there.
These results, I think, bear on the OCM program in a number of ways.
First of all, those practices that have not received a performance‑based payment by Performance Period 4 must either elect the two‑sided risk model or drop out. We currently have a number of practices that have not received a bonus, or performance‑based bonus payment (PBP), and they are currently evaluating whether or not they should stay in.
We assist that by projecting how practices are likely to perform in 2020 and what the probability is that they would have to make a payment back to CMS and then model the application of stop‑loss insurance to mitigate some of those risks. That is something that they are going through right now. But there are a percentage of the practices that will drop out of the program.
There are some tangible factors of the program that could make it a strong program, such as making the attribution logics more transparent to providers for lightening the load for the data aggregation acquisition reporting requirements and addressing the target setting, which is, arguably, artificially low in some cases, because the stage of the disease and comorbidities are not generally fully accounted for in the risk adjustments. The novel therapy adjustment and other factors might not be fully appreciated or accurate.
Some of those things can be addressed, and it is my expectation and hope that CMS will continue to evolve the program going forward, because the basics of the program do work.
I think we will wait to see what CMS comes out with and what the feedback has been with their own performance results. There have been a number of alternatives, from entities like COA and Come Home and others that have suggested modifications to the program. But I suspect that the program will continue to evolve, hopefully fixing some of the issues with it.
We have actually started to measure the impact of specific interventions, because practices may think, well, what can I do to impact the price and the quality? What interventions or practice transformation initiatives actually matter and get results?
We find that there are several things that have an impact. One of those is complex case management for high‑risk patients. Combining predictive analytics to identify patients prospectively with a high risk of either a hospitalization or death allows you to case manage those patients closely to prevent adverse events and hospitalizations and prolong their life. We find that that works.
We find also that end-of-life programs work and have measurable results. We find, for example with case management of high‑risk patients, that there is about a $2,200 episode savings with a program for those kinds of individuals, or about a little over $400,000 per year for a practice for every 100 episodes. That works.
We also saw that end-of-life programs consisting of palliative care, hospice, bereavement and other factors also produces an impact through the end-of-life period and overall.
Overall, there is about a 2% impact that can be made on case management, a 2% impact overall that can be made with palliative and hospice care and about a 2% impact that can be made with efforts to modify drug spend by choosing biosimilars and cheaper generics.
I think those are important take-home messages. The program works. It is a little difficult and complicated and expensive to administer, however.