A blueprint for high-volume, high-quality lung cancer screening that is detecting cancer earlier—and helping to save lives
It has been a little more than a year since UnitedHealthcare launched a pilot program to test a bundled payment model for oncologists, and the insurer and its five pilot sites have worked through many challenges to make it operational. The program pays a flat fee that factors in what the physicians would have made by charging the health plan a margin on the chemotherapy drugs provided to patients. It also requires the practices to standardize the regimens of care they provide within their groups. Next month, the pilot sites will begin comparing their results with one another to look for opportunities to improve care. That will be the most important part of the effort, says Lee N. Newcomer, MD, MHA, senior vice president for oncology, women’s health, and genetics at UnitedHealthcare. “When you look at this kind of measurement, you can always find things that can be improved and costs that can be lowered. It’s inevitable.”
A board-certified oncologist who practiced for nine years, Newcomer is a former chief medical officer and vice president for health service operations at UnitedHealthcare. In between, he helped form Vivius, a consumer-directed health care company, and previously was a medical director at Cigna Health Care of Kansas City. He is the former chairman and a current board member of Park Nicollet Health Services. Newcomer earned a bachelor’s degree in biology from Nebraska Wesleyan University, a medical degree from the University of Nebraska, and a master’s degree in health administration from the University of Wisconsin at Madison. He completed an internship and a residency in internal medicine at the University of Nebraska, a fellowship in oncology at Yale University, and a fellowship in administrative medicine at the University of Wisconsin at Madison. He spoke recently with MANAGED CARE Editor John Marcille.
Managed Care: Can you tell us about your new payment program for oncologists?
Lee N. Newcomer, MD, MHA: We are working with five medical groups around the country, each of which picked its own treatment regimens for 20 different clinical conditions for patients in breast, colon, and lung cancer. There is universal agreement on the broad strategies for treating each of these conditions — chemotherapy, radiation, or surgery — but there is variation and disagreement about what type of chemotherapy, radiation, or surgery. Because there aren’t absolute answers, we didn’t try to force a solution on them. We asked them to decide as a group which regimen for each of those categories was most effective and then to treat all of the patients in their practices with that regimen.
MC: Do treatment regimens often vary within group practices?
Newcomer: In most practices today, one partner may use one regimen and another may use another, depending on where they trained. We got these groups to commit to treating their patients uniformly in each of the categories. There are a lot of strongly held opinions about these therapies, and several of the groups had to lock the doors and just have it out to come to a consensus about what treatments they would use. The groups, obviously, haven’t chosen the same treatments for the same kinds of patients.
MC: Is the goal to create clinical care pathways?
Newcomer: A lot of people have looked at the program as a way of saying we got everyone to treat on the same regimen. There are multiple companies out there doing pathways, but that is not going to be enough to solve the problem. Pathways are just a beginning. It’s nice to get consistent therapies within a group, but what is critical is then measuring the performance of those therapies and seeing how you can make them even better. Without measuring performance and without constantly going back and looking at the data and saying, “What could I improve?” pathways are only going to get us a tiny blip in savings for one year and it’s going to disappear.
MC: Does this program simply substitute the flat fee for what the doctors make on marking up the drugs, or am I missing the point?
Newcomer: Because we coalesced a lot of payments into a single payment, this has mainly been known as a bundled payment project, and there is that component to it. But there were two objectives of the program. The first was to create a payment structure that separates the selection of drugs from the income that an oncologist would make. The second, which folks have lost sight of, is that we are going to create a learning environment where we can measure our performance and improve it.
MC: Is improving outcomes even more important than cost control?
Newcomer: It is, because in the end, looking at the data, understanding the performance, and then changing the care we provide is where you will see both better outcomes and cost control.
MC: What have you accomplished so far?
Newcomer: Most of the work in this first year has been on the operational aspects, such as working with practices to find patients, making sure they notify us right away, and making sure that we’re not missing folks who they started treating but forgot were UnitedHealthcare patients, and so on. That’s been the hardest part of this in the initial 12 months. We have also been developing measures of performance, and they are all finished. The data are being run now. The groups will meet in January to look at that data and discuss where they might see improvement. It will give us a chance to compare the results of one approach versus another in groups that are uniformly treating patients. It is essentially real-time comparative effectiveness research.
MC: What did you learn while pulling this project together?
Newcomer: I learned that I couldn’t come to a group and say, This is a great idea, you should do it. Instead, their leadership had to agree, and then they were the ones who pushed it. There had to be a physician champion in the practice who said this is the right thing to do. The second thing was that it takes a long time. And the third is, I don’t think there is any right answer yet in oncology. Many executives have the impression that we ought to be able to just decide that it’s regimen A and that’s what we ought to cover. For almost every cancer, there is a regimen A, a regimen B, and a regimen C, and they are all effective. We just don’t know which one is superior.
MC: Getting the opportunity to see data from other group practices must have been appealing to physicians.
Newcomer: It was, because every one of these folks is highly competitive. They are very interested in knowing how their performance compares to others’ and how they can get better. They don’t have access to that kind of information through any other source. One of the very exciting things about this for the medical groups is that we are going to be able to sit down in an environment where there’s no penalty and talk about getting performance better.
MC: By no penalty, you mean that there is no longer a financial incentive to use a particular therapy?
Newcomer: That’s correct. Let’s say we find that a generic combination of drugs, very inexpensive, is just as good or maybe even a little better than any combination that uses very high-priced brands. If we decide to change to the generics, the episode fee stays the same, so the money they used to make as a profit margin on the drugs is still there as patient-care money. Their patients will be getting a better result, but their income is not going to go down.
MC: Are your physician costs lower than your drug costs?
MC: What’s the ratio?
Newcomer: We spend about 40 percent of our budget on chemotherapy, about 40 percent on hospital care, and the remaining 20 percent on everything else. What we spend for the medical oncologist is pretty slim compared to what we have to spend for drugs.
MC: So when I look at income charts for oncologists, what looks like a lot of money is really irrelevant?
Newcomer: Part of that money comes from the margin on drugs.
MC: But you are preserving that amount.
Newcomer: We are going to preserve that. We have to face the realities of the marketplace, where there are fewer and fewer oncologists to serve more and more patients with cancer. If I walk in and say to the medical oncology community, I have a great deal for you here: I am going to take your income down 20 percent, I am not going to get very many volunteers. If I can say to them, I am going to keep your income whole, but I would like you to work with me to reduce the costs of care and keep the quality better or the same, we are both going to win. That’s what this program is intended to do.
MC: Where do the savings come from for the insurer?
Newcomer: There aren’t any savings for the first year. All we’ve done is drawn a line that says, “Here’s your patient-care fee. This is a reasonable amount of money, and it includes the money that you used to make on drugs. We’re not going to increase it every time a new expensive drug comes out, and we’re not going to decrease it if one of these drugs becomes generic.” We will save some money because physicians will not have an incentive to use more expensive drugs. Far more powerful is that in comparing performance, we may find regimens that are not only as good as all the others, but far less expensive, and we’ll switch to those. There may be times when we find that the most expensive regimen is producing better outcomes, in which case we are going to move to that. As we get more and more knowledge about best practices and how to reduce costs, that is when this will pay off.
“The registry is our silent partner. We are developing a much better picture of cancer care in the United States than we could do with claims alone.”
MC: You have also established a registry to look at some of the variation in care across the country. Are the physicians in the pilot payment program using the registry?
Newcomer: They are. The registry is our sixth silent partner when we look at the data. We are going to compare our performance in these pilots to national performance as well. If we find somewhere where the national norms are still better than what we are getting in our five pilots, that raises another set of discussions.
MC: Can you tell us more about how the registry works and what kinds of information you have derived from it?
Newcomer: The registry now has 3,200 breast, colon, and lung cancer patients. It combines information supplied by physicians, such as stage of cancer, genetic information, and the current status of patients, with data from claims. It creates a small, longitudinal electronic medical record so that we know where the patients started, what drugs they were treated with, what radiation they were given or not given, what kinds of surgeries they had, and what kinds of complications they had. We can begin to put patients in buckets that are similar, and that helps us understand what is appropriate treatment for those groups. We are developing a much better picture of cancer care in the United States than we could do with claims alone.
MC: Can we look at this as a variation on a disease management program?
Newcomer: We are trying to identify gaps in care nationally that we can correct nationally. We are not going back to individual patients, because claims are always several months behind the fact. But if we see that across the country, for example, Avastin is consistently being used incorrectly, we can put in programs to keep that from happening.
“We found that 15 percent of patients being treated with Herceptin did not have the [HER2 overexpression] test, or they had the test and the gene was not overexpressed. We are now at 100 percent testing, and the drug is being administered correctly.”
MC: What programs have come from the registry?
Newcomer: Several years ago we discovered there was a wide variety in how Herceptin was being used for breast cancer, so we did a formal chart review. The first thing we looked for was, Did these patients have the HER2 gene? If the HER2 gene is overexpressed, Herceptin is a very effective drug. If that gene is not overexpressed, the drug will not work and could be very harmful for no gain whatsoever. We found that 15 percent of patients being treated with Herceptin did not have the test, or they had the test and the gene was not overexpressed. We now require the HER2 gene report to be submitted before we pay for Herceptin. The good news is that we are now at 100 percent testing, and the drug is being administered correctly.
MC: What role should insurers play overall in making sure genetic tests are given and the results are followed?
Newcomer: When there is a biomarker that is clearly associated with drug effectiveness, the test should be a standard authorization criterion for the drug to get coverage.
MC: Is that becoming standard in managed care?
Newcomer: It will be. The criteria are often very crisp and clear criteria, and there really isn’t a reason that you should give some drugs without those markers.
MC: Do you expect that a majority of cancer drugs will soon have these biomarker tests?
Newcomer: I don’t know. We are at the beginning of the genetic revolution. One of the things we are going to have to be careful about is not jumping to a conclusion about genetic testing too quickly. We have to be rigorous about making sure the science fits before we start assuming that a genetic test is associated with a drug or a cancer. We can learn from Duke University’s experience. They moved too fast. We all want this to work, but we have to do the same rigorous testing that we do on drugs.
MC: Where do you look for guidance on these kinds of decisions?
Newcomer: The FDA has been fairly clear about this on new drugs, so you will see it on FDA labels. We also use the National Comprehensive Cancer Network as an information source. If the NCCN says you should do a genetic test before using a certain drug, that would become part of our criteria.
MC: You work closely with the NCCN. What more could it do to help physicians and health plans?
Newcomer: If I look at regimens for breast cancer, I will find 14 or 15 recommended. It would be great if the NCCN could begin to say, Here are the two or three most preferred because they have the best results, the least toxicity, or whatever the rationale might be. That’s the next step, rather than listing every regimen that has some evidence of efficacy.
MC: Would you recommend that other health insurers operate registries of their own?
Newcomer: Our registry contains only UnitedHealthcare patients because they are the only patients I have claims data on. The American Society of Clinical Oncology is working on building a common registry across all payers, and it starts with physicians submitting data. In the long run, that will be the better approach. But it’s going to take longer, and it’s going to be more expensive. We started ours to get something happening, to make something move.
MC: Is this a piece of the long-awaited medical record?
Newcomer: It could be. The long-awaited medical record, though, is only going to tell you what happened within that physician’s office. The advantage of using claims data is that you find out what happened to the patient in the rest of the community.
MC: We often hear that the benefit of an electronic medical record is that it can exist across providers.
Newcomer: In Minneapolis, all four of the major health provider systems all use the same common medical record from Epic Systems Corp. But there is no common database, so a researcher can’t come into Minneapolis and say, I’d like to look at all the 45-year-old diabetics across all systems. We haven’t created the infrastructure yet to allow those kinds of registries to occur, even in a community where everybody uses the same electronic record. We have a way to go.
MC: One of the things you’ve talked about in the past is the high cost of oncology drugs and the need to evaluate them, perhaps based on life years gained. How do you see this playing out politically over the next several years?
Newcomer: The first step is going to be what we talked about earlier — better precision about who gets the drugs. In the old world, we would get a new drug for lung cancer, we would treat every lung cancer patient with that drug, and 20 percent of them would respond, so we treated 80 percent unnecessarily. With genetic biomarkers, we can narrow that down. That’s going to be the first approach to trying to reduce the cost structure. It won’t be enough. The second thing we are going to have to deal with is whether we get enough value from a drug.
MC: Do you have an example of a drug that should be evaluated in this way?
Newcomer: Provenge is a good example. Men with prostate cancer live three to four months longer with that drug on average, but it costs $100,000. Society is somehow going to have to decide about whether or not that is affordable. That will not be done at any private insurer, and the government will really struggle with that process. The British set a monetary limit for the amount they would spend for an extra year of life for cancer drugs, and we may have to do a similar thing if we are going to keep our system viable. My fear is that a bankruptcy of the system will finally force the discussion. I would love to see it occur way before that.
MC: Will the Affordable Care Act be effective?
Newcomer: It doesn’t do anything to truly bend the cost curve. The comparative effectiveness part of that bill is very, very minor, and that’s the kind of work we will need to do to truly bend the cost curve. I had a senator say to me, We took care of costs. We are going to be paying hospitals less. That is true, but that hasn’t helped hospitals understand how to cut costs. Hospitals will need direction, as will every provider. That’s where the act is silent, and we’re going to have to come back to that discussion.
MC: The last time you did a Q&A with MANAGED CARE, in 2000, you were at Vivius, a start-up consumer-directed health care company. What did you learn then that you apply to your role today?
Newcomer: Our cancer unit at UnitedHealthcare is a very small team of 10 people. That is deliberate. We are very flexible, very fast. We don’t think about claims or contracting or benefits. We find the cancer problems, then we go out to the appropriate departments within the company and we work with them to solve the problem. We had the same experience at Vivius. We got to think about what we could do that was different, and how we could make it happen. I also learned that you have to be willing to fail.
MC: Where have you failed recently?
Newcomer: We had a spectacular failure in women’s health. We discovered that many babies are delivered electively before 39 weeks of gestation, which causes problems. The babies are not ready to be born, and they have more neonatal ICU admissions. We sent a very competent obstetrician/gynecologist out to 18 hospitals where our data showed there were big problems with these premature deliveries. She gave the data to the hospitals and talked with the physicians. They all continue to deliver early. We learned that education isn’t enough. We are now looking at financial incentives for the hospitals, either positive or negative, to help encourage the development of policies that will improve this.
MC: Thank you.
A blueprint for high-volume, high-quality lung cancer screening that is detecting cancer earlier—and helping to save lives
Multiple Sclerosis: New Perspectives on the Patient Journey–2019 Update
Summary of an Actuarial Analysis and Report