In May of this year, The American College of Radiology (ACR) Council elected Alan D. Kaye, MD, FACR, as its new president. Dr. Kaye is a managing member of Advanced Radiology Consultants, LLC of Connecticut, former chairman of radiology at Bridgeport Hospital for 22 years, and on the medical staffs of Bridgeport, Yale New Haven and St. Vincent’s Hospitals.
He is currently a member of the ACR Budget and Finance Committee, Radiology Integrated Care Network and numerous ACR commissions, including Government Relations, Economics, and Informatics. As ACR president, he will be a member of the College’s Board of Chancellors and Executive and Budget and Finance committees. Kaye previously served as a speaker of the ACR Council and member of the Council Steering Committee and many other ACR commissions and committees.
Dr. Kaye received his medical degree from the University of Connecticut, and has interests in administrative radiology and health policy. He currently serves as president of the Radiological Society of Connecticut.
Advanced Radiology interviewed Dr. Kaye shortly after he was installed as ACR president, to discuss his goals for the coming year and thoughts on the near future of radiology.
AR: Dr. Kaye, you’ve been involved with ACR for many years, in several different roles, how did that culminate in your becoming president?
Dr. Kaye: As the long-time leader of Advanced Radiology, with our hybrid structure – private community offices, hospital departments, and training of young radiologists – I have had an opportunity to interact with and impact all aspects of the ACR’s functions. I became more involved with the ACR in the late 1990s. I first ran for election to the Steering Committee in 2000, and was a member of that committee for six years. I followed that with two years as vice-speaker, two as speaker, and finally six years as a member of the Board of Chancellors before being elected president. I really enjoy being involved with ACR and believe I can make a significant contribution to the organization and help it serve radiologists and our patients. The various roles that I have had, both in the ACR and in private practice, allow me to bring the independent physician’s viewpoint to the ACR. I feel that my relationship with ACR has been mutually beneficial. I enjoy serving and making a difference, which is why I pursued the position.
AR: What do you see as the biggest issues you will face during the next year?
Dr. Kaye: With all the interest in health care reform, there are a lot of ideas as to how doctors, including radiologists, can best serve patients. I don’t have a lot of confidence in the federal government’s ability to assign meaningful, appropriate metrics to judge physicians’ performance and, ultimately, how doctors should get paid. There is a lot of talk about outcomes, but then we see things being measured that don’t necessarily affect outcomes. I think there is something of a “training wheels” method of using metrics, i.e., getting people used to stating what metrics they are using I may have chosen different training wheels instead of things like, for example, the criteria used to determine a stenosis on a carotid ultrasound, as an indicator of what you should be paid. Not that it’s a bad thing to do, but it doesn’t really relate directly to outcomes, and can be trivial.
As far as I am aware, no one has come to the ACR and asked, “What should we use as metrics for radiology If they had, we could have done better.
I believe that no matter what happens, in the foreseeable future fee for service is going to be the primary basis of payment. Whether the method includes a provision that twenty percent of your payment is based on those other metrics, you’ll still be paid for the piecework that you do. There are other alternative payments models being developed in which there will be things that are put at risk. How radiology will be involved in that, where we don’t have as much involvement in direct patient care, is going to be problematic. When there is a bundled payment that must be split with other doctors, there will be infighting among the doctors. On the practice side of things, I think that’s a big issue.
AR: Is this one of the reasons that Clinical Decision Support will soon be required? It was surprising to me that the responsibility for monitoring referring physicians was placed on radiologists.
Dr. Kaye: Yes, that is one of the reasons. Unfortunately, they could have done better with that, too. First of all, who doesn’t get paid if the referring physicians don’t use Clinical Decision Support? We (the radiologists) don’t get paid. We are put at risk for somebody else’s behavior. And this has been a recurring issue over the years. An example is capitation. Early in my years leading the practice, we dabbled in it to see how it would work. With one of our payers, we agreed to take a capitated rate, which, as calculated, wouldn’t have been much different from our regular rate, provided volumes stayed the same. If volumes went down, it would be a higher per unit reimbursement, because the overall payment was fixed. What we didn’t take into account was that the only other group of physicians that was capitated were the primary care doctors. Now, if you’re a primary care doctor and you get paid $4 or $5 per month per patient, and you never have to see the patient, you make out well. You go home and have dinner with your family every night. If a patient calls and complains of chest pain, you tell them to go get a CT scan. The doctor doesn’t get “dinged” for doing the scan, because they’re not costing the system any more money – because the radiologist does not get paid any extra for doing an additional test. But the primary care doctor would get “dinged” for referring to a specialist, who would get paid and, therefore, cost the system money. We saw primary care doctor’s referrals increase significantly. The primary care doctors would rather see the imaging study first before examining the patient, so they can their patient not to worry, or request another test. So we were at risk, and primary care doctors had the exact opposite incentive. That didn’t work out, and we withdrew from it.
So when you start implementing bundled payments, or other payment models, you have to take many things into consideration and be prepared for unintended consequences. The problem with the new payment models is that the metrics are unproven, and tend not to be, at least in my experience, valid metrics. The metrics also change frequently. The point is that new payment models are uncharted territory, and don’t have a very good track record, from my perspective.
AR: Do you see the ACR being proactive and approaching the government organizations who assign those metrics?
Dr. Kaye: Absolutely! The ACR is one of the premier physician advocacy organizations. Whether the issue relates to economics, health policy quality, safety, and informatics. The physician volunteers and staff are dedicated and brilliant. An important reason I have stayed involved is that I learn so much from them – and it has benefitted ARC.
AR: What are other issues you see the ACR as needing to face in the coming year? The second big issue, maybe a longer term one, is machine learning and artificial intelligence. This is related to Clinical Decision Support in that radiologists are now supposedly going to be more valuable to the system because we are now being gatekeepers, we’re going to help save money, lower radiation dose, get patients taken care of earlier and more appropriately. The problem is that, with CDS, machines (actually, software) will be doing much of that, and everyone will consider what the machine says and not consult the radiologist. That worried me about decision support, and that was before the whole furor over artificial intelligence. When we add artificial intelligence and neural networks to the equation, what will happen? A machine might know better than a radiologist what is appropriate. I often research things that I know I need to know one hundred percent instead of ninety percent. Suddenly there is a machine that knows it all before I even know to ask the question.
The ACR brings unique capabilities to the future of AI, and I do not think that includes the ACR doing the research and development. Although we have very smart people, we don’t have the resources that corporations like Google and IBM and other organizations do. The ACR should focus on the regulatory, economic, quality and safety issues. So now we have the Data Science Institute which is going to do that, and I’m on the advisory board for that. My goal is always to make patient care better, and I can’t imagine that a machine, any time in the near future, will be able to bring to the equation the combination of science, empathy and engagement that a trained radiologist can.
One of the things I intend to do as ACR president is make sure that the members and the Council have an opportunity to weigh in on this and be aware of what’s going on. We need to determine if the Council needs to be more active in this than it has been in the past.
AR: Do you see the role of ACR in AI research as consultative with the IBMs and Googles of the world?
Dr. Kaye: We can help guide them to do more appropriate research and focus on better patient care and helping radiologists be more effective. I hope to find ways that we can work together with the large corporate research groups.
AR: Your term as president is one year. Is that enough time to accomplish what you’d like to?
Dr. Kaye: True, that is not a lot of time, but there are opportunities to begin processes that can be lasting.
AR: And it’s not like you’re going away after a year…
Dr. Kaye: Well, most presidents do go away. I know that I will not be as involved as I’ve been the last several years. By then it will be another person’s chance. Just like in our practice. This year will be the big push, and I’m going to try to push in the right direction.
AR: When you look back at the end of your year as president, what will you hope to have accomplished?
Dr. Kaye: As you can tell from my comments so far, there are two essential current efforts of the ACR, the Data Science Institute and the Health Policy Institute, where we can focus our resources on making sure radiology adds value and how we can provide data for the alternative payment models.
AR: It seems as if you’ve arrived at this position at a very interesting moment.
Dr. Kaye: There is never a time that’s not interesting. But this is the biggest turning point, by far. I’ve always thought about the future more so than about the present or past. That’s why informatics is so important in our practice, and why I have advocated for it in the ACR for at least eight years now, and why AI is such an important topic for me. For the future. we need to assess radiology’s value and evaluate where we can make the most difference.
AR: That raises one more question. I’ve read recently that the number of medical students choosing the radiology specialty is declining. Why do think that is, and what can be done to redirect that trend?
Dr. Kaye: Understandably so, because of three things. First and second are alternative payment models and artificial intelligence, which we’ve talked about. Those have caused uncertainty for radiology groups. That uncertainty acts as the catalyst for a cycle of events. It causes practices to seize up, and not hire. When the job market goes down, people turn to other specialties. That goes on for a year or two until the practice gets too busy, or doctors retire. Then they need people and the trend reverses. I think we’re at that point of catching up now. I have a feeling that we’ll be able to test, in the next couple of years, whether the downturn in applicants is due to the job market versus the continued concern over payment models and artificial intelligence. My prediction is that this cycle may be different than previous cycles. I think that AI and alternative payment models are new factors that are much more threatening. While these issues could make this downturn more severe, we have felt similarly in previous cycles.
The ultimate determining factor in the success of a radiology practice is delivering quality care and great service to referring physicians and patients. Cost is also a contributing factor, but before that becomes an issue, taking care of those two groups is the most important thing. Service has to be incorporated into a practice’s culture. I repeatedly remind our people that, although we examine 1,000 patients a day, that test may be, for each patient, the most important thing he/she does that day, that week, or maybe even in his/her lifetime. Every member of our practice, from the radiologist to the people who schedule appointments, to the ones who keep the office clean, is important to our operations and needs to understand that. When I read an exam, I think, if that were my child, or my family member, how would I want them to be treated? There was a study published some years ago that showed if you included a picture of the patient with their images, it had a significant positive affect on how the patient was treated. Empathy, I think, is what differentiates us.