Fifteen oncology practices-called "vanguard" practices-will begin providing patient records to CancerLinQ, the cancer care learning system being developed by the American Society of Clinical Oncology, later this year.
ASCO says that in the long term, the society envisions that CancerLinQ will provide information for clinical guidelines and decision support for oncologists who want to use their patients' individual characteristics to determine the best treatment option.
As ASCO plans the system launch, ASCO CEO Allen S. Lichter, MD, FASCO, spoke with OT about what CancerLinQ means for oncology practices in the foreseeable future.
Where is CancerLinQ on the development timeline?
"We are deeply in the design phase right now so we can provide what practices need in order to improve the quality of their care and to improve their practice. We are having a series of workshops, and we have had one meeting of our Physician Advisory Committee, which includes a physician from each of the vanguard practices. We will be continuing those sessions until we get to the ASCO Annual Meeting, where we plan to show a demo of the product.
"In the fall, we will begin taking in data from practices and feeding back quality and practice analytics reports to them-and getting feedback and making revisions. We will first release this product to the vanguard practices and make sure that people feel that CancerLinQ works for them and that it adds value.
"My guess is we will need some months to make sure that we have the product functioning well. As we get into 2016, our hope is to begin to allow the many, many other practices that want to join this program to begin to participate."
How will practices submit data to CancerLinQ?
"Our ideal goal is to create an interface that allows the data from a practice's electronic medical record [EMR] to be extracted seamlessly. One might think of a feed going into the CancerLinQ program on an every-night or every-other-night basis that happens in the background.
"One of the challenges in a project like this is that there are so many different EMRs, and even in the same EMR, there are often many versions of it. So it is quite a challenge to begin to extract the data in a meaningful way. We are meeting with EMR vendors to try to enlist their aid in this process. We plan to be able to take data from any EMR, but we are starting with the EMRs used by the 15 vanguard practices. Then when the product is ready to scale, we will move on to tackle other EMRs."
What kind of reports will oncology practices receive from CancerLinQ?
"The first things are reports on their performance compared against ASCO performance measures. Many of those measures will be drawn from the Quality Oncology Practice Initiative [QOPI(R)] project, where we have validated a large number of performance measures. Some of them will be new because we'll have access to more data and be able to feed back quality reports on things that we could not get out of QOPI.
"As of right now, data comes into QOPI through manual chart extraction. CancerLinQ will extract the data from the EMR, so that's a huge difference in the amount of practice effort required to receive this valuable data.
"CancerLinQ dramatically differs from QOPI in the sense that it will also include outcome data. QOPI always has had performance-measure data, but we never knew what happened to the patient. So we could say that Mr. Jones got "the right therapy at the right time," but we never knew what happened to Mr. Jones, nor others like him, to know whether those performance measures were meaningful for patients like Mr. Jones.
"And, of course, cancer patients differ dramatically. When we create performance measures, it's an average across huge groups of aggregated patients. But when you have lots of patients and can start separating patients into smaller and more meaningful subgroups based on age, comorbidities, past medical history, etc., and then add genomic and other biomarkers, we can begin to ask whether these performance measures are, in fact, the correct ones across different subsets.
"That will allow us to make our guidelines and our performance measures more accurate and more meaningful. That's the learning part of a learning health care system. We can get smarter by studying the outcomes of hundreds of thousands of real-world patients."
What data will CancerLinQ use to report on patient outcomes?
"We recognize that the tracking of patients is going to be an important part of this program. There are plans to create a patient-facing side to CancerLinQ. We not only want to know how they are doing, but we also hope to provide long-term services to patients to help them through their cancer therapy and through the subsequent survivorship experience.
"We hope to convince patients of the importance of contacting us and staying in touch with us over time, because the consequences of cancer therapy can play out over many, many years. So we will use the patient-facing side to try to keep our records as up-to-date as possible, and we will tap into other databases to help inform CancerLinQ. For example, there is a national death registry that can be cross-matched with our database to make sure we understand as much as possible the status of patients.
"The initial release this fall to the vanguard practices will not have a patient-facing side because there is only so much we can do at once. My guess is that it's going to take another year or two for us to get to the patient-facing side. We have a committee of patient advocates working with us on CancerLinQ, and we are taking in their suggestions and trying to understand their needs."
What other kind of reports will CancerLinQ provide in its first iteration?
"We will take in practice management data and begin to show practices how their resource utilization compares with other practices-not by name, but in the aggregate.
"We would like to capture as much resource utilization as possible, including things like unplanned hospitalizations and visits to the emergency department. This is something that practices have a great need for, not only to understand their own data, but because this information is not well known across the whole field. Even if a given practice understands, for example, that for a certain type of cancer, it typically requires CT scans so many times per year, very few people can say how that compares with what others are doing.
"If a practice could see, for example, that their outcomes are in the top quartile of peer practices but resource utilization to achieve those outcomes is much higher than the peer practices, then that is extremely valuable information. That would allow practices for the first time to rationally adjust utilization of resources to maximize value for patients. So we think that this will be a tremendous resource as we move forward."
When will CancerLinQ be fully functional?
"There were about 15 major pieces of CancerLinQ when we put this program together. The initial release will have five or six of them and, over time, we will be adding more and more capabilities to the system.
"Building CancerLinQ is, in many respects, a project that never will be done. It will always be improving, adding new resources, deleting things that are not helpful, and adapting to new technologies and to the rapid changes in oncology."
CancerLinQ Who's Who
The first eight practices that will begin providing patient records to CancerLinQ later this year are:
* Inova Comprehensive Cancer & Research Institute, Falls Church, Virginia;
* South Coast Centers for Cancer Care, Bedford Falls, Massachusetts;
* New England Cancer Specialists, Scarborough, Maine;
* Medical Oncology Hematology Consultants, Newark, Delaware;
* Cancer Treatment Centers of America;
* Marin Cancer Care; Greenbrae, California;
* Space Coast Cancer Center, Brevard County, Florida; and
* Michiana Hematology-Oncology, Northern Indiana.
Another seven cancer centers are expected to join the effort soon so that about 500,000 patients will be represented in the first version of CancerLinQ.
Former ASCO President Douglas W. Blayney, MD, of Stanford Cancer Institute chairs the CancerLinQ Physician Advisory Committee. Other members:
* Paolo F. Caimi, MD, Case Comprehensive Cancer Center;
* John Deeken, MD, INOVA Comprehensive Cancer and Research Institute, Falls Church, Virginian;
* Jack Erter, MD, Tennessee Oncology, serving middle Tennessee;
* Mohamed Farhat, MD, Michiana Hematology-Oncology, serving several communities in Indiana;
* Susan Kim, MD, Southcoast Centers for Cancer Care, Bedford Falls, Massachusetts;
* Eric Martin, MD, Medical Oncology Hematology Consultants, Newark, Delaware;
* Patrick Mergler, manager of cancer informatics, University Hospitals Seidman Cancer Center, Cleveland;
* Jamal Misleh, MD, Medical Oncology Hematology Consultants, Newark, Delaware;
* Michael N. Neuss, MD, Vanderbilt Ingram Cancer Center, Nashville;
* John Niederhuber, MD, Inova Comprehensive Cancer and Research Institute;
* Ritwick Panicker, MD, Cancer Treatment Centers of America, Phoenix;
* David J. Perry, MedStar Washington Hospital Center, Washington, DC;
* Amitabha Sarma, MD, Virginia Cancer Specialists, (US Oncology Network), Leesburg, Virginia;
* Keith Thompson, MD, Montgomery Cancer Center, Montgomery, Alabama;
* Mark E. Thompson, MD, The Mark H. Zangmeister Cancer Center, Columbus, Ohio;
* Tracey F. Weisberg, MD, Maine Center for Cancer Medicine, Scarborough, Maine;
* Allison Zibelli, MD, Sidney Kimmel Cancer Center, Philadelphia; and
* Solomon Zimm, MD, Space Coast Cancer Center, Brevard County, Florida.