By Sarianne Gruber
Twitter: @subtleimpact
The COTA story is The Little Engine That Could tale, for a group of oncologists and hematologists at the John Theurer Cancer Center, beginning with “I think I can” build a database to arriving to “I thought I could” transform the delivery of healthcare. I had the pleasure to hear firsthand from Dr. Stuart Goldberg, a Hematologist/Oncologist and COTA Board member, and learn what it took to cross the data frontier starting with the gaps in information access to evolving into an information focal point for physicians, patients and payers. Dr. Goldberg has been at the center for approximately 17 years where he first treated blood cancer patients at the Bone Marrow Transplant division. In 2002, he founded the Acute Leukemia Section, which also treats serious blood disorders such as myelodysplastic syndrome. He has performed over 3,000 bone marrow transplants and provides care for over hundreds of patients with acute leukemia and several hundreds of patients with other types of leukemia. The John Theurer Cancer Center (JTCC) at the Hackensack University Medical Center is one of the largest leukemia centers in the United States, located in Hackensack, New Jersey. Here is how their story begins.
The Daily Hurdles Directed the Database Design
Many of the center’s hematologists and oncologists have a very strong background and interest in research, and are actively conducting a lot of clinical trials. About 10 years ago, the need became apparent for these doctors to do their daily work of taking care of patients and their research, they would need a functional database. ” One of the problems in medicine has been that we really don’t bench mark ourselves, and we don’t keep track of statistics on how well we are doing or how well patients are doing. If somebody asked an average oncologist how many patients you seen with breast cancer, they may say they see a lot of them, but they couldn’t give you an idea of how many or how they were doing. So it became clear to us that we needed that kind of basic information to run our daily business, but also to do some of the research we wanted to do”, stated Dr. Goldberg.
An expert in Chronic Myeloid Leukemia, Dr. Goldberg runs many studies and the task of collecting patient study samples for clinical trials had become exceedingly arduous. He said that he would get calls from pharmaceutical companies all the time asking him if he would like to do a new kind of trial. The first question he was asked was how many patients do you see. He would say he saw a lot but had no idea how many. Nor could he tell anyone how well they were doing or how many had problems that maybe the new drug would be able to solve, unless he started to pull charts and start looking for them. Dr. Goldberg pointed out that EHRs didn’t solve this problem either, since they had become electronic versions of charts and that you still had to go back and try to find the charts.
If you asked a doctor how many patients they have with a certain disease, the first thing they would do is probably go to the ICD-9 codes. “The problem is that the ICD-9 codes and even the new ICD-10 codes are very limited. Basically, for breast cancer there is one code – breast cancer”, remarked Dr. Goldberg. He explained that they now know, and it was obvious when they did research, that breast cancer is not just one disease but hundreds of diseases. There are women who are hormone receptor positive, some are hormone negative tumors or others may have HER2/neu and antigen on the outside, which is very important when selecting treatments. Next level of disease stratification would be the size of the tumor, the stage of the tumor (if it is an earlier stage or metastatic, stage 1 through stage 4) and none of this information is captured in the ICD-9 codes. “If you ask me how many breast cancer patients there are, I may be able to hit a couple of buttons on the EHR, go to the insurance company, or the federal government and pull out the SEER data or Medicare data, but all I would get is the number patients with breast cancer. I would not get a more detailed diagnosis”, declared Dr. Goldberg. It was at that point in time, the center set its sights on forming a database within their own hospital to help them with their daily work. What evolved was a company called COTA, which stands for Cancer, Outcomes, Tracking and Analysis.
From Textbooks to a Taxonomy of New Codes
A taxonomy of new codes was created that would classify each patient into prognostically relevant categories. When seeing a patient from any center that is using COTA tools, information is pulled out of the EHR, and organized it in a way that makes oncologic sense. “When we will pull out records on breast cancer patients, we can see the size of the tumor, how many lymph nodes were involved, whether it has metastasized, if it is hormone positive, if it is HER2/neu positive, etc. Now we have thousands of codes for breast cancer like a dewy decimal system. We put a number on every different breast cancer and now we have thousands of codes for all breast cancers. And we can digitally search”, emphasized Dr. Goldberg. The center spent over three years going through the breast cancers looking at all the things they wanted to count and those they didn’t need to count. They resorted to text books to look at all the relevant information that they would need to know about the disease. The system was structured so doctors could query the data base. “Once you are able to group and sort patients into a finite prognostic groups, well you now have the ability to do a lot of high end analytics. And really be able to advance medical care and also value-based medicine. We can get into cost analysis because you can now truly compare apples with apples. How are your advanced stage 4 with breast cancer hormone positive doing against mine? We can compare truly like patients, “ declared Dr. Goldberg. What this does mean when you want to analysis the data? It removes the biologic variance and allows the physician variance or the treatment variance to be measured. This is essential when evaluating value-based medicine. Dr. Goldberg described an analysis that profiled one physician group with 5 different doctors. They selected one specific type of breast cancer to evaluate, the most common type that group was seeing. Out the 5 doctors, they were using 2 or 3 of the same chemotherapy regimens, but the costs were dramatically different while the outcomes were very much the same. When they delved down deeper into data, they found that one doctor was using a lot of growth factors inappropriately which mapped to patients being admitted to hospital more often. Accessing the same patients types, not sicker or less severe, all prescribed the same first line of therapy, with the computer is easy to do. Each doctor can download a “very short list” for every single cancer, unique to each cancer. Considering the different cancer divisions, leukemia items are different than what information the lung cancer doctor would care about, and so forth for each disease. “We pull out the 10, 15 or 20 different factors that are important for that disease, and digitalize that information. We follow what treatments they got and what the outcomes were. Very simple. With that we can find interesting patterns, “proclaimed Dr. Goldberg.
About the John Theurer Cancer Center, Regional Cancer Care Associates
The John Theurer Cancer center is a disease-specific organization. Twelve divisions include multiple myeloma, gynecologic oncology, lymphoma, neurological oncology, gastrointestinal oncology, thoracic oncology, breast, leukemia, skin and sarcoma, geriatric oncology, head and neck and urologic oncology. Each specific expert group has over 30 physicians with their own specific areas and expertise. On average 7,000 brand new patients are treated per year at the JTCC, ranking them in the top 10 cancer centers in the United States. The JTCC oncologists and hematologists participate in the Regional Cancer Care Associates. The practice extends across the entire state and treats approximately 60% of all cancer cases in New Jersey. More information is available on their website.
To learn more about COTA and Dr. Stuart Goldberg, please visit the Cancer Outcomes Tracking and Analysis website.