As Computers Learn to 'Talk' to Each Other, Patient Care Will Improve 

A Conversation with Edward P. Ambinder, MD

Get Permission

Because cancer is so data-intensive, oncologists need software that captures and automatically presents key data about patients in a standardized report or template, which can be sent to and understood by computers in other hospitals, office practices, patient homes, cooperative research groups, and registries. This requires agreed-upon standards between the data generators and recipients.

—Edward P. Ambinder, MD

Last fall, Edward P. Ambinder, MD, Clinical Professor of Medicine, Hematology, and Medical Oncology at the Icahn School of Medicine at Mount Sinai in New York, and a member of ASCO’s Health Information Technology Work Group, spoke about “The Information Age: Cyberspace and Cancer,” at the Chemotherapy Foundation Symposium in New York. During his presentation, Dr. Ambinder discussed how advances in information technology are changing the way cancer care is delivered.

Improvements continue to be made in software functionalities and in terms of interoperability between computer systems. Nevertheless, more work needs to be done to make health information technology more seamless and universally used, said Dr. Ambinder.

The ASCO Post talked with Dr. Ambinder about the importance of developing computer-compatible electronic health records, advances in DNA sequencing, and the implementation of ASCO’s CancerLinQ rapid learning system.

Technology-based Changes

Is technology making it easier for physicians to care for patients with cancer, and is it making it easier for patients to understand their disease and treatment?

Mobile hardware and operating systems that use touch and voice (eg, voice-to-text, voice-to-text-to-data, voice-directed navigation, and conversation—think Apple’s Siri) and incorporate user educational tools are beginning to transform how physicians interact with computers. They require a fraction of the time to learn, are much less costly to buy and use, can be easily updated using cloud-computing technology, and improve our office workflow compared to the older Mac or PC systems. These older programs are good for documenting patient information, capturing charges for billing, and helping with administrative tasks, but they are far from ideal to meet the current needs of the practitioner, office staff, and patient.

For patients, technology is making it easier to access information about their disease and find clinical trials. But the problem is the overwhelming amount of information available online and the difficulty of not having good guidance about where to get reliable, accurate information. Patients really have to rely on their physicians and nurses for help; clinical decision-support tools and links from the electronic health record to appropriate websites will encourage this trend.

Computational Genomics

How is using technology to decipher a person’s genetic mutations in cancer cells improving cancer care?

There are 3.2 billion base pairs or genetic “letters” and over 20,000 expressed genes present in our DNA. What we are learning is that most cancers that we treat have gene mutations and we can now collate these mutations, giving us diagnostic, prognostic, and predictive information about the cancer. By predictive, I mean we are beginning to be able to predict which medications, chemotherapy drugs, or monoclonal antibodies will be a specific and effective treatment for the cancer.

As the price of utilizing this technology falls and dramatic advances in DNA sequencing continue to be made possible by computational genomics, most patients with cancer will automatically get a complete genomic analysis of their disease, and we will be able to provide more targeted therapy. We will be overwhelmed with data and must rely on computers to maintain this patient information and help guide us in choosing the best cancer treatment.

If we were asked to stage a patient with cancer 9 or 10 years ago, we would basically use the TNM staging system. We would look at the size of the tumor, whether or not lymph nodes were involved, and whether there was metastatic disease. Now, in addition to all that, we are beginning to get significant amounts of genomic, proteomic, and other molecular data that will begin to guide us in better staging the patient with cancer and more precisely choosing the patient’s therapy.

Communication between Systems

At the Chemotherapy Foundation Symposium, you talked about the need for developing an electronic health record that would contain common data fields for different cancer types and treatments and that could be used universally across different computer platforms. Is that happening now?

Because cancer is so data-intensive, oncologists need software that captures and automatically presents key data about patients in a standardized report or template, which can be sent to and understood by computers in other hospitals, office practices, patient homes, cooperative research groups, and registries. This requires agreed-upon standards between the data generators and recipients.

That’s what we are beginning to do with ASCO’s Health Information Technology Work Group and ASCO’s CancerLinQ rapid learning system ( Hopefully, with the assistance of electronic health record vendors, the office and hospital data from our electronic health records will be shared with CancerLinQ and other cancer stakeholders, permitting us to create standardized reports and have templates shared by the various cancer registries. ASCO’s Oncology Standards Summit, which took place in January, will bring together 21 of these groups to begin this process.

Right now all this information is captured manually or in an electronically proprietary way. If we can use the computer to do this for us automatically, it will save a lot of time and effort and make a major contribution to significantly improving the workflow and efficiency in all our offices. Ultimately, we will have a health-care system with a continuously learning feedback process.

Future of the Electronic Health Record

Do you envision that the electronic health record would be a complete document that holds the patient’s medical history as well as information about clinical trials and survivorship care plans?

No. The way I envision the electronic health record is that it will keep the data on patients, document our encounters, and assist with billing and other administrative tasks. But it will need to be electronically interoperable with other sites for obtaining clinical decision support data, providing registry reporting, and other information sharing with all health-care stakeholders including the patient.

Some people are looking at the Apple app model, by which different apps could work with other apps with the electronic health record database. For example, you might have apps that are easily updated with the ability to query the electronic health record and capture the data necessary for the electronic prescription app, the clinical and genomic staging app, the cancer guidelines and care plan app, the clinical trial app, the cancer registry app, and ASCO’s CancerLinQ app.

Thus, the raw patient data resides in the electronic health record, but the software to use that data for a specific purpose will reside at an external site. The app would act as an intermediary to get the data from the electronic health record, bring it to the site that needs the data for analysis, and then come back to the electronic health record with information that is appropriate for the particular patient.

Information for Patients

Could there also be an app for patients to maintain their own health records, including their treatment and survivorship care plans?

Sure. At ASCO a few years ago we began to define a chemotherapy treatment plan based on specific cancer types that we could give patients before they started their chemotherapy. It would give patients information about the chemotherapy drugs they would be getting, their treatment schedule, and the side effects they could expect.

We also developed a treatment summary report for the cancer survivor. This would give the patient all the basic information we had already created for the chemotherapy treatment plan and also have information about the current state of the cancer, what the response was to treatment, what significant adverse events occurred during treatment, and how patients should be monitored after finishing treatment.

These plans are now combined in the Chemotherapy Treatment Plan and Summary templates ( This information is helpful for both the patient and the medical team caring for the patient post-treatment.

One of the big criticisms of cancer care today is that there is this disconnect between the time a patient completes treatment and his transition back to the primary care physician. Without a seamless way to electronically communicate the patient’s medical record to all the necessary parties, it makes it very difficult to maintain continuous, coordinated care, and that is one of the major goals of our health-care system. ■

Disclosure: Dr. Ambinder reported no potential conflicts of interest.