Headshot Big Data and Oncology CareA recent report in JAMA Oncology said 14.9 million new cancer cases were diagnosed worldwide in 2013 along with 8.2 million cancer deaths and 196.3 million disability-adjusted life-years of living with cancer.

Those are big numbers that quantify a big problem that requires a big solution. Big as in Big Data, or advanced data analytics.

The timing appears just right. The percentage of health care organizations surveyed by HIMSS Analytics that have a clinical and business intelligence solution plugged in rose to 52.1 percent this year from 46.2 percent in 2013. Separately, the American Society of Clinical Oncology unveiled its CancerLinQ Big Data initiative that initially will aggregate treatment and outcome data from 15 health care organizations starting later this year.

McKesson spoke with two oncology data experts about the role Big Data can play in cancer care: Debra Patt, M.D., director of policy at Texas Oncology and director of informatics at McKesson Specialty Health and The US Oncology Network, and Heather Morel, vice president and general manager for health informatics and reimbursement/access services at McKesson Specialty Health.

Info Big Data and Oncology Care

How would you define Big Data as it applies in the health care setting?

Patt: Big Data in health care really is the aggregation of patient information, treatment protocols, therapy results and comparative effectiveness research in one place and then using it to understand clinical risk and improve individual and population health.

Morel: That's exactly right. It's not just about the fact that you can aggregate data and bring it together. It's about the ability to then present it in a meaningful way in the workflow of clinicians as they go about their tasks day in and day out to make the best medical decisions for their patients.

How can Big Data help oncology practices provide better care to their cancer patients?

Patt: Oncology in general is a very data-driven medical specialty. We all want a good bit of evidence for the things that we do. Most oncology care lends itself to evidence-based decision making. So what Big Data has given us is the ability to better understand how treatments and other therapeutic support impact patient care and clinical outcomes. The data we use to justify patient treatments and for the FDA to consider approval of therapeutic interventions is based on clinical trials. While clinical trials are critically important in the advancement of cancer care, their application in the phase II and phase III setting is usually limited to healthier patients without a history of other cancers. This means that when we then use these treatments in the real world clinical setting, we are supporting utilization of therapeutic intervention in a healthier group of patients on average than what we see in day to day practice. Big data allows us to use real world data to learn what influences the outcomes of real cancer patients and adds a great deal to what we learn from clinical trials.

Morel: New research and studies on cancer prevention, screening and treatment come out almost daily, That can be overwhelming for oncologists and oncology practices. It's difficult to keep up with the latest information. Big Data can simplify an enormous amount of data for oncologists and bring up the most important information at the right time in the right place.

What specifically is Big Data's patient job to be done in oncology?

Patt: Big Data enables oncologists to apply all that aggregated information to a specific patient and tailor a treatment plan for that specific patient. It tells you what will and won't work for an individual patient. I compare it with art. We were using crayons and markers to draw up general treatment plans. We now know how individual risk factors – age, genetics, comorbidities – will affect efficacy and toxicity that, in turn, influence outcomes. We're using different brushes and color palettes for each patient to paint individual portraits of cancer care. Using big data and predictive analytics to understand how to influence patient outcome in a more meaningful way will allow us to improve individual and population health—the application of the art of medicine.

Morel: Big Data also lets clinical trial researchers comb through patient health records and identify a practice's patients with select medical characteristics who potentially could benefit from a new drug, therapy or treatment being tested.

How do financial considerations factor into Big Data's oncology formula?

Morel: The data set from which oncology practices can pull information to make treatment decisions ideally would include the patient's economic situation, his or her insurance status, insurance benefits and potential out-of-pocket costs. All those things are variables in a very complex decision that needs to be made between doctor and patient. If, for example, the oncologist finds out after the fact that the selected course of treatment is unsustainable financially for the patient and will lead to other socio-economic problems, that treatment option may not have been the best choice. Perhaps another course of treatment that produces some clinical benefit at less cost may have been better.

Patt: Having financial information is useful because it makes the data multi-dimensional. It helps us understand each patient's risk of financial toxicity in addition to clinical effectiveness and medical toxicity, and to intervene as necessary to truly change how we're delivering care. As therapeutic innovation in cancer care is skyrocketing, so too have the cost of some cancer drugs. For some patients, the financial toxicity that could ensue is treatment prohibitive. As clinicians we need to help patients understand all their options so they can make the right choices for their own treatment.

How can Big Data help oncology practices succeed financially under value-based risk contracts?

Patt: Oncology practices are taking on the financial risk for the health of their patient population under these alternative payment models. Big Data – specifically predictive analytics – lets practices identify patients at high risk for emergency care or hospitalization. Oncologists can then modify the treatment plans for those patients to reduce that risk and improve outcomes. ER visits and inpatient hospital stays are big cost drivers and can eat up a lot of margin under those contracts. In the future, home monitoring and sensor technology will send oncologists real-time data on patients' condition that will allow them to address a problem before it gets worse. For example, Big Data could alert me to a patient whose has postural instability and a drop in blood pressure. I then have an opportunity to intervene and bring her to the clinic and hydrate her before she falls, breaks her hip, goes to the ER and then needs surgery and rehabilitation. We will have the ability to intervene before something bad happens thereby preventing a bad outcome and the resultant healthcare utilization and cost.

Morel: Debra comes up with these great ideas, and it's our job to figure out how to make them real. That's good for oncology practices and good for their patients.

For more information using data to improve oncology care, please visit the oncology solutions section on the McKesson Specialty Health website.

To learn more about how alternative payment models for cancer care are affecting oncology practices, read “Prepping Oncology Practices for Value-Based Reimbursement” on McKesson.com.


About the author

McKesson editorial staff is committed to offering innovative approaches and insights so that our customers can get the most out of the health care solutions they have and identify areas for operational improvement, revenue growth and improved patient satisfaction. If you have a suggestion for a blog topic you’d like to see covered, let us know in the comments.