The Data Chief | Episode 9

Moffitt Cancer Center's Dr. Dana Rollison on accelerating scientific discovery with data

Dr. Dana Rollison

VP and CDO

Moffitt Cancer Center

Current EpisodeEP9: Moffitt Cancer Center's Dr. Dana Rollison on accelerating scientific discovery with data

Episode Overview

Joining Cindi today is  Dr. Dana Rollison , Vice President, Chief Data Officer, and Associate Center Director of Data Science at Moffitt Cancer Center. Dana has been with Moffitt for over 16 years in varying capacities since earning her PhD in Epidemiology from Johns Hopkins University in 2004. In this episode, she and Cindi discuss her work at Moffitt, how data is accelerating scientific discovery, and what leveraging data from the lab to the clinic looks like in practice.

Key Discussion Points:

  • What makes Moffitt Cancer Center unique? Moffitt is the only National Cancer Institute-designated comprehensive cancer center in the state of Florida, and its level of data and analytics maturity leads in healthcare, a field that traditionally lags.
  • Moffitt’s Total Cancer Care protocol seeks to enroll every one of the center’s patients in a research study to find new prevention and treatment strategies through collection of unique data sets. 
  • Trust matters. Dana’s win/win approach to data collection earns trust from patients with privacy concerns. This helps her gather the most complete and useful set of information for doctors and researchers to arm themselves in the battle against cancer.
  • Data has a place in every department. Leveraging data to work in the best, practical interests of all departments and partners provides better integration for everything from clinical effectiveness to operational logistics.

Key Quotes:

We are on the right path toward investing in data. … Some of the efforts to harmonize capturing of data across medical records and sharing data through health information exchanges has gone a long way toward improving outcomes with chronic illnesses [such as diabetes and heart disease]. Cancer is a little more complicated. It often involves imaging tests or other circulating biomarkers. … I think initiatives like CancerLinQ and others that are working towards standardizing how we collect cancer data so that we can truly learn from each other’s experiences and improve outcomes is a work in progress and something we should definitely continue to advance.
It’s not just how accurate the algorithm is, but also how well the physicians understand it. They’re less likely to trust a black box. If they understand what’s behind the black box, then they may be more inclined to use it.
The combination of those two approaches: educating so that people understand what these algorithms are doing and how they’re created, how they’re validated, as well as positioning them to really be accelerators of the physician’s day-to-day activities and not seen as a replacement, are key to the adoption of these technologies.
If [patients] know why we’re collecting the data – that it can both help us treat their cancer better, but also help learn about what may have caused their cancer or what may be a predictor of how they do with their cancer that can help other – then I think people are more likely to take care and the completeness of the response and not view it as an administrative burden to fill out this form.
It’s those who can bring together the group of experts to tackle that problem that are going to be successful. So you have to know a little bit about all of those areas you’re trying to bring together enough to speak the language of those different team members and to relate to them about what it is you’re trying to do and what role you think they can play.
If the cancer patients have failed conventional therapy and are on their fourth or fifth line of treatment and really have no treatment options left, there can be utility in looking at the different signaling pathways or mutations that are active in the tumor and seeing if other patients in our Moffitt history have had similar mutations, maybe in a different cancer. And if so, did a particular drug work for that patient? Because that could be a clue for what could work for this patient. And that’s almost a last resort, but having those data available to us through hundreds of thousands of patients may provide that glimmer of hope for the patient sitting in front of us today.

More About Dana:

Dr. Dana Rollison earned her undergraduate degree in Biology with honors from the University of Miami in 1997 and completed her master’s (1999), PhD (2002), and postdoctoral training (2002-2004) in Cancer Epidemiology at Johns Hopkins University. Dr. Rollison joined Moffitt Cancer Center in 2004 as an Assistant Member in the Cancer Epidemiology Program and began serving as the Scientific Director of Moffit’s Cancer Registry in 2005. She was promoted to Associate Professor with tenure in 2010, and became a Senior Member in 2017. Her independent research focuses on the application of data science techniques to enhance cancer surveillance and etiologic studies of cutaneous viral infections, environmental exposures, and immune function. 

Building upon her experience with linking Cancer Registry data to other administrative, research, and clinical data, Dr. Rollison began serving as Moffitt’s Vice President, Chief Data Officer (CDO) in 2010, setting an executive vision for Moffitt’s enterprise-wide data warehouse and analytics strategy to support team science, clinical pathways, accountable care analytics, data sharing partnerships and the practice of personalized medicine. As CDO, she oversees three Departments within Health Data Services, including the Cancer Registry, Data Quality, and Business Intelligence, and Health Informatics. Current initiatives within Health Data Services include the use of artificial intelligence for mining text within the electronic health record, platform-agnostic curation of molecular data, and migration of enterprise analytics to the cloud. 

Recognizing the need to grow intellectual capital in parallel with technical infrastructure, Dr. Rollison expanded her leadership portfolio in 2017 by becoming Moffitt’s first Associate Center Director (ACD) of Data Science and Division Chief for Quantitative Science, overseeing the academic departments of Integrated Mathematical Oncology and Biostatistics and Bioinformatics, as well as two data-related Shared Resources. To better facilitate the development of novel quantitative methods and foster team science, Dr. Rollison pioneered the creation of a new academic track, the Integrated Scientist track, and established a new, third Department of Machine Learning, focused on advancing cancer research through the analysis of complex data sets, including imaging and the IoT. 

Leveraging her dual role as CDO and ACD, Dr. Rollison seeks to optimize Moffitt’s extensive data assets across clinical and research areas of the organization, thus accelerating scientific discovery and translational research across the cancer continuum, and ultimately advancing the prevention and cure of cancer.