Oracle Applications Enable Personalized Cancer Treatment

Moffitt Cancer Center in Tampa, Fla., has seen the first fruits of its partnership with Oracle, which helped Moffitt develop its Health and Research Informatics (HRI) platform. The data warehouse and analytics system, which went live last October, has already enabled Moffitt to identify patients suitable for clinical trials much faster than in the past. As a result, the cancer center has been able to qualify for additional research grants, according to Mark Hulse, Moffitt's vice president and CIO.

That's only the first step, Hulse told InformationWeek Healthcare. Oracle's healthcare data model and high-level analytics give Moffitt the ability to create personalized approaches to treating cancer, he said. The HRI platform allows the institution to aggregate and analyze diverse clinical and outcomes data. The resulting information will be fed into "clinical pathways" to help physicians determine the best and least toxic methods to treat individual patients.

Moffitt's Total Cancer Care program, which aims to foster personalized medicine, involves Moffitt and 17 other hospitals in 10 eastern states. The participants supply data that allows Moffitt researchers to look at particular types of cancer in particular stages to see which kinds of therapies are most effective in individuals. This requires analysis of not only clinical data, but also genomic and molecular data.

[ Most of the largest healthcare data security and privacy breaches have involved lost or stolen mobile computing devices. For possible solutions, see 7 Tools To Tighten Healthcare Data Security. ]

When Moffitt chose Oracle as its partner a year ago, the challenge was to figure out how to combine all of this data in such a way that researchers could use it. Some of this information was collected in cancer registries at the state level, Hulse noted, and this data had fewer disparities than did the data in the source systems.

But there was also information from disparate electronic records, labs, and tissue sample repositories, as well as patient self-reported data. The latter, which so far is confined to Moffitt but will be expanded to other facilities, includes information from a structured questionnaire that patients fill out.

To provide the solution that Moffitt was seeking, Oracle applied elements of its enterprise healthcare analytics suite, including a healthcare data model that the company had been developing for several years.

The first step in this approach is to "normalize" the terminologies used in each of the source databases, said Kris Joshi, vice president of health product strategy for Oracle, in an interview with InformationWeek Healthcare. In other words, disparate terms for the same medical concept are mapped to a normative terminology that encompasses but goes beyond systematized nomenclature of medicine (SNOMED), Joshi said.

Second, Oracle assembles the information into a single database by assigning metadata tags that link back to the source data. This "metadata management" also tells users where the data came from and what version of the system generated it.

The end result is to create a unified database that's fit for clinical research. Although this relies mainly on clinical data at Moffitt, Joshi noted, it could also incorporate ADT and billing data if researchers wanted to measure the cost of interventions.

Oracle has many analytic tools that the researchers can use, Joshi pointed out. The company has also built an "app exchange" and has invited third-party developers to develop analytic software that integrates with the Oracle platform. This is good for the developers, because it helps them get noticed, said Joshi. It's also good for customers who want to buy a "best-of-breed" application.

"They can deploy it on top of this platform without worrying about the back-end integration, which is often the most expensive part of putting in a new application," he noted.

Marc Perlman, global vice president, healthcare and life sciences for Oracle, said that the company is trying to increase the size of the market for its enterprise analytics tool. "We're trying to reduce the cost of ownership and increase the value. Analytics have been an area with an elusive value in the past. But we think the strategy will work, and Moffitt's a good proxy for that."

Moffitt also has plans to commercialize its research. A for-profit subsidiary of the cancer center is working with pharmaceutical companies, which hope to leverage Moffitt's discoveries to help develop new cancer drugs.

Hulse says it can take a pharma company two or three years to find enough patients who fit the criteria for a phase two or three clinical trial. "But using the HRI platform, we have that data electronically on the Total Cancer Care patients. So in some cases, we've been able to condense that down to a couple of hours."

Healthcare providers must collect all sorts of performance data to meet emerging standards. The new Pay For Performance issue of InformationWeek Healthcare delves into the huge task ahead. Also in this issue: Why personal health records have flopped. (Free registration required.)