Canadian student uses BI to study cancer survival

Could it be genetics, molecular alteration, medical treatment or socio-economic factors? Physicians and researchers are still not sure why some breast cancer patients tend to live longer than others.

In an effort to determine what factors play a crucial role in patient survival, a second-year Canadian graduate student has turned to a suite of software tools more often used by accounting firms and financial businesses.

Darren Brenner, a sciences graduate student at Brock University in St. Catherines, Ont, is using business intelligence software from SAS Institute to crunch through and analyze massive amounts of social, genetic and biological data on about 2,000 cancer patients. He hopes his "holistic approach" in the study of the long-term survival of some breast cancer patients will eventually lead to better treatment strategies.

"I hope my findings will help aid patient and clinician decision-making and guide more intensive therapy and patient monitoring and ultimately, improved breast cancer patient survival," Brenner said.

Business intelligence (BI) software gathers data from various sources and is able to analyze the information to provide organizations with a more comprehensive knowledge of the factors affecting their business. Using data such as sales metrics, production and internal operations, BI applications can also provide predictive analysis that aid decision making.

Banks and other businesses have relied on BI tools to mine and analyze data to improve production, performance and marketing campaigns. Governments and insurance companies are also using the tool to analyze "social networks" and uncover money-laundering operations, organized insurance scam rings and tax evaders.

Brenner is culling information from three large data sets: clinical studies; census reports; and the cancer registry. Variables including genetic factors, molecular changes, clinical measurements and social factors such as educational attainment and income bracket will be fed to SAS BI software tools for analysis and correlation.

The researcher hopes to determine the factors related to long term survival in cancer patients.

Similar studies have been done in the past, but none have included the combination of data sets and predictors that is being used in this study, Brenner said.

The next step, he said, is to use BI tools in analyzing various factors that have a role in the development of cancer. Such a study will help improve detection and prevention of the disease.

Canada spends at least half a billion dollars a year on cancer research. The money comes from various sources such as fundraising events and tax payments. Only two per cent of the funds go to research bolstering prevention, 22 per cent is used for improving treatment and 56, per cent in researching the biology and etiology of cancer, according to a recent study by the Canadian Cancer Research Alliance.

The use of BI tools in medical research is a very logical transition, according to one Toronto-based analyst.

"The process of mining data from various sources and analyzing it remains the same. Only the context has changed," said Carmi Levy, senior vice-president of consulting firm AR Communications Inc.

He said the strengths of BI technology, although proving very useful in the corporate world, lends itself to the medical research realm. "The technology allows researchers to find that needle in the haystack – the useful knowledge residing in those stacks of data."

Using BI software could also be a cheaper alternative to less accessible high-end supercomputing, Levy said.

He noted that in some research efforts supercomputers such as IBM's Deep Blue were employed. "Not all researchers have the funding or the access to these big computers." He said BI technology, however, is still "very young" and vendors as well as users have not "fully explored its potentials" in such areas as medical research.