The only way to generate new knowledge is through the integration of clinical data with research data.
That, in any event, is the prevailing view in research communities. And looking ahead, it’s also what’s needed to expand health care capabilities, according to the head of leukemia-cancer research at CHU Sainte-Justine Research Centre – Mother and Child University Hospital (CHU) in Montreal.
“That probably sounds easy on paper, but it isn’t all that feasible in real life,” said Dr. Daniel Sinnett. “But, if we are capable of doing it, it will be a major achievement in the treatment of disease.”
The challenge for Sinnett and his charges is to integrate the “huge amount” of research data they compile on a daily basis. It’s a process necessary to understand and treat complex deadly diseases like cancer.
Sinnett brought IBM aboard in early November to supply the technical expertise he needed for an informatics infrastructure.
CHU is working with IBM to develop tools to access clinical data wherever it is in the hospital, in whatever format, Sinnett said. CHU will then have new knowledge generating capacity that will reveal new diagnosis methods and predict who is susceptible to a given disease.
“We chose to concentrate on childhood leukemia because it is a very successful area of progress at the hospital,” Sinnett said.
For its part, IBM said it is teaming with CHU because it recognizes the critical role IT can play in 21st century health care. Specifically, IBM will help CHU better define the genetic markers for acute lymphoblastic leukemia (ALL).
Sinnett likens his new informatics infrastructure to many villages connected by highways.
For example, he said, radiology, biochemistry and pharmacology can all be considered individual villages that, in the past, had no means of traveling from one to the other. What he needed to figure out was a way to put highways between the villages.
“That’s what I call integration,” Sinnett said. “The technology is the highway, that’s the simplest way to put it.”
Now CHU’s researchers will have real-time access to a body of quality data, using a technology that is easy to navigate and accessible to the research community, Sinnett said.
Clinical data that was once manually extracted from the hospital’s patient file will now be electronically transmitted and merged with genomic data to create a Medical Information Repository (MIR).
“When you link the data it is possible to discover patterns and predictors that will define who will respond or not to a given treatment, who will develop side effects, or who will have a higher or lower risk to relapse,” Sinnett said. “It is the integration of all the data that will develop this patient knowledge. But if we can’t get access to the medical information the process is impossible.”
The MIR hopes to reduce the query process from days to minutes and allow researchers to develop personalized therapies for patients and to keep longitudinal records.
“Hopefully, by the end of this year or the beginning of next year, we should have a pretty good idea whether or not it is going to work the way we planned,” Sinett said. “It is an essential tool to do research here.”
Sinett said clinical data, if updated more quickly, will also have a profound impact on patients whose conditions could worsen. It would be easy to imagine that the technology would be useful if you wanted to follow a group of obese individuals because you know that some of them will develop diabetes.
There was a long period of planning and discussion before CHU partnered with IBM.
“We knew what we wanted, but to get there, that’s a long ways,” Sinnett said.
“(IBM) had to understand all the processes that are used in (CHU’s) environment,” said IBM Life Sciences spokesperson Yvan Foster. “That’s the reason why it took so long to implement the technology.”
Foster said it is important to know what researchers are doing in their day-to-day work, what kind of data they wish to use and how the data will be used in the future because information is always expanding – the researchers are always thinking about new ways to use and combine the data.
“You have to build a database that will be flexible for future diagnostic targets,” Foster said. “In the future (CHU) may want to integrate environmental data, including family histories, or if (the patient) is a smoker, things like that.”
The more you add and update data the more you need a database that can accept the new data, Foster said. Then you can add analytic tools to mine the data.
Brian Eaton (email@example.com) is associate editor with InterGovWorld.com.