All our patient data has value, and the oncology department, for instance, holds on to data for at least 10 years, whatever the course of the illness," says Robert Cecil, business-technology director for radiology at Cleveland Clinic Foundation, a health-care company that provides services through 11 hospitals in Ohio. "Even when patients expire and we have the ability to delete the information, the value of it in research could change the life of a new patient."
Cleveland Clinic faces multiple challenges: It must find ways to manage and store large amounts of data, deploy systems that can access and retrieve needed data quickly, and have the capability to analyze that medical data to find new ways to help ailing patients.
Cecil has storage needs to rival that of any business-technology executive. He expects the amount of data the clinic collects to grow at a rate of 2 terabytes per week by next year. "The screens get bigger and the equipment works faster, and a 5-Mbyte to 10-Mbyte image could catch an illness earlier," he says.
To handle the data, Cecil is looking at hierarchical storage management, in which information is kept on different types of storage media based on how current the data is and how quickly it might be needed. Over time, rarely used data is shifted to lower-cost media such as tapes, while data that might be needed quickly is kept on high-performance hard-disk drives. Cecil is now overseeing the migration of some of the clinic's less-used data to higher-density and lower-cost storage systems.