Why Medicine Needs Big Data
The domain of big data and its associated algorithms make it possible to access and analyze vast stores of structured and unstructured data, leading to informed decision-making and the efficient allocation of resources. While many businesses are benefitting from this new breed of tools, the medical field has not experienced the same level of adoption, but the promise of personalized medicine may change that.
Even while medical science embraced the computational era wholeheartedly, the larger healthcare industry resisted digitalization. For example, the industry has yet to adopt clear standards for sharing patient electronic medical records for research purposes.
Reflecting on the status quo bias, Dr. Clifford Hudis, current president of the American Society of Clinical Oncology (ASCO) and chief of the Breast Cancer Medicine Service at Memorial Sloan-Kettering Cancer Center in New York City, makes the observation that “medicine is astonishingly archaic functionally.”
“As doctors, we write down our observations, but we have not pulled all these disparate data points together to make references, draw conclusions and make decisions,” he states in a recent Tech Page One blog entry. “Even with revolutionary electronic medical records, in many cases we’ve simply electronified the old systems. Medicine has yet to catch up with the business world and its use of big data.”
There are signs of hope however. There is a clear movement by healthcare institutions toward personalized medicine approaches that provide treatments tailored to individual patients. The University of Pittsburgh Medical Center (UPMC), for example, is conducting a five-year enterprise analytics study using big data to advance discoveries in the personalized medicine arena.
UPMC’s new enterprise data warehouse has enabled researchers to electronically integrate the clinical and genomic information of 140 breast cancer patients. One of the primary questions that researchers sought to address was whether there was a notable difference between pre-menopausal and post-menopausal breast cancer.
“We are interested in this question from a research standpoint because we are moving toward personalized medicine, and personalized medicine is all about finding subgroups of patients who have a specific type of disease for which we could develop novel therapies,” notes Adrian Lee, Ph.D., a renowned expert in the molecular and cellular biology of breast cancer, and director of the Women’s Cancer Research Center at the university’s Cancer Institute.
The research aims to accumulate as much data as possible. Tumor samples are collected from a large number of cancer patients, and then analyzed using sequencing and other techniques. Tumor types are then characterized, which enables targeted therapies to be developed. Once the data has been analyzed and quantified, it is shared with an international team of researchers.
The greatest barrier to advancing these types of studies is the sheer volume of data. There are billions of measurements and measurement combinations, and all the data needs to be correlated and aggregated. The discipline requires that new algorithms to be developed and run on high-performance computers.
Lee believes that medicine is on the brink of a transformation, such that advances in personalized medicine will lay the way for pharmaceuticals that are as unique as the patient taking them. The UPMC study has already made some important discoveries and it still has three more years to go. The team found “intriguing molecular differences” in the makeup of pre-menopausal and post-menopausal breast cancer tissue. It’s an important detail that brings scientists one step closer to the development of targeted therapies.