ORLANDO, Fla.--How physicians and researchers make genomic and molecular data available to stratify patient populations based on various criteria (such as the diagnosis of a patient’s stage of cancer) may be able to help clinicians group patients to determine best practices for clinical decision support (CDS), stated David Fenstermacher, PhD, chair of the department of biomedical informatics at the Moffitt Cancer Center, at a Feb. 22 educational session on personalized health and genomics at HIMSS11.
“Genomic information will be a large asset to personalized medicine as the data could help physicians understand what the patient's determinates are and how the proposed actions of [CDS] may affect the biology of the patient setting,” said Fenstermacher in an interview.
The Moffitt Cancer Center is a nonprofit stand-alone cancer center in Tampa, Fla., which includes a hospital, clinic and research institute.
Using genomic information in a cancer patient for example, clinicians can look at a tumor using new technologies such as next generation sequencing that will allow clinicians to understand what the genetic make-up and biological determinates of that tumor are. “As more and more drugs are targeted for specific genes or molecular pathways, we can determine whether or not a person will respond to a drug given a particular chemotherapeutic agent just by the genetic make-up of that tumor,” said Fenstermacher.
“Hopefully, we could additionally assess through biomarkers whether or not the patient is responding to the therapy to demonstrate usefulness.”
However, Fenstermacher noted that genomics added to personalized health is not just about gene expression but also about linking relevant data together across the biological system to understand what are the consequences within genomic pathways and networks. This will require a huge amount of data to get the biological systems linked together using rich annotation resources so clinicians and researchers can quickly understand the context of the molecular data.
Fenstermacher made no mistake about it: This integration will not be a cakewalk.
“Taking advantage of genomic information will play a huge role for better treatment of patients, but gaps in patient-based data are currently preventing making this a reality,” said Fenstermacher.
For one, the current state of EHRs is a mixture of discrete data and unstructured texts so physicians still have to dictate notes and a transcriptionist puts those notes in a text form before going in a medical record, Fenstermacher stated. “A lot of information in those texts documents are imperative if we are going to do personalized medicine and to exchange a patient’s past history data in real time across a national exchange.”
Another gap is that clinicians cannot follow patients longitudinally easily. At Moffitt, for example, physicians can’t, in real time, track a patient’s medical history if he or she came in for an initial chemotherapy treatment and went back to his or her primary care physician for general care before returning a year later for a recurrence. Instead, paper copies are physically transferred by the patient or via fax.
Adding to the barriers includes the lack of a national health ID and the inconsistent use of data standards, stated Fenstermacher.
Fenstermacher called for creating a different type of data warehouse for comparative effectiveness research. “To obtain a better understanding of genomic and clinical data, we need to make sure we use differing standards in an interchangeable manner,” said Fenstermacher. “We need to build an infrastructure that looks at filling the information gaps by linking to personal health records while taking advantage of population based data sets--for example, from the Centers for Medicare & Medicaid Services and Surveillance Epidemiology and End Results linked to patient cohort data.
“We need to use today’s best practices to move forward and make innovations to create a network that will allow added value in the context of personalized medicine, CER as well as drive creations in health IT,” Fenstermacher concluded. “The proposed universal exchange language will only get us so far without other health IT innovations that solve the information gap and expand health data networks that are able to integrate heterogeneous data from multiple sources using centralized and federated data models.”