BOSTON – Health care CIOs might question why they're implementing a massive wave of technology to enable meaningful use criteria, ICD-10, new payment models and two new diagnostic languages for payer claims, and SNOMED for standardizing EHR problem lists. Life sciences experts believe the answer lies in clinical data analytics.
Speakers at this year's Bio-IT World conference -- most of them representing pharmaceutical and biotech companies -- offered a glimpse into where meaningful use is taking us. All this health IT, they said, comes with massive potential for curing disease, creating courses of personalized medicine for the individual patient and mining EHR data to create much larger pools of drug knowledge than tightly regulated, small-sample clinical studies can accomplish at present.
While they're waiting for health care providers, payers and HIEs to get their acts together in the latter stages of meaningful use adoption in order to collect clinical information worthy to be called Big Data, companies like Merck, AstraZeneca and Pfizer have launched small pilots and partnerships that show its potential. One longstanding project, which predates even personal computers, involves cancer patients at Moffitt Cancer Center in Florida sharing their medical records and biopsy tissue samples with Merck to enable the pharmaceutical giant to create deep research databases for drug development.
The project, said Merck Executive Director of R&D IT Jason Johnson, has evolved in two decades-plus to the point where patients willingly sign over access to their EHR files to Merck. About 80% of patients give consent to the program upon admission to Moffitt, he said, thanks to a promotional campaign that included endorsements from celebrities and even former Florida Gov. Jeb Bush. The endorsers, he said, push the value of the program as benefiting future patients, not necessarily those signing away their records.
"We have about 20,000 samples in the system…with the medical records," Johnson said, adding that other hospitals participate in the research program -- enabling researchers to follow patient progress once they've completed their treatment at Moffitt. "I think it's one of the better models of consent. It's nice we can go back to the patients and go back [and follow up with patients], it's one of the great assets of the program."
Genomics + HIT = Clinical decision support
Personalized medicine is what the health care system will do with all this Big Data generated by current health IT initiatives, the speakers agreed. But how will physicians use it in everyday treatment, considering the overwhelming volume of data and clinical data analytics that can be done with it? The answer is clinical decision support, according to Zhaohui Cai, biomedical informatics director for AstraZeneca.
Physicians need HIT as an assistive technology that gives them more bandwidth to pay attention to the patient in the short time they're given for face-to-face encounters, Cai said; porting genetic data into decision-support tools for tailoring personalized care plans is the best way to deliver that. These decision-support tools will help physicians determine which of the rapidly expanding (and often expensive) choices of genetic test are the most relevant for the patient in front of them, given the individual data entered into the EHR. And he thinks such tools will also be used by payers, too, to validate those clinical choices.
Cai also gave a validation for both ICD and SNOMED adoption in health care. ICD, a system of diagnostic terminology that supports billing systems, yields some research results. The ability to predict much more granular results of a given patient care path, however, will come with the diagnostic codes in EHR problem lists -- which will be driven by SNOMED.
"You need information from the [EHR]," Cai said, to create truly risk-adjusted, outcomes-based personalized medicine plans.
Clinical data analytics could lower costs
Michael Cantor, a practicing physician, NYU medical professor and Pfizer senior director of biomedical informatics services, took it one step further. He believes genome data will be the key to "bending the cost curve" for U.S. health care, unlocking higher quality health care -- individualized for patients -- at lower costs.
With meaningful use in its earliest rollout phases now, he said, health care providers are focusing on low-hanging fruit such as controlling diabetes and hypertension. That will expand greatly through later stages, as the health care system harnesses bigger, better data sets.
In an earlier keynote address, however, data storage expert Chris Dagdigian, co-founder of life sciences IT consulting firm BioTeam Inc., said that the expansion of data in life sciences and health care is far outstripping hardware vendors' ability to innovate its technology and serve customers' needs.
"We're at an inflection point where it's getting cheaper and cheaper and cheaper and easier and easier and easier to acquire ridiculous amounts of data…but we're not able to buy storage," Dagdigian said. "Storage is not getting bigger at the rate we're expanding our storage hunger. That is absolutely going to cause catastrophic problems for certain organizations this year."
The yin and yang of decision support and health care data analytics
CIOs make data analytics a top 2012 priority
The success of ACO model hinges on clinical data analytics