After teasing a "Precision Medicine Initiative" in his State of the Union address, US President Obama today released the details of his $US215 million plan: A massive database containing the genetic profiles and health records of at least a million volunteers. Genetics sequencing! Big data! These all sound like good things! But there is potential problem: Electronic heath records are a goddamn mess.
The big winner under Obama's plan is the US National Institutes of Health, which gets $US200 million to develop the million-large volunteer cohort and carry out cancer research. (To read more about the research, head on over to our sister site io9.) A tiny slice of the initiative's budget — $US5 million — is also going to the Office of the National Coordinator for Health Information Technology. According to the White House's fact sheet, the ONC is going to "support the development of interoperability standards and requirements that address privacy and enable secure exchange of data across systems."
That sounds deathly boring, I know, but the ONC is building the scaffolding that makes this Precision Medicine Initiative possible. If it fails, the whole thing is not going to get off the ground. Here's why.
The NIH is not going to recruit a million volunteers anew; instead, it's going to pull data from a pool of over 200 existing cohort studies that could range from the Framingham Heart Study in Massachusetts to a database at Kaiser-Permanente in California. Each of these studies has been collecting data in their own way in their own systems. Smooshing all those individual databases together into one centralized one will be a gargantuan task.
"They're going to have severe problems because the federal government refused to demand data standards," says Ross Koppel, a sociology professor at the University of Pennsylvania who specialises in healthcare IT. He is referring to the lack of interoperability between EHR systems built by different companies. For example, says Koppel, a simple question about smoking can be asked in many different ways: Do you smoke? Have you ever smoked? How many times a day do you smoke? When did you quite smoking? How do you combine answers to all those questions into one coherent database?
The ONC currently has a 10-year road map for interoperability, which Koppel calls "nine and a half years too late." The $US5 million is welcome boost to the ONC's $US75 million budget, it's a tiny droplet in the ocean for the $US3 trillion that Americans spend on EHR every year.
But there are reasons to be optimistic because, honestly, we do all want this to work out. The NIH's eMERGE Network successfully combined medical records from nine different healthcare centres. The Precision Medicine Initiative has the real potential to revolutionise medical research, so it'd be a grand shame for it to be hobbled by IT problems.
Top image: Christian Delbert