The Maine HealthInfoNet is aggregating and analyzing health information exchange data at the population level, with the aim of finding trends and specific figures that currently evade most tools of epidemiology.
In a pilot project funded by the Centers for Disease Control and Prevention, HealthInfoNet, Maine’s statewide HIE, is collecting and assembling data for the Maine CDC, using the open source software popHealth. The project focuses on 13 Meaningful Use clinical quality measures using the ABCDS — aspirin therapy, blood pressure, cholesterol and diabetes control and smoking cessation.
It should let public health researchers find out, for instance, what percentage of Mainers with diabetes have sugar levels under control and what percentage of hypertension patients had their blood pressure checked during their last medical visit.
[Related: CDC to use Direct protocol for health safety network.]
These and other public health measurements, at least of large populations, have eluded researchers for a while, said Stephen Sears, MD, Maine’s state epidemiologist.
“If you want to know how many diabetics within a database there are within a certain age group,” Sear said, “that’s almost impossible to get right now unless you have a data set like the Maine HealthInfoNet registry.” Plus, you need the technology to sift through it all while staying HIPAA compliant.
Sears is cautiously optimistic that they’ll be able to find all of what they’re looking for in the various clinical areas. “What I’ve seen is that it suggests that for certain parameters it looks like its going to be able to work.”
The data basically runs from HealthInfoNet and its vendor, Agilex, to popHealth and then to the Maine CDC. The U.S. CDC’s role is mostly funding and support, through its program Demonstrating the Preventative Care Value of Health Information Exchanges.
The project and a lot of data collection started a little less than a year ago. Now HealthInfoNet and the Maine CDC are essentially testing their capabilities.
[See also: A look at how Maine’s HealthInfoNet is turning grant money into actionable outcomes data.]
“For every person, each month we’re producing different measures for people based on conditions,” said HealthInfoNet CEO Devore Culver. They’re able to make comparisons like how many diabetics who’ve had a hemoglobin lab test scored under 9, an indication of diabetes control.
“The state CDC expends a significant amount of energy and effort trying to gather and compile data that allows them to draw conclusion about trends of health in Maine,” Culver said. “Up until now, it’s been a fairly manual-intense process and the data is not always clear.”
Culver noted that the project is first of its kind and could be a boon to the Maine health department and CDC.
“It’s really a first foray into whether you can repurpose information, not violate patients privacy or expose providers, and use it for something of value and see how health is progressing in the state,” Culver said. “This is a very low cost, with a lot of value.”
If Maine can prove the analysis works, the goal is to eventually take the model to other states, said Dr. Taha Kass-Hout, Director of CDC’s Division of Informatics Solutions and Operations.
[Feature: A new age of biosurveillance is upon us.]
“We’re building a way for state health departments to use the data that’s circulating around their state health information systems,” Kass-Hout said. “The whole goal here for us is to be able to create shared services and platform for local and state health department to use this data.”
First, though, HIE organizations need to be able to meet certain technology and policy criteria that lets them navigate potentially rough waters, he said. HIPAA compliance is a major challenge, as is maintaining providers’ privacy.
“Maine Health InfoNet has the right governance, the right policies and it’s independent, non-for-profit — free of much political or industry influence,” Kass-Hout said.
The pilot project runs until the end of the year. Kass-Hout wouldn’t say if the CDC will renew funding for another year.