NATIONAL HARBOR, Md. -- At the eHealth Initiative's National Forum on Data and Analytics in Healthcare last week, stakeholders discussed the importance of data and analytics in implementing health reform, as well as the challenges associated with it.
Jennifer Covich Bordenick, CEO of the eHealth Initiative, said, "Our survey, CIOs and members kept telling us that they are concerned about analytics. They don't feel they have the tools necessary to meet the demands of accountable care and meaningful use." She noted that "93% of the CIOs believe it is very important, but 72% don't feel their organizations have what they need to meet the analytical needs."
By convening experts in health data and analytics, the forum aimed to highlight organizations that are leading the way and facilitate conversations around the need for improvement, she said.
Federal Government Touts Data and Analytics To Support Health Reform
Niall Brennan -- director of the Policy and Data Analysis Group at CMS -- told attendees that health data analytics is "absolutely central" to everything that has to do with health reform.
Brennan -- who stepped in for U.S. Chief Technology Officer Todd Park to give the afternoon's keynote speech -- highlighted the federal government's efforts "to make the data more helpful," while not compromising individuals' privacy.
He said that in the past CMS was "overly conservative" in terms of data release. However, in the last few years -- in large part because of the Affordable Care Act -- the agency has made great strides in liberating health data, he said.
"It might not look like it on the outside, but we are literally constantly pushing the envelope," Brennan said.
He cited the Blue Button Initiative and HealthData.gov as examples of the government's efforts to make privacy-protected health information available to the public.
He also noted that Section 10332 of the ACA authorizes the release of Medicare fee-for-service data to qualified entities if they agree to combine the CMS data with claims data from other sources to compile performance reports.
While Brennan touted the federal government's release of more health care data, he acknowledged, "You can have all the data in the world, but if you don't have the right analytics ... it's just a bunch of useless numbers."
Brennan said the federal government -- from CMS to HHS to the White House -- is "very, very committed" to data and analytics.
Survey Finds Industry Still Has Far To Go
At the forum, Jason Goldwater -- vice president of programs and research at eHI -- offered a sneak peek into the results of a survey eHI conducted with the College of Healthcare Information and Management Executives to get a picture of the types of data and analytics being used in health care.
The survey, which was conducted in July, focused on four areas:
- Types of data used;
- Types of analytic functions used;
- Types of functions needed; and
- Challenges to the use of data and analytics.
When asked what data their organizations actively exchange:
- 76.6% of respondents said lab results;
- 74.5% said demographics;
- 70.2% said discharge summaries;
- 46.8% said allergy information;
- 36.2% said continuity of care documents; and
- 36.2% said problem lists.
The survey found that most health care organizations are focusing their resources on retrospective analysis, with 58.3% citing that as the area in which they direct the majority of their analytical resources. According to the survey, 16.7% of respondents said their organizations direct the majority of their analytical resources toward real-time decision support, 13.9% said optimization and efficiency and 2.8% cited predictive analytics.
When asked what type of analytical functions their organizations primarily use:
- 87.5% of respondents said ad-hoc queries;
- 61.1% said data mining;
- 56.9% said data warehousing;
- 34.7% said exploratory data analysis;
- 30.6% said on-line analytical processing; and
- 23.6% said predictive modeling.
Goldwater said that despite health care organizations' interest in data and analytics, respondents cited several challenges, including:
- Lack of standardized data across systems;
- Lack of a system infrastructure to support analytics;
- Cost of analytical software;
- Concerns about privacy and security of the data; and
- Limited utility of the results to the organizations.
Goldwater said that eHI, CHIME and McKesson will host a webinar on Aug. 30 to discuss the survey results in more detail and that eHI will release an issue brief in the fall.
Speakers Highlight Challenges Associated With Health Data and Analytics
Several speakers offered real-life examples of the challenges cited by respondents to the eHI/CHIME survey.
Jason Williams -- vice president of business analytics at RelayHealth -- said that health care cost and quality reforms require inter-stakeholder transparency and that there needs to be "more emphasis" in that area. He also cited the demand for business and technology analysts and the need to find a balanced approach to privacy as areas for improvement.
Micky Tripathi -- president and CEO of the Massachusetts eHealth Collaborative and chair of eHI's Board of Directors -- said that variations in EHR systems can be problematic for analytics.
Brendan Mullen -- senior director of PINNACLE, the American College of Cardiology's outpatient registry -- noted that its system integration tool has to look at 41 locations in the NextGen EHR system to determine if a physician provided patient education on heart failure.
Tripathi said that as vendors are working to address the issues, new measures are coming down the pipeline.
For the Healthcare Information and Management Systems Society Conference in February, MAeHC compared the results of its certified Quality Data Center with the Office of the National Coordinator for Health IT-sponsored, open-source popHealth tool to evaluate meaningful use quality measures. The tools used the same exact data, but they did not produce the same results for any of the 44 measures, Tripathi said.
He explained that further investigation found several reasons for the discrepancies, including the definition of the continuity of care document, coding and mapping, and the interpretation of certain measures, such as age.
Tripathi said, "We have a ton of work to do," adding that the industry needs to keep pushing along.
Covich Bordenick told iHealthBeat, "It was clear by the end of that day that this was just the start of a conversation that is going to take years to explore," adding, "eHealth Initiative is going to help unpack this issue."