Medical Informatics: Apps, not data warehouses, are wave of future

BOSTON--Beware of the perils of data warehouses, Jon D’Amore, MS, founder of Clinfometrics advised an audience at the Medical Informatics World Conference on April 8.

Providers are still a long way off from meeting Stage 2 Meaningful Use (MU) that requires the achievement of advanced clinical processes and better population health management through EHR information capture and exchange, said co-speaker Dean Sittig, PhD, professor at the school of biomedical informatics at the University of Texas Health Center.

“We call it incentives, but it’s really pressure,” said Sittig, speaking to federal penalties for poor performance. Medical data standards are slow to advance as fewer than 25 percent of hospitals exchange summaries and interoperability remains a daunting challenge, he said.

D’Amore said instead of large data warehouses, health apps that function within an EHR are a solution to better data integration and analytics and may hold the key to unlocking MU Stage 2.

Data warehouses are cumbersome, he said, as providers often encounter difficulties retrieving information. “What if you had to call every time you go to its website and ask, are you ready for me?“ Sittig said of typical exchange protocol, noting that less than 10 percent of providers get a positive ROI when using them.

“We need apps to solve specific problems, not warehouses to store data,” he said. For analytics to live up to its true definition, he said it must drive decisions and actions. Apps, not warehouses, offer that potential. For an app to be effective, it must include many sources of data to layer onto an EHR--such as medications, labs, orders, problems, vital signs and allergies.

“Any EHR can document care, but the apps will make it smart,” he said, noting that the Office of the National IT Coordinator is investing $15 million to develop healthcare apps.

Technically, he said continuity of care documents (CCD) can be generated from all MU certified EHRs and incorporated into an app, with the XML packet containing demographics, medications, allergies, lab results, problems and procedures. Such an approach would require both structure and standard terminologies, he said.

Both speakers also spoke about the need for prediction tools to prevent readmission, and how clinical documentation apps can alert providers to potential documentation gaps in processes in treating an array of conditions, including asthma, CAD, COPD, CHF, diabetes, osteoporosis, renal failure and stroke.

In one study using the apps, D’Amore said, “30 to 50 percent of alerts were accepted by ambulatory physicians, dramatically increasing the documentation rate.”