‘This is scary.’

 October 11, 2013

Before you know it, the ICD-10 go-live date of October 2014 will be here, and for those 55 percent of physicians who have yet to begin implementation, it’s looking like a rocky road ahead.

Speaking at the MGMA 2013 annual conference in San Diego this week, Holt Anderson, executive director of the North Carolina Healthcare Information & Communications Alliance, shared his group’s experiences with an ICD-10 end-to-end testing pilot.

Initial results? They’re just plain “scary,” said Anderson, who explained that worker productivity decreased by 50 percent when attempting to identify the proper ICD-10 codes.

[See also: ICD-10 readiness lagging, says MGMA.]

And Anderson considers NCHICA more ahead of the game than the rest. The group includes more than 300 member organizations and established its own coding task force back in 2010.

The pilot, he said, all started to see where coders were at and to share experiences with others in hope of preparing more providers. “We know what 5010 should end up being paid,” said Anderson. “We send it, and see how they adjudicate it and how it comes back. Did it match or did it not? Did the clearinghouse in between do anything?”

For 20 dual-coded and peer reviewed scenarios, the wave one results were 55 percent accurate, with many scenarios falling well below that. This was not a number Anderson, NCHICA and the participating pilot members were expecting.

These were actual clinical cases together with the top coders. “We selected some of the best of the best coders in these organizations,” he said. Then, they did peer review and realized these coders weren’t even coding correctly in ICD-9, let alone ICD-10.

In wave two, the results were considerably better, averaging to a 63 percent accuracy rate, but “single, live births born in hospitals, delivered by Caesarian section were 100 percent coded wrong,” said Anderson.

So far, the task force has conducted more than 200 highly relevant dual coded and peer reviewed scenarios, and another 100 scenarios are still in the process.

Not only was the coding proving inaccurate, it was “taking about twice as long to code,” added Anderson. Coders were averaging two medical records per hour with ICD-10 compared to four or more under ICD-9. “Can you lose that kind of productivity?”

Larger healthcare organizations with the financial wherewithal to move aggressively forward with ICD-10 get it. They’re allocating the resources and getting the staff they need. But for smaller healthcare practices who don’t necessarily have the resources, it’s just worrisome. “We’re not sure how aware they are,” he added. “They’re concerned about meaningful use and getting their EHRs up to speed.”

And the end of the day, however, this is a major cash flow issue, he said, and providers need to “group up” and “get on it.”