10 Critical Myths and Realities of Master Data Management
Prevalent myths surrounding MDM alongside an explanation of the realities.
This week, I had what-oddly enough-turned out to be a major disagreement with a friend in health care. I wanted to know why I had to copy my insurance information onto a form when they were then going to turn around and copy the card.
"Well, we do that so we can verify the information," she said.
"How does that verify the information?" I asked.
"But that would just verify whether I copied it correctly. Or memorized from the card correctly. I think this form just hasn't been redesigned since the invention of copiers. I mean, really-do people sometimes write down a completely different insurance on the form and then give you a card that says differently and you're like, 'Aha! You said you have Anthem, but you actually have Humana?' What's the point? Wouldn't you trust the card over their memory anyway?"
I was told I didn't understand because I'm not in health care, but she is in health care and, trust her, the form matters.
She's right-I don't understand. And among the things I don't understand is how this duplication of effort is a legitimate health care issue OR a legitimate effort to control data quality. From where I stand, it looks like yet <em>another</em> example of the health care industry's antiquated and error-prone approach to (mis)handling data.
It's not as if I would know if my insurance card is wrong, since the card is the source of my information. Oh, sure, if I'm smart, I double-checked the information on the card against my policy letters. But, I'm guessing most people don't do that, so basically, most of us are relying on that card to be right. If we've memorized our insurance information, we used the card to do so.
Therefore, asking me to copy that information-or write it down from memory-just invites errors into the data. Oh sure, if I don't have the card, you've got to ask me to write it down, but otherwise, why bother? An even more ridiculous, but increasingly common scenario: They ask you to fill out your insurance information on a form and then scan your card directly into a computer.
But, honestly, for all my ranting about IT/business alignment, I can see why IT gives up. If a health care employee can't understand why using a copy of the card is a more reliable source than a handwritten form ... well, why bother?
Here's a good reason why somebody needs to bother: $314 billion. That's how much middleware expert Hollis Tibbetts of Artemis Ventures estimates fraud or inaccurate billing costs the U.S. health care industry each year.
I'll grant you, he arrives at that number through some mathematical gymnastics using the U.S. GDP and a remark by the U.S. Attorney General's office that 14 percent of health care dollars are wasted in fraud or inaccurate billing. The figure gets worse if you add in other ways data problems can raise the cost of health care.
But if IT doesn't deal with the bad data problem, then the cost gets pushed downstream to the 'business,' where the business costs are geometrically higher. The model is not that different from that of 'healthy software,' where it costs $1 to uncover a defect during developer/unit testing, but $100 to fix that defect if the software is released to the end-users.
He outlines three best practices for improving data quality within IT:
This isn't just a health care integration issue, of course. Lots of organization would do well to introduce data quality into these three key processes.
I appreciate that Tibbetts specifically mentions deduplication, because I've often wondered why deduplication isn't discussed more often as a way to improve data quality during integration. I've been told most data integration platforms wrap up deduplication as part of the integration process, but if you're a hand-coding shop, check to make sure it's being incorporated into the process.
And while we're on the topic of deduplicating, if you're in health care IT, would you do me a solid? Introduce your business users to a card scanner and "deduplicate" those outdated forms, because what is this, the 70's?