Everybody wants data. Big Data. And then they want big-ticket analytics software systems to glean insights that predict the future of patient care, operations, revenue cycle from that data, of course. With so much chatter about predictive and even prescriptive analytics, c-suite executives are taking notice. What's perhaps less understood, however, are the practices, processes and technologies often needed to support those analytics investments. 

The twin imperatives of master data management and data governance, in fact, can help tame troves of clinical information and hone the datasets into shapes and sizes best suited to those who need it. 

"Data governance efforts can be hard to get off the ground and to get funding for," said Kim Jackson, Vice President for Strategy, Products and Governance at Providence St. Joseph Health. 

"It's not always popular. Because it's essentially housekeeping." 

There have been clear benefits to implementing those systems at PSJH – breaking down silos of data ownership, better data stewardship practices across the organization in general – but at first Jackson had to work hard to convince C-suite decision-makers to pay for the programs, which don't typically have a well-defined or obvious return on investment.

When it comes to getting buy-in from execs who may be mystified about the need to for governance and MDM, the key, Jackson said, is to "really have an ROI focus. A lot of governance initiatives start out with, 'We have too many report requests so let's sit in a huge room with 30 people and prioritize reports.' But that doesn't really provide ROI. It just gets all these people in a room to talk and talk and talk with no action. I am someone who believes in having deliverables and ensuring that we make progress. I wanted an ROI based solution."

Jackson's years of experience at PSJH meant she knew all the senior executives and "could tell stories about how everyone has had to struggle with data because of data governance and quality issues," she said. "In each data domain – whether it be finance, whether it be clinical quality, whether it be marketing – for every senior leader, I could say, 'Remember when this happened, and we weren't successful because of data quality? Remember when this happened and it was because we lacked data governance?"

By highlighting those "specific stories and pain points," she was able to secure funding for the decidedly unsexy but entirely valuable jobs of master data management and governance.

"Be clear about what your ROI is," she said. "I focused on reducing duplicates, on data quality – and then attaching that to outcomes."

MDM came first: "We started with an enterprise patient master index, we have created patient ID across our entire enterprise" said Jackson. "We made huge gains in that. On average it cost about $100 to clean up a record, so I could point out, 'Here are the risks and problems that arise if you have a duplicate patient record: safety issues, wasted money, there is impact in lots of different areas."

The ROI? "We've been able to decrease our patient duplicates by 67 percent – the national average is 10 percent," she said. Starting the physician master was very similar. "When we first put all the records together there were 300,000 unique physician records and I know we only have about 25,000 providers. So that shows you how bad the duplicate rate was in the organization. We have reduced these duplicates over 90 percent."

That's just a couple of examples of dirty data that highlights the fact that more isn't always better. In fact, it rarely is. What matters is having the right data, easily accessible in the right place.

"The customer's' perception is that they need more and more data, and it's a never-ending thirst," Jackson said. "Sometimes I like to say that an analytics program can be a hoarder's paradise." 

With the combination of deactivation and consolidation, matching and merging, PSJH has managed to streamline its data resources in a real and meaningful way. 

"We continue to grow in creating hierarchies and crosswalks so we're not hard-coding reports all over the place and we're really reducing the amount of analytics requests we have by ensuring that our data is standard and more consistent," Jackson said. "That's the moral of my story."

Jackson will be speaking in the session, “Breaking Down Barriers with Master Data Management and Data Governance,” at 1 p.m. March 8 in the Venetian, Palazzo G.

Twitter: @MikeMiliardHITN
Email the writer: mike.miliard@himssmedia.com