Q&A: CMS Chief Data Officer Niall Brennan

'It’s just absolutely critical that we have all our data ducks in a row'
By Mike Miliard
10:06 AM

UPDATE: On Friday, June 19, Niall Brennan will be the morning keynote speaker at the Healthcare IT News Big Data and Healthcare Analytics Forum, which kicks of June 18 in New York. There, he'll talk about his role as CDO and offer insights into how CMS is deploying big data techniques to improve access to quality, lower-cost care.

This past November, the Centers for Medicare & Medicaid Services launched its new Office of Enterprise Data and Analytics – helmed by CMS’ first Chief Data Officer, Niall Brennan.

OEDA’s mission is to help CMS make smarter use of its huge – and fast-growing – troves of claims and clinical data, helping devise new approaches to analytics and finding ways to open access, putting that data to work driving better, less costly and more patient-centered care.

[See also: CMS launches chief data officer position]

Healthcare IT News spoke with Brennan about his first few months as CDO, and asked him his advice for those providers just beginning to tap into their own data to spur efficiency and improve care delivery.

Q: 2015 has been called the "Year of the Chief Data Officer." Why is this a position whose time has come?

A: I think the reason is that over the past few years, and this obviously goes beyond the emergence of the CDO in the healthcare sector or the government sector, it’s become clear to people that in order to fully harness and leverage the available data that a given enterprise has – and data is multiplying at an exponential rate – organizations have been most successful when they focus on data as a core function distinct from the IT component of stitching it all together and putting it in a data warehouse.

I certainly know the CMS and HHS view was that data was becoming more and more a core function of our day-to-day activities. That’s not to say it wasn’t important in the past; we always generated and collected a lot of data. But as we move from a relative passive purchaser of fee-for-service care, to value-based care with an emphasis on holding providers accountable to various performance thresholds and measuring quality much more aggressively, it’s just absolutely critical that we have all our data ducks in a row.

Q: Especially since HHS has recently put forth some pretty ambitious goals.

A: Exactly. And one of the central pillars of that announcement Secretary Burwell made a couple week ago is better access to health information.

Q: What have been your first orders of business in setting up the OEDA?

A: This new organization didn’t emerge out of thin air. There was a body of work that was already being conducted, and that body of work was then aligned under OEDA. We had a solid foundation upon which to build. The marching orders were to develop analytic centers of excellence internally, covering our various business lines – be that Medicare, Medicaid or the health insurance marketplace. We’re working aggressively in that area. You need to have your data stored and architected and accessible in a way that talented analysts can go in and have fun and test hypotheses and ultimately generate valuable information to guide the enterprise as it goes forward.

Another internal component is better data governance at the agency – just ensuring that one part of the agency knows what the other is doing, from a data perspective. Improving data standards, data quality, metadata and the like.

And then from an external facing perspective, we recognized that demand for our data is probably greater than it’s ever been. Obviously, a lot of the data is extremely sensitive, and we take great pains to protect patient privacy. But to figure out other ways of pushing the envelope in terms of making data available externally. Building on the success of things like the physician data release last April, where we released 9 million records covering 800,000 physicians; the hospital charge data release; our county-level Medicare geographic variation data, etc.

Q: Recognizing that it’s been early going so far, have there been any interesting or gratifying discoveries or surprises since you became CDO?

A: In many respects, it is too soon. But I think the gratifying thing is probably how positively the role and the position has been received, both internally at the agency and among external stakeholders. I think often, with the creation of a new office, there can be the danger of people rolling their eyes and saying, "Oh, look, it’s another chief." But people have been very positive – and again I think that’s in part due to the fact that we were somewhat of a known quantity, prior to the announcement and alignment, and we had a good record of delivering tangibles and adding value to other folks at the agency.

Q: Looking forward, what are some of the data-driven initiatives you’re most excited about at CMS?

A: We’re looking forward to incorporating some very important new data streams into our enterprise. We have Medicare Advantage encounter data – we’ve been collecting that for a little over two years, and we’re looking forward to finishing up validation of that data and integrating it into our day-to-day analyses. There’s a similar effort with Medicaid data: we have struggled somewhat in the past in obtaining timely and valid Medicaid data from states, so there is a new effort called T-MSIS – "Transformed MSIS," or Medicaid Statistical Information System – that will be launching soon to collect more and better data from states.

Obviously, as the health insurance marketplaces continue to stabilize and grow, there’s a lot of information that can be learned there.And then starting to tie data across multiple silos – better integrating quality data with our Medicare claims data and the like.

Q: Any challenges so far, either with regard to the data itself or the analytics tools?

A: It would be naive of me to say there aren’t challenges, but I don’t think there are too many challenges on the technology side as long as you deploy that technology in a smart and strategic way. I think one of the challenges – and this is certainly not unique to me, it probably applies to every CDO – is acquiring the necessary human resources with the right skills to leverage the technology. You can have all the IT in the world, but if you don’t have the right folks to optimize it, that can be challenging.

Q: We’re seeing more and more CDOs on the provider side. Do you think that’s a trend that will continue?

A: I think it will. We realized that we needed to begin to recognize the critical value of data and analytics to our core mission. I think it’s probably similar for many health systems. Again, we’re moving beyond a system where it’s, "Send us the claims and we’ll write you a check," to one where these organizations will have to effectively manage patient populations, interact with CMS and other payers in providing all kinds of quality and other performance data; will have to learn to interpret performance data that they’re receiving back from payers; integrating clinical data from the their electronic health records – you know the investments that have been made in these systems, and unlocking the data that lives in EHRs will be key for many organizations to attain success in this rapidly changing landscape.

[See also: Chief data officers come to healthcare]

 

Q: A lot of providers are just starting to dip their toes in the water with this stuff. As someone swimming in these waters for a while, any advice, for data wranglers on the provider side?

A: My single most important piece of advice would be to hire smart people. And my second most important piece of advice would be that you have to make your analyses tangible and relevant to the powers that be. Be careful – you could have the greatest analysis in the world that could put someone asleep in five minutes. You need to make sure it’s relevant to the challenges the organizations’ leaders are facing.

Want to get more stories like this one? Get daily news updates from Healthcare IT News.
Your subscription has been saved.
Something went wrong. Please try again.