Q&A: Good Governance Lends Structure to Tech Changes in Healthcare
Healthcare provider Catholic Health Initiatives is using governance and enterprise intelligence to help meet the challenge of rapid changes in healthcare.
- By Linda L. Briggs
- May 20, 2014
Pushed by federal legislation and the evolution of healthcare itself, healthcare organizations are moving rapidly to adopt technology to cope with electronic patient records, meet federal healthcare mandates, and define reasonable use of patient data, among many changes. At Catholic Health Initiatives, a large national healthcare provider based in the Denver suburb of Englewood, a clear focus on governance is helping to protect consumer data, organize and manage technology holdings, and move the organization forward.
In this interview, Deborah O'Dell, senior director of enterprise intelligence operations at the organization, discusses why both data governance and overall governance are important to CHI, and how governance initiatives are being implemented and managed. CHI reaches across 18 states with healthcare systems and services that range from large hospitals to rural facilities as well as doctors' offices and nursing colleges. The large non-profit is a $19 billion company; if it were publically held, it would be a Fortune 200 company.
BI This Week: Data governance has evolving meanings. What does it mean in particular at Catholic Health Initiatives?
Deborah O'Dell: Actually, we perform various types of governance; data governance is really just one type of governance work we do. We perform governance over all information management. For us, information management is a strategic initiative; that's how we use information and analytics as an organization, and our leadership helps to set the direction.
To us, data governance alone means stewarding individual data elements. For example, if we're going to develop an analytic around length of stay for our hospital patients, then we need some way to look at that. To develop that analytic, we might meet and choose a standard definition of length of stay. However, if we start collecting data for length of stay and we find out that it's not what we expected it to be -- say the number is much higher than we expected -- then we need to dig into the actual data elements. We may find that we aren't capturing the admission date very well, which is affecting length of stay, so we'll put data governance around that specific element to try and increase the accuracy, completeness, and validity of that data element.
We also do data governance around our analytic portfolio. In that case, we send data out to our vendor partners; data then comes back to us from those vendors. We govern that data as well.
When you spoke at the TDWI World Conference in Las Vegas earlier this year, you said, "For us [at CHI], it's all about governance." Moving beyond just data governance, can you talk about overall governance at CHI?
Yes, it really is all about governance for us. As I said, data governance is a narrow piece of it. Our governance efforts are based around a broad strategic initiative at CHI that says we value information as an asset. Someone said recently in a meeting here at CHI that data is the new capital. We have to own that data and steward that data, just as we do the dollars for capital improvements. That's what is going to get us to the next level.
In terms of overall governance, we have strategic governance around what we refer to as information management -- how we manage information as an asset in the organization. Part of that becomes, at one level, data governance, and at another level, governance of our analytic portfolio vendors. We also have project governance -- when we have to build a dashboard or bring in new data, a situation that calls for another level of governance. Then we need the right people to sit around the table with us. The concept is more about governance in general -- and the level of engagement that you need from stakeholders.
When you say, "data is the new capital," do you think that is especially true in healthcare right now?
I do. In healthcare, new initiatives at the federal level, along with the evolution of healthcare itself, are driving us toward electronic health records. People sometimes say that the data in healthcare is not as mature as in other industries because we haven't had integrated systems that have been collecting data electronically for 20 or 30 years. It's true -- data in an electronic format is new for us.
Who participates in your governance efforts? What is the structure, and what kinds of issues do they consider?
We have different groups throughout the organization. At the highest level, our president and chief operating officer, as well as our chief medical officer and senior VP of finance, all participate in what we refer to as the Information Oversight Committee, or IOC. They meet about once a quarter and provide high-level oversight and direction. Under that, the Information Management Council is made up of VP-level folks who represent both national functions such as finance and human resources and also our regions. Underneath that, we have a [new] director-level group that we're calling our Prioritization Workgroup -- that group will help us look at the strategic initiatives from the oversight committee, break them down into projects, and prioritize those projects.
For example, at higher levels of the organization -- at the VP levels -- we may decide that we need an initiative around length of stay, but actually developing that in the data warehouse requires a whole new level of detail. Do we have the data in the warehouse? Is it organized in the right way? Are we going to provide a dashboard? Are we going to provide self-service access? That can become four or five different work efforts. Some of it may connect to other efforts and projects, so that's what the Prioritization Workgroup can help us with.
We need the business people engaged in that conversation and talking to each other. What we have discovered is that our group, Enterprise Intelligence Operations, is often the center of all of these conversations about initiatives and work. That's because all of the groups need data and information. The business units aren't always connected at the data analytics level, so we're more or less the hub for it all. We might talk to finance, and then we talk to [the] strategy [group] -- and discover that they need to be talking to each other.
Governance helps us get all those people around the table together and having conversations with each other. We convene people and say, "Hey, you seem to be talking about similar things. How can we make this work together?"
Can you talk about your bottom-up rather than top-down approach to governance?
The bottom-up approach really comes into play around data governance. In fact, when we look at all the governance across [the organization], I think we have a good balance. We have the right high-level strategic direction and we have directors who help us to lay out the details. As we move down into details such as data governance, that means having the right-level people in the conversation with us.
For example -- and this is a real example -- we might think that we can do a better job of collecting e-mail addresses. If we want to keep people healthy, we need to be able to communicate with them. If we don't have their actual phone numbers or e-mail addresses, or know their preferred language, that's going to be really difficult. However, when we look at how we capture e-mail addresses, talking to the vice presidents isn't really effective. We need to interface with the people who talk to the patient right at the doctor's office. When someone goes to the doctor's office and signs in, how are we collecting e-mail addresses?
In an organization as large as we are, we need to find those people who have those interactions and make them understand the importance of this effort. We need to help them understand that because we really do care about helping keep people healthy, having an e-mail address is really just one more way to do that. We can't push that down -- we need to have that conversation with the people at the front desk. We need to drive it from the bottom up and help the right people understand the value.
Within the governance discussion, can you talk about deciding to outsource certain analytics? How is governance maintained in that situation?
We believe that the data warehouse will not be everything to everyone. There are vendors out there with expertise in certain categories, and we need to leverage that and partner with them. This started two or three years ago, as we were building out the data warehouse. We had categories of analytics that we knew we were going to need in order to support this new model of healthcare. We had categories around care management and other areas, such as the supply chain. We mapped in vendors that we knew would fit in those categories. Today, our Information Management Council really oversees the analytic portfolio, which includes 23 vendors we work with, most of which receive data from the warehouse. If anybody at this point needs a new vendor to work with, they need to go through the Information Management Council to explain why they want a new vendor.
We also have an analytics management portfolio, to make sure that each new vendor goes through that governance initiation. Part of managing the portfolio is managing vendors, writing service-level agreements, and spelling out very clear deliverables and penalties. We do that through the contracting phase, then we operationalize it, so there's a lifecycle for how we manage the portfolio, and the Information Management Council has oversight of that.
Because healthcare is a newly evolving area, there are a lot of vendors out there offering services. CHI is a very large organization, so our people are constantly being contacted with solutions. Sometimes those solutions are extremely valuable to us, because it means that we don't have to build them internally, but sometimes it's technology that hasn't been proven yet. In that case, we can help them look under the covers and see what value it may or may not provide or we might find that what someone needs is within the product suite from a vendor we already have. That way, we can [negotiate and] get a better value proposition.
How does your analytics portfolio drive and enhance the value of the data warehouse?
The warehouse is only as good the quality of the data. The data is only good if it's used. If we just load the data every day, it doesn't do anything. When people want to use an analytic vendor, and we send data from the warehouse to that vendor, and it improves the quality of the data. First, we're using it. Whether we're sending it to the vendor every day or every month, it's being used -- it's a tool that the business wants and needs.
Also, with many of the solutions in our analytic portfolio, we require that the vendor send us data quality reports in return. Different vendors can validate for different things -- does the gender match the procedure? Does the charge make sense against the total account charges? They do the additional validations and send us reports. We then distribute those to our data stewards within our organization, to see why those validations are coming out that way. That's valuable feedback -- how does my data look within the data warehouse itself?
Can you talk about how you've put such a strong governance initiative in place at CHI?
One thing that's very valuable for us is that we don't sit within the IT organization. We sit within Clinical Services and report up to the chief medical officer. Having that alignment with the business connects us much more closely with the business people and helps us get that governance in place. That has helped us. What we do is fairly technical, and we need the business to say what they need and how they need it.
Our job is to make sure that the business people are talking to each other so they can be making the most effective decisions. That structure and those conversations have proven their value as we've grown over the years.
What we're doing -- integrating data and providing complex analytics -- isn't easy. Part of it isn't just getting direction from the business. We also need to explain to everyone the complexities of what people are asking for, so everyone can understand why certain decisions are made and why it sometimes takes longer than they would prefer.