Steven C. Barlow is CIO and co-founder of Healthcare Quality Catalyst of Salt Lake City, UT.
Tell me about yourself and the company.
Healthcare Quality Catalyst was established three years ago with the mission of helping organizations accelerate the process of outcomes improvement using information and measurement to support that process.
I worked at Intermountain Healthcare for about 18 years. The last 10 or so years there, I was the director of the enterprise data warehouse team. I had the great opportunity of working with some clinical visionaries who really knew how to apply proven quality improvement principles in healthcare and lead the information in evolution of analytics to support that process.
After spending that time with Intermountain Healthcare, my business partner and co-founder Tom Burton and I launched and started Healthcare Quality Catalyst to bring those same kinds of principles and products to organizations across the country.
Data-driven quality improvement solutions usually involve some combination of technology and consulting services. How is your approach different from what your competitors do?
We are more of the technology and products company, with services to support the installation and configuration of those tools. We bring a set of tools.
Let’s say an organization just deployed or has recently deployed an EMR system such as Epic and now needs a rational solution to get access analytically to those data. We have a data warehouse core starter set that we have built for EMRs like Epic that can get a client up and running on a data warehouse core in a matter of a few short months.
The services we provide involve orienting the client organization around proven guiding principles in data warehousing. Data warehousing is one of those disciplines that has a few leading schools of thought that haven’t necessarily been proven all that effective in healthcare. We bring some proven and rational incremental approaches to healthcare data warehousing in this domain.
The products are the data warehouse and then a few other tools, such as a Wiki-based metadata repository; a centralized data capture tool that allows analysts and clinicians and researchers to, in short term, capture information that may not be readily available in the EMR; and a tool called the Key Process Analysis Tool that we’ve developed which helps organizations prioritize and identify where they have the greatest opportunity for quality improvement, which typically is demonstrated through the amount of variation in a given domain. For example, why do we have an average length of stay for a hip replacement DRG with the same severity-adjusted population of x days in this facility and x + 5 in this facility? We have a tool that helps bring visibility into those opportunities for improvement.
Data warehousing seems to be something that hospitals think should be easy and inexpensive and they get scared away when they find out that it’s not necessarily either. Do you find that that’s an issue in convincing people that your solution will work for them?
Yes. We do typically see clients who are a little anxious about the size of these projects.
Our competitors take an approach where a lot of those lengthy project timelines really are required. The products that we bring to the table really accelerate the time from project initiation to usable information in an integrated data warehouse. Now we don’t prescribe or pretend to think that we’re going to have, in a few short months, a completely robust data warehouse that includes data from every possible disparate system within the organization, but we get them a long way down the path.
We also teach them some very tried and true pragmatic principles, both design and process principles, and get them well on their way. In a few short months, three to four months, we can have a client up and running and knowledge workers in the organization beginning to actually discover knowledge and do analytics.
Many companies have offered data warehouses and business intelligence dashboards geared toward quality and cost. Do you think hospitals have seen the results they expected from those? Why do you think that is or isn’t the case?
You know, I don’t think they have. There have been a lot of reports circulated over the last decade where across verticals — not just in healthcare – over 50% of projects are deemed as failures. Based on our experience, I think I would highlight a few reasons we see those still failing.
One, it’s a technology project looking for a business sponsor. The “If we build it, they will come” approach will never be deemed successful by the business.
Two, I think there’s a lot confusion about, “What architecture should we use for this data warehouse?” There’s a lot of fits and starts. A couple of the predominant approaches have been proven to be very effective in industries such as finance and manufacturing and retail where the data are a bit simpler and much less complex than they are in healthcare. We prescribe to one of those approaches, as opposed to, “Let’s think pragmatically and maybe adjust one of those approaches to fit the need in healthcare.”
The final reason I would highlight based on our experience is the motives behind building a data warehouse often are misguided. For example, we see often organizations either acquiring analytical products or deploying measurement systems for the purposes of identifying where there are outliers and reining those outliers in. It’s used in a more punitive way, as opposed to a learning way where, “Let’s identify in this organization side by side, technologists with clinician, where we have some great things being done and let’s learn why it’s being done consistently in this area and permeate those changes across the system.” It’s punitive versus learning motivation.
With that in mind and knowing as you said an organization needs a business sponsor as well as the technology, how can you tell if a prospect is really going to be motivated to take the actions that the data are going to indicate?
That’s a great question. I think when we go into a client organization we really like to visit with both sides of that fence. We like to visit with the technical leadership as well as the operations and clinical leadership. We can quickly get a feel within the organization how motivated and engaged they will be with a business in a clinical sponsorship driving the technology and how open the technology folks will be in that kind of a relationship. It’s quite easy to tell in a few short visits.
Can you give me a few of the specific outcomes that customers have seen as a result?
We’ve seen some of our clients, as they deploy the technology very quickly, they also begin to deploy the methodology that we would prescribe. They begin to have opportunities open up to them — the provider organizations — to speak with payer organizations and say, “Hey listen, we’re working on these quality improvement initiatives and we see opportunities to share in the savings that will result from these initiatives.” There are some exciting discussions and relationships beginning to form between payers and providers where they co-fund these initiatives and begin to share in the savings.
There are also real clinical improvement measurable results that we have seen with some of our clients. As an example, one of our clients set some goals to reduce the elective induction labor before 39 weeks. We know based on research that if labor is induced before 39 weeks gestation, the risk of NICU days goes up and the average duration of labor is increased. The goals are to reduce the percentage of the time that labor is induced electively before 39 weeks gestation. One of our clients went in less than a year from a 15% elective induction rate down to a 2% elective induction rate. That’s just one example of some interesting improvement initiatives that we see happen in our clients.
You mentioned Epic. Kaiser is doing some pretty amazing things with their information from HealthConnect, which is Epic. What kind of work are you doing with Epic customers and what’s the benefit to them beyond what Epic offers out of the box?
Epic did a great platform, a great EMR system. What we can provide is really helping clients now who have deployed Epic. We’ll leverage that rich resource of information and very quickly, in a matter of a few short months, they can have those data available in a data warehouse — a very scalable, usable data warehouse platform into which they can also integrate other disparate data sources.
We have a product roadmap where we’re going to bring in some interesting visualization tools and other ancillary tools to support that process around clinical domains. As a clinical diabetes team or cardiovascular team gets together and identifies their opportunities for improvement, we’ll have the data and the visualization tools to support those efforts based on evidence-based work done inside and outside the client’s area.
The company has a pretty large and well-credentialed management team. What’s the strategy going forward?
Our strategy going forward is to increase the number of connectors, if you will, to the various EMR systems and to continue developing and enhancing our current ancillary toolset. As well as creating greater knowledge into these targeted data sets around conditions that we see after doing some interesting Pareto analysis from client to client. We patterns bubbling up the top 10 or 15 clinical conditions and we’re going to fill out and make more robust those information assets that we’ll make available to our clients around each of those conditions.
How do you see the role of data-driven quality and cost initiatives changing with healthcare reform?
There are so many pressures from all angles that impose on provider organizations from the regulatory perspective, from a payer perspective, that I see the appropriate application of measurable outcomes improvement initiatives using a rich information repository is going to be an absolute fulcrum to make all of this possible.
I think we’re going to see – and are seeing — more and more healthcare organizations, from both from internal and external pressure, forced into paying much more attention to data coming out of these systems that we spent so much time getting data into over the recent years.
Do you have any concluding thoughts?
Healthcare is a very dynamic industry. We feel that Healthcare Quality Catalyst is positioned very well based on our rich experience and heritage and set of products that we bring to the market. We are just very excited to help healthcare organizations benefit from our experience and our toolset to accelerate their time to improving clinical outcomes for the patients they serve. It’s a great opportunity, and we certainly feel a responsibility to do our part to help in this solution to the healthcare problems that we face in this country.