The Changing Role of Healthcare Data Analysts—How Our Most Successful Clients Are Embracing Healthcare Transformation (Executive Report)
system they are being asked to rely on will not scale. This concern has historically been justified and validated, because traditional EDWs have been built using dimensional or enterprise architectures that present significant challenges in a healthcare environment. Here is a brief overview of why these architectures have not scaled well in healthcare:
- Enterprise model: In this approach, the goal is to model the perfect database from the outset—determining in advance everything the organization would like to be able to analyze to improve outcomes, safety and patient satisfaction. This is the right approach if the organization is building a new system in a vacuum from the ground up. But in the reality of healthcare, organizations are not building a net-new system when they implement an EDW. They are building a secondary system that receives data from systems already deployed. This model tends to disregard the realities of the data a healthcare organization actually has available. Furthermore, the model binds data to rules and vocabularies very early, and once data is bound, it becomes very difficult and time-consuming to make changes. In healthcare, business rules, use cases and vocabularies change rapidly. An effective EDW must be flexible to allow change.
- The dimensional (or independent data mart) approach to data warehouse design is an approach where organizations start small, building individual data marts as they need them. If the organization wants to analyze revenue cycle or oncology, they build a separate data mart for each, just bringing in data from the handful of source systems that apply to that area. This model has three major drawbacks. First, with all of these isolated data marts in place, the organization doesn’t have an atomic-level data warehouse from which to build additional data marts in the future. Typically, data marts do not contain data at the lowest level of granularity, so it becomes impossible to dig deeper to discover the root cause of data trends. Second, this model bombards source systems repeatedly and unnecessarily, and it requires the IT team to build redundant feeds from each source system to feed these data marts. Finally, like the enterprise model, this approach binds data quite early in the process. As data is brought into each independent data mart, it is mapped into the predefined data model- inhibiting the adaptability of the analytics solution.
Fortunately, an EDW architecture has now been developed for healthcare that avoids these pitfalls and allows the system to scale easily. This model—called a late-binding EDW—is an adaptive, pragmatic approach designed to handle the rapidly changing business rules and vocabularies that characterize the healthcare environment (Figure 9). This architecture takes data in its atomic form from source systems and brings it into source marts within the EDW. In the source marts, data is only bound to core data elements that are fundamental to any analytic use case, such as patient and provider identifiers.
Data is only bound further when a specific business driver or use case calls for it. Specifically, more volatile rules and vocabularies are bound as late as possible. This late-binding approach enables not only scalability but also the pragmatic, incremental development of an analytics system. Job security concerns As the CHIME survey introduced at the beginning of this paper shows, healthcare analytics is on the rise, and executives see data analysts as playing a valuable role in data-driven healthcare transformation. Removing the report queue from their duties will not put data analysts out of work. Instead, analysts have a tremendous opportunity to move from gathering, collecting and provisioning data to being part of multidisciplinary teams and applying their skills to improving performance outcomes. Increased workload because clinicians have more data The potential for an increased workload may seem daunting. Fortunately, the combination of a late-binding EDW and easy-to-use visualizations will take a lot of pressure off of healthcare data analysts. These technologies enable self-serve analytics—clinicians can look at the data themselves, including drilling down into the data and filtering information. Therefore, instead of demanding numerous reports from analysts, these clinicians can simply turn to a single analytics application to find the answers they need. This will allow analysts to work on more interesting and sophisticated analytical needs.
MAKING CULTURAL TRANSFORMATION HAPPEN
Tools and technologies alone cannot single handedly transform the role of the healthcare data analyst. Real cultural transformation is required, but cultural change and organizational transformation are never easy. So how does an organization make positive change happen? The following examples—one from the healthcare sector and the other from a Fortune 10 IT company—provide advice and insight about successful change management in analytics.
Changing Care Delivery in a Large Health System
The first example is shared by Dr. John Haughom, currently a senior advisor at Health Catalyst. Dr. Haughom was senior vice president of safety and quality and, later, CIO for a health system that spanned three states in the Northwest. His job was to support 23,000 physicians and 11,000 employees. Dr. Haughom led a 400-person department, 70 percent of whom were IT. During his tenure, the health system began to implement analytics technology to drive better quality. Before implementing the technology, his group was producing tens of thousands of reports. A large percentage of these reports went into binders that nobody looked at—not an effective use of his team’s resources nor of valuable data. To make the new analytics technology and strategy successful, Dr. Haughom had to change his organization’s culture so that reports, and the analysts who understood them, were integral parts of teams affecting patient care on the frontlines. Every member of the care improvement team—including the healthcare data analysts—needed to share a common goal focused on improving the quality, safety, efficiency and cost of care being delivered to patients. The improvements needed to be scalable and sustainable. This necessitated an approach and a structure that could assure these goals. To accomplish this, he implemented a three-system approach throughout the enterprise. The three systems approach involves implementing:
- An analytics platform, which includes the technology and the expertise to gather data, make sense of it and standardize measurements
- Evidence-based content to standardize decision-making at key points in the care delivery
- An organizational structure for implementing change, including multidisciplinary frontline teams and a governance structure
Dr. Haughom shares the following insight about his experience implementing this approach: “It can be a hard slog introducing that kind of change. It certainly was for us until our analysts crystallized their role. The light bulb clicked on for our analysts as they started to see improvement projects succeed because of the support they were offering: direct correlation between the data they provided and clinicians saving and improving lives. Clinicians couldn’t do this without them, and clinicians began to recognize and acknowledge that reality. When the teams started to gel in that way, it was very exciting.” There are several change theories and models for leading change. One model that Dr. Haughom has found particularly helpful is John Kotter’s eight- stage process for transformational change. Kotter’s process comprises the following stages: 1) establishing a sense of urgency, 2) building a guiding coalition, 3) developing a vision and strategy, 4) engaging stakeholders, 5) enabling action by removing barriers, 6) generating short term wins, 7) consolidating gains and producing more change, and 8) sustaining acceleration. Kotter (1996) contends that all eight stages are essential, that change is dynamic and messy, and that effective leadership is critical to successful change. He describes engagement and building relationships in the second and fourth stages of his model, as shown in Figure 10.
Transforming Analysts’ Roles in a Fortune 10 Technology Company
The second example comes from the technology sector where Paul Horstmeier, currently the chief operating officer and a senior vice president at Health Catalyst, served as a senior vice president for Hewlett-Packard. He oversaw a large organization of 720 people in 78 different countries with over 2,000 distributed IT systems. In a very complicated technical environment, Paul led his organization through a series of transformations, including restructuring the role of analysts. When Paul took over the analyst group, he was dissatisfied with the traditional ticket-oriented, report-queue model in which they operated. The group was isolated as an all-purpose utility resource and was drowning in producing reports. This situation concerned Paul for several reasons. He worried that those requesting the data from his team were turning the accountability for making sense of the data to those creating the reports. He also knew that spending all their time creating the reports kept the analysts from doing any actual analysis. Finally, he was concerned that, because of the reporting backlog, his analysts were becoming irrelevant—it took so long to produce reports that people simply went around them or stopped asking for reports. As a result of this backlog, the group’s constituents’ satisfaction was very low. What he found was a complete disconnect between the reporting his team was doing and the actual use of data to make a difference. He decided to change the situation by putting his analysts on teams so that everyone involved in working with the data was accountable for making a difference. That way his analysts could play a role that had a specific business impact. The first step in driving this change was to create a better technology infrastructure. His group implemented a self-service OLAP tool that enabled people with simple data requests to get the information on their own. At first, his analysts felt threatened by this technology and worried what they were going to do with their time. To help them use their time more effectively, he put them on teams. Working on these teams, his analysts were able to apply their data expertise directly to business problems. However, getting the rest of the organization to use the infrastructure they had implemented and to bring the analysts on board presented a challenge. He had to make a clear business case to get leaders in the organization to think holistically—and to see the value of the new infrastructure and processes. These are the steps Paul took to overcome this challenge:
- He found a senior leader who was empathic to the big picture (in a healthcare setting, this might translate to finding an executive sponsor in a clinical or operational department with a holistic mindset). He explained his vision to her, and she became invested in pulling his team members in to work with her group.
- He worked with this senior leader to refine his vision and to create a compelling message for those who would be involved. They first identified the people on her team who would be most receptive to the change and then targeted their message to them. They tried to think from these individuals’ perspective: Why would the business be better if we did this? Importantly, their message focused on the positive. They did not declare, “You have to do this.” Rather, they created a compelling enough case for change that these team members would at least be interested in trying it.
- He piloted the new system with this group and made sure they had a good experience. He and his team went overboard to make sure the group still received the same level of service or better.
- He made sure that the group they worked with became the heroes of the project—even at times when his own analysts did 80 percent of the work. This helped the group believe they could make the change work. When they saw their success, they became evangelists to other groups about the success of the new approach.
At that point, Paul knew they had successfully ignited change. But that didn’t stop challenges from popping up. Even when he thought he had sold the approach up and down the chain of the organization, someone would suddenly raise an objection. He had to keep feelers out for these situations and be ready to address them. Positivity, repetitiveness and results were key to driving the change forward and making it stick.
TEXAS CHILDREN’S HOSPITAL: USING AN EDW AND MULTIDISCIPLINARY TEAMS TO DRIVE CHANGE
One Health Catalyst client that is having considerable success using analytics technology and multidisciplinary teams to improve quality, patient experience and cost is Texas Children’s Hospital. To address the challenge of the impending transition to value-based reimbursement, Texas Children’s Hospital launched a quality and safety initiative in 2006 to develop a comprehensive and integrated enterprise-wide data management infrastructure. The first step was to implement an electronic health record (EHR) to collect raw clinical and financial data from across the enterprise. Although the EHR proved tremendously valuable as the means of digitizing care across the hospital, Texas Children’s IT leaders soon discovered that the newly digitized clinical data was hard to extract and combine with other data sources in a timely manner. Clinicians and quality teams still lacked access to meaningful information they could use to guide clinical quality interventions and improvements. “Our clinicians thought that the EHR would be a silver bullet to get the data they needed for quality improvement and operational reporting, and they blamed IT when the information wasn’t forthcoming,” recalls Texas Children’s Senior Vice President of Information Services Myra Davis, M.E. “The comment I would hear is, ‘I can’t get the right data from them,’ or ‘they don’t understand what I need from them.’ It created nothing but frustration.” For her part, Davis was frustrated that the IT department was quickly becoming a “report factory” for the rest of the hospital. Beginning in September 2011, the hospital worked with Health Catalyst to implement a healthcare EDW designed to unlock meaningful data trapped in the EHR and other applications to meet clinicians’ expectations. Today, Texas Children’s IT team uses the EDW to create near real-time reports from data aggregated from a range of clinical and business systems. With the EDW in place and self-serve analytics rolled out to clinicians, the IT department receives fewer report requests and experiences faster reporting times—which has made a world of difference for that team. Rather than spending all their time responding to an endless queue of report requests, department analysts now are able to function in their intended role – uncovering patterns in data to reveal the most productive operational and clinical improvements. Thanks to these improvements, business partner satisfaction with the IT department has increased dramatically. With this infrastructure established, Texas Children’s has been able to dedicate data analyst resources to multidisciplinary quality improvement teams. Healthcare data analysts are serving as the data experts on these teams to tackle clinical and operational projects effectively. They are successfully:
- Improving clinical care outcomes
- Driving labor cost savings and eliminating capital expense
- Implementing better processes for rolling out evidence-based guidelines
- Streamlining operations and care delivery in the radiology department
- Integrating their patient satisfaction data to deliver better care and improved operational efficiencies
- And more
While reaching this level of success with analytics may seem like an overwhelming task to organizations just starting out along the path of healthcare transformation, it doesn’t have to be. With the right technological infrastructure as a foundation and with organizational structures that make data analysts key members of multidisciplinary improvement teams, healthcare organizations can successfully and sustainably improve quality and cost and meet the challenges of value-based care.
Would you like to use or share these concepts? Download this presentation highlighting the key main points.
REFERENCES American Hospital Association. (2014). Price transparency efforts accelerate: What hospitals and other stakeholders are doing to support consumers. Retrieved from http://www.aha.org/research/reports/tw/14july-tw-transparency.pdf Centers for Medicare and Medicaid Services. (n.d.). Readmissions reduction program. Retrieved from https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/Readmissions-Reduction-Program.html. Centers for Medicare and Medicaid Services. (2010). National health expenditure projections. Retrieved from https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/NationalHealthExpendData/downloads/proj2010.pdf. Curley, A.L., & Vitale, P.A. (2012). Population-based nursing: Concepts and competencies for advanced practice. New York, NY: Spring Publishing. Hall, H.R., & Rousell, L.A. (2014). Evidence-based practice: An integrative approach to research, administration, and practice. Burlington, MA: Jones and Bartlett Learning. Health Catalyst. (2014). Survey of CHIME members ranks analytics the number one IT priority. Retrieved from /news/analytics-outweighs-accountable-care-population-health-icd-10-as-an-it-priority-say-health-system-execs/. Institute for Healthcare Improvement (IHI). (2014). IHI Triple Aim Initiative: Better care for individuals, better health for populations and lower per capita costs. Retrieved from http://www.ihi.org/Engage/Initiatives/TripleAim/Pages/default.aspx. Kotter International. (2104). The 8-step process for leading change. Retrieved from https://www.kotterinternational.com/8-steps-process-for-leading-change/. Kotter, J.P. (1996). Leading Change. Harvard Business School Press.
NOTES 1. The newsletter respondents included CIOs and other executives, as well as vice presidents, directors and analysts in BI departments.