Interactive Healthcare Dashboards Are Gaining Momentum
Back when I ran a large anticoagulation clinic, I could have really used the valuable insights an interactive healthcare dashboard provides. The dashboard certainly would have helped me be a more effective manager and clinician because I would have had access to near real-time data about my patients’ quality of care.
Instead, I had to submit report requests to a busy analyst and wait. The analyst then had to manually compile each report based upon my requests. I might get a report fulfilling my requirements several weeks later. By that time, my needs had likely changed, and I would need to resubmit a new request. Because the analyst was so busy with requests to build many different reports from different departments—and the analyst lacked the right tools to pull the data efficiently—he rarely had time to, well, help analyze the data. So I had to rely on perception to gauge the well-being of my patients rather than use objective, holistic data. A far cry from evidence-based care!
Even after I received the analyst’s report, I still had to take the time to sift through the data to interpret it. Why? The data was displayed in an Excel spreadsheet with raw line-item data that lacked visual displays of trends or comparisons. What clinician has time to dig through raw data to answer questions like the following: How are my patients doing compared to last week, month, or quarter? Is our patient volume increasing or decreasing? What is the breakdown of my patients by diagnosis?
The same inefficiencies still plague today’s healthcare industry. While it was frustrating back then, these days, with the move to value-based care, decreasing waste and improving care is a high priority for health systems.
The Need to Improve an Inefficient Reporting System
Analysts’ reports are at the heart of how health systems are run. Many people within the organization depend on these reports to understand how the areas and the patients they’re responsible for are doing. From healthcare executives to department managers to clinicians, data analysts receive many report requests each day. But this process ties up many highly trained analysts with the manual work of gathering and compiling data using tools like Access and Excel. Such tools are only capable of producing a static report, which has limited use as an improvement tool. This is because static reports lack the interactivity and visual display capabilities needed to help those who requested the reports make sense of the numbers presented.
Static reports are deficient in other ways. They often lack the context users need to understand how and where to implement change to drive better performance and practice evidence-based medicine. After all, clinicians are trained using data and taught to critically evaluate patient therapies based upon data in the literature. They want to do the right thing for their patients, yet static reports don’t provide the necessary insights clinicians need.
Static reports also don’t enable clinicians to compare their outcomes to their peers or to national standards. For example, a physician may believe that his or her length of stay (LOS) metrics reflect quality care. Without access to other LOS outcomes, physicians can’t compare metrics and discover areas for improvement.
The Case for Interactive, Healthcare-Specific Dashboards
Healthcare dashboards are critical for users (e.g., clinicians, CEOs, and improvement teams) who need quick and insightful answers to their questions in an easy-to-understand visual format. It is much easier for workers to glance at a line in a green range on a dashboard (see Figure 2 below) to see if metrics are still in the desired range rather than trying to digest a monthly line item report of patient data. Plus, users don’t need to know SQL or other querying languages to dig into the data to find valuable insights.
Because of the near real-time data dashboards display, workers can visualize where they are, where they are going, and how fast they are headed there. This enables quicker course correction if needed.
Dashboards also show users if they are holding their improvement gains for previous initiatives or if they are slipping as a previously targeted project falls off the radar. For example, Texas Children’s Hospital was able to produce a 15 percent reduction in unnecessary chest X-rays for asthma patients in just one and a half months by using a dashboard. Another health system, MultiCare, developed a modified early warning system (MEWS) dashboard for sepsis patients to quickly identify those who were trending towards a sudden downturn. Their dashboard now serves as an early-detection tool for caregivers to provide preemptive interventions. In just twelve months, MultiCare was able to reduce septicemia mortality rates by an average of 22 percent, leading to more than $1.3 million in validated cost savings.
Dashboards also serve another purpose—they give daily reinforcement to workers that their work is making a difference. This boosts confidence in their ability to make improvement changes. It also boosts job satisfaction because workers are able to maintain those gains and see the results of their efforts in the dashboard.
The Dashboard as a Single Source of Truth
A healthcare dashboard helps spread a much-needed single source of truth across an organization. Everyone—from executives to frontline clinicians—can look at the same data. When everyone in the organization has access to a consistent, reliable source of truth, then everyone can speak the same language, spread a system-wide standard of care, and work together to implement improvement initiatives. For example, because a dashboard offers a single source of truth, the dashboard can be used to maintain the definition of each specific goal. Is the organization’s goal a decrease of 20 percent or 30 percent? Having the objective defined in a consistent way across the system helps everyone remain focused on the right target.
5 Key Qualities of a Good Dashboard
There are many key aspects to building a good dashboard. But as a clinician who has worked with both high-value and low-value dashboards, I’ve found the following features critical:
- Be easily accessible. A dashboard should be easily accessible to each user who will need to tap into its insights. The typical arrangement where analysts email reports to a user is ineffective because the user then needs to save the report somewhere or search their inbox to find the report. If the report isn’t easily accessible, they are unlikely to reference the report when making decisions.
- Display reliable data. Users need reliable, trustworthy data; if they don’t trust it, they won’t use it. Including those who use the data in the build and validation process can significantly help with team buy-in.
- Contain relevant data. A good dashboard should only contain the factors users need. If the dashboard has the ability to report on 50 metrics, but the user only needs five, the extra 45 metrics just clutter up the user’s abilities to focus on what’s important. It’s better to highlight five key metrics than water down the dashboard with 50 metrics.
- Use timely data. A dashboard needs to contain near real-time data so users can address challenges promptly. For example, it’s difficult for providers to follow up on why a patient treatment is outside a particular protocol if the data arrives weeks to months after the fact. But if the provider or department can see near real-time information about a patient’s episode of care, it’s easier to intervene while the circumstances are still fresh in the team’s memory.
- Include trends and/or benchmarks. Trends and benchmarks show users where they’ve been and where they are going. If improvement efforts don’t move the needle on cost or quality, then users need to know, so they can change whichever intervention isn’t effective. Or if the improvement initiative isn’t having an immediate impact, users can still view the trending data to see if their efforts are making a difference. This keeps them engaged and motivated.
Figure 3: Screenshot of a readmission summary for selected pneumonia patients
Why Healthcare Dashboards Need an Enterprise Data Warehouse
Of course, a dashboard doesn’t just fall from the sky. Healthcare dashboards are built on the foundation of a late-binding enterprise data warehouse (EDW). Binding data later means delaying the application of business rules (such as data cleansing, normalization, and aggregation) for as long as possible until a clear analytic use case requires it. This approach to an EDW is ideal for “what if” scenario analysis and is best suited to the ever-changing world of healthcare data.
Because the dashboard is built on top of the EDW, teams can pull data from multiple source systems (e.g., EMRs, financial, patient satisfaction scores, and research) into the dashboard. This gives the team a well-rounded view of their performance metrics. Users can also monitor a clinical intervention and watch the impact on cost—all at the same time they’re following balanced scorecard performance metrics for patient satisfaction.
The beauty of adopting a late-binding data warehouse is that the time-to-value for the setup of the EDW enables users to tap into the data’s insights much sooner than with the setup time for a conventional EDW. For example, clinicians do not have to wait for the completion of all data mappings in the entire EDW (which can take months or years), but can jump in on areas of interest within weeks or months of implementation of the EDW to begin to see relevant insights and make real interventions. Clinicians can then focus on the immediate needs of their patients and the organization, rather than trying to define every possible scenario they might be interested in querying at some point in the future.
Improving Outcomes with Access to Easy-to-Use Dashboards
Healthcare is constantly changing and growing. Dashboards built on a late-binding data warehouse platform give users the ability to be agile in their approach to analyzing and using data. Late-binding data warehouses are also scalable and adaptable to the healthcare industry’s need to improve quality and decrease costs. In addition, business rules and vocabularies in the industry are changing at an unprecedented rate. The late-binding data warehouse provides fast time-to-value, plus it has the agility required for the demands of today’s healthcare analytics solutions. Combined with an interactive, healthcare-specific dashboard that’s easy to use and provides near real-time data, users now have access to the deep insight their data holds and can use the knowledge to drive improvement initiatives.
Do you use a static report or an interactive dashboard? If you’re working only with static reports, can you see the value of adopting a healthcare dashboard?