Earlier this year, I wrote a newsletter about emerging efforts to publicly report the amount of care being delivered to hospital patients (Worked Hours per Patient Day, WHPPD) as well as the quality of that care. I also noted the use in this reporting of a prototype created by HDV to explain and display the information. At the end of that newsletter, I observed that I would be crazy to say that any of this work was “one and done,” because it wasn’t and couldn’t be, and that in addition I was already thinking about more, different, and better ways to communicate data.
Enter our friends the good folks at Juice Analytics. They’ve created a new data-reporting and visualization platform called Fruition®. We have been watching the development of Fruition for a while now, and find the visualizations insightful and easy to use, with touch-friendly apps that work great on tablets. Given our goal to find new and different ways to engage the public in all publicly reported data, we decided to see what we could create with help from Juice and Fruition. Below are a few prototype screens (not yet interactive).
In addition to a growing volume of patient-satisfaction and quality data, hospitals are increasingly sharing information on staffing levels with each other and the public, to help everyone become better-educated and more proactive healthcare consumers. The challenge, however, is to communicate this complex information simply and clearly, allowing people to make informed decisions about how, when, and where to receive care. We know that this type of communication requires a multifaceted approach:
- serious attention to translating healthcare jargon and statistics into plain language
- combination of data in meaningful ways for display
- use of best practices of data visualization
Re-using the same example data from several hospitals that we’d shown in our first display, we created visualizations for two very different audiences: patients and their families; and healthcare professionals. In both displays, the information is communicated with a narrative at the top of each page along with several examples to explain the healthcare statistics and calculations presented.
Unlike our first display, however, this version includes high-quality photos that give it a very friendly and inviting look and feel, enhanced by large navigation icons and visualizations that make it easy to see how the data can be filtered and compared.
To communicate with the public, we have embedded summary information about each hospital, emergency department, and unit or floor right above the graphs, thereby eliminating the need to search or leave the page to find that information. In Patient View, we display completed-years data about three different quality metrics (pressure ulcers, falls, and falls with injury) for the selected hospital. The viewer can also compare these results to those in the same quality metrics for other hospitals. Below this display, we allow selection of the type of unit each user is interested in, the average number of patients cared for each day on the unit, and the average number of caregiver hours each patient receives.
We tried something slightly different (perhaps even a bit provocative in some circles) to display the Emergency Department (ED) quality metrics in the view below. The different metrics are displayed on what is sometimes called a lollipop graph for the selected hospital in blue and for the state average in dark grey. This display makes it easy to look across the section to see performance along several metrics for both groups. The last bar graph displays the average number of minutes patients spent in the ED before being sent home, with a brief statement as to whether the selected hospital is above or below the state average.
Healthcare Professionals’ View:
In the view designed for healthcare professionals, we created a graph of the difference between the staffing plan and the actual number of worked hours per patient day for previous years. (This display was not presented to patients, since it could be misread as a quality-of-care indicator instead of a record of actual-versus-budgeted. Using this display as a quality-of-care indicator would require extensive analysis to ensure accuracy.) As it is, the display allows healthcare professionals to see how many hours of patient care were delivered and how well their forecasts (and those of other hospitals) matched their actual worked hours per patient day (WHPPD).
Tackling this project again with a different reporting tool (Fruition) and a different group of designers was interesting and fun. Although many of the data visualizations are similar to the first ones we created, I love the warm photos and the overall engaging views. To me they feel less technical or sterile; I can really see myself sitting down with a loved one to learn about the information presented here, to have an informed discussion about where to seek care. This was a wonderful experience in thinking through the user experience more fully.
And yes, I freely admit that my idea of fun hasn’t changed much since our first run at this! There is still nothing like the thrill of tackling a really hard problem and exploring alternative solutions until you land on something that works. I (still) find it exhilarating!
P.S. Big thanks to Lindsay and Michel at Juice — nice work!