Three Simple Words That Will Improve Your Data Displays

When you hear the words “just do it” you immediately get the message — stop your whining and complaining and get your lazy butt moving.

“Just do it.”

When it comes to healthcare data displays the three words that matter most are…

“At a Glance”

In sharp contrast to the pain we often (and correctly) associate with “just do it,” the mantra “At a Glance” should serve to remind you that viewing your reports and data displays should be a pain-free experience.

Here are a couple of re-designed tables and graphs that illustrate my point.

In the first example below, you see a table of data. Tables are very helpful when you need to look up specific values; not so much when trying to quickly and elegantly impart information about trends or actual performance compared to goals. For “at a glance” insights, a well-designed graph does the job much better.

People are ridiculously, crazy busy-multitasking at illegal levels busy. And so you can count on getting their attention for maybe all of five seconds (maybe being the operative word).

So now imagine, you have five seconds to explain to your boss how well the operating rooms (ORs) are being utilized compared to target using the following table. Not happening.

improve your data display 1

Restart the clock – now you have five seconds and another chance to show your boss you know what the OR utilization rates are compared to target using the following graph:

improve your data display 2

Wow, much better. Remember, just because the data and messages that we need to communicate are complex, the display of them need not be.

By displaying the information about OR utilization in a horizontal bar graph and adding a line for the target, it is easy to see the information in about five seconds — or less. The addition of the monthly utilization rates in the base of each bar provides the viewer with additional information that they often request (and yes, I know it is a bit redundant but we must make some minor concessions!)

Next, using the following graph, can you tell the boss what the medical center’s medication occurrence count, rate, severity and (to keep it interesting) most frequently noted causes of occurrences for the second quarter of the fiscal year that began October 2010 were?

improve your data display 3

I concede that part-to-whole relationships can be a challenge. But even here there are ways to show the information “at a glance.” First off, it’s fine to break down the component parts and then put them back together so that the information is easy to see and understand. (In other words, avoid piling on all in one graph.) Think about the parts of the story that you need to convey and then consider if they are better de-constructed and then re-constructed, as shown in the graph of the same information below:

Take a “glance” at this:

improve your data display 4

I used a simple line graph to show and trend the actual count of medication occurrences on the primary Y-axis and the rate per 1,000 patient days on the secondary Y-axis. I used a line graph in this case because it shows the trend over time nicely and I could use the Y-axis on both sides. The viewer of this graph can easily see the trend — the high-level summary.

Because it is not only important to understand the trend but also the severity of the occurrences I used a simple horizontal bar graph which is a good way to compare the different severity categories and to order them hierarchically from no injury to death. Then I positioned the graphs directly beneath the line graph so that they correspond with the quarterly timeframes.

And finally, in order to eliminate medication occurrences it is important to know the reason why they happened. So, I created three vertical bar graphs for each quarter, which details the cause of the occurrences. I chose a vertical graph for this part of the display because I want the viewer to be able to see how they may change over time and if I had chosen a horizontal graph the viewer might think that they were also categorized by severity given their proximity to the graphs directly above them. And by only labeling the first graphs Y-axis I have taken advantage of the fact that we read from left to right and used a limited amount of space to pack in a lot of information. The use of a vertical graph makes it clear that the reasons are for all of the occurrences.

I created all of these graphs using Excel and simply pasted them on the page to create one graph.

“At a Glance” reminds us how we should able to see and understand the message in charts and graphs. So get in the habit of taking the five second “At a Glance” test — I guarantee you will be amazed at how effective it is in helping you to assess and improve your charts and graphs.

Okay, so you know what I’m going to tell you next… “just do it.”

Leave a Reply