A Picture Is Worth a Thousand Words

It’s common for people to make decisions about an endeavor based on general impressions about its performance. We get a sense of a phenomenon occurring, and are likely to take action — or not — based on related assumptions. While no one would deny the value of listening to our instincts, the fact is that isolated data or vague notions about an enterprise can often be inaccurate.
Sometimes it’s best to make a picture of the data, and let the picture do the talking. Plotting data over time offers insights and maximizes the learning from any data collected by revealing patterns and improvement opportunities.
Let’s take, for example, a teenager’s routine complaints about being stuck with too many hours of homework every Sunday night.* The child’s father (a statistician) suggests that she try to figure out why this happens by writing down how much time she spends on homework each day for a month. The daughter agrees and at the end of the month has collected the following data on the number of minutes spent studying:
These data are interesting, but they really come to life and tell a story when plotted on a run chart:
Now, what do the data tell us? This picture clearly reveals that the Sunday night crunch is tied to the fact that she does no homework on Friday and Saturday. So she comes to realize that if she could get into the habit of spreading the load by cracking the books for an hour on Friday night or Saturday morning, she might end up with less pressure on Sunday.
There are several methods of displaying data graphically to aid in analysis. These include a “histogram,” which plots observations to show their distribution; an ordered bar chart (also known as a Pareto diagram), which illustrates relative frequency of occurrence; a scatter diagram, which plots observations to show the relationship between two sets of data; and a map, which plots data across geographic locations. Plotting data over time in a run chart, such as the one above, is an especially powerful tool in quality improvement. Improvement requires change, and change is, by definition, a temporal phenomenon. So tracking the path of key measures over time is an essential component of a quality program. And the visual display of data can offer insights that lists of figures alone simply cannot.
The application of this approach to improving health care is limitless. Consider, for example, a group practice of four internists in a suburb of a major city.* The group has identified a problem with patient waiting times, and they’ve decided to face the issue head on by collecting data and plotting them visually. They track the maximum waiting times in minutes each day over three weeks, with the following results:
Again, the data are interesting, but plotting them visually tells the story:

This chart clearly identifies two issues that had not previously been apparent to the group of clinicians: first, Monday is routinely their busiest day because demand for services builds up over the weekend. Second, one data point (Wednesday in Week 2) is well out of the normal pattern of variation. The group investigates and learns that this is when one of the providers is on call at the local hospital. Awareness of these issues will allow the practice to adjust scheduling and staffing to better accommodate patient demand patterns and the providers’ full range of duties.
Tracking and observing the path of key measures over time, displayed in chart or graph form, can be a compelling tool in an improvement campaign. The exercise can help your team make more informed decisions, incorporate contingency and back up plans, and better endure industry uncertainties. Being able to see trends and patterns in any area — length of patient stay, volume, visit demand, patient satisfaction, clinical outcomes, staff turnover, and so on — is a prerequisite to achieving continuous improvement. Progressive institutions learn to keep an eye on data at every stage, and instinctively aim to review the data in revealing graphic form wherever possible.
*Thanks to “The Improvement Guide” by Langley, Nolan, Nolan, Norman, and Provost.
Average Content Rating
(1 user)
Please login to rate or comment on this content.
User Comments