#Design #Quiz — How often should you take this medication?
Here is the medication label.
You should take 1 capsule by mouth every day, right?
But look at what it actually says:
You actually should take 1 capsule by mouth every day.
I am sure this was not intentional and I can imagine there are a lot of factors that influenced this design outcome: a smaller than normal medication box, a printer that requires a certain sized label, typesetting that led to a carriage return for text to the very right edge of the box if a medication has complicated instructions.
It’s a clear example of how typography matters for health, as I have shared before in a previous post.
And typography is just as important for the electronic health record as it is for the prescription label.
Here is a real-world example of how a test result in the electronic health record looked like on a provider’s mobile telephone screen:
#Design #Quiz — Is the result of this test positive or negative?
Here are the same test results viewed on the desktop:
Did you assume that the result was negative based on the mobile view? It was actually positive!
The design of the text, the carriage returns, and the organization of the sections all contributed to an erroneous conclusion on the mobile view. Let’s dissect out the design of these displays.
First let’s look at the Mobile display:
The primary results of the test do not have adequate space or size, so it’s easy to miss the most important result, a positive test!
The reference ranges are misaligned and mixed in with actual results, and the results are split among multiple lines, leading to possible reading errors. Is the result the reference range or is the reference range the result? I would interpret the first “Normal” as a negative test result even though it’s really just part of the “Normal Range” header!
Now let’s look at the Desktop Display, which was slightly more interpretable but still has issues:
The reference range, positive result, negative results all look the same (and not just for desktop but also for mobile)! There is no emphasis with color, symbol, weight, or size, leading to the result blending in with the surrounding text.
Finally, the primary reference range is all the way at the bottom separated from the initial result all the way at the top. The visual space gives this line seemingly more importance than anything else. This might confuse the viewer into thinking that the reference range is actually the result!
To me this is a frightening example of how bad design can lead to medical errors. The provider assumes a negative result, the patient is never notified, and the opportunity for proper followup and treatment is delayed, leading to adverse long-term consequences for the patient.
This is a major healthcare #design #fail.
I deeply appreciate both the academic and industry colleagues who shared with me some of the best practices for visualizing laboratory results in electronic health records (see below for reference links):
- Keep multi-component test results together when displaying them along with other test results.
- Display result information in columns wide enough for users to see the full value and abnormality level without having to adjust the display.
- Use consistent format for abnormal results, regardless of source.
- Results outside normal reference ranges or abnormal in some other way must be visually distinguished from normal results.
- Don’t just use color to signify abnormal flags. Color and an icon or type styles at a minimum.
- Make sure reference intervals are present and appropriate.
Here are examples of better designed interfaces, thanks to my design colleague Jake Dwyer!
The next time you look at the design of your mobile or computer screen, think about its implications for patients and providers. There clearly is a greater need for focus on the design of Electronic Health Record!
Click here for information about creative commons licensing. Disclosures: Medical Advisory Board of GoodRx.
College of American Pathologists Laboratory General Checklist courtesy of David McClintock
UT Health School of Biomedical Informatics Electronic Health Record Usability
Electronic Health Record Association Electronic Health Record Design Patterns for Patient Safety Sept 2017
Perrotta et al. Validating Laboratory Results in Electronic Health Records: A College of American Pathologists Q-Probes Study