Test result display - User research & usability testing

Updating the display of test results

Background

The test results page is one of the most visited pages in Patients Know Best (PKB). Aside from the positive feedback about showcasing historical data and allowing patients to see the trend of all their test results at a glance, we had also received feedback from patients and professionals that they would also like the option to see a list of just their latest test results in a table on the page.


After running an initial round of interviews to better understand our users’ pain points, we began usability testing of possible updates to the page. We decided to use various methods, interviews, prototype testing, preference testing and first-click testing. We tested the updates with both existing users and those who have not accessed PKB before, users were aged 20-75+ from around the UK.

Findings

Initial set of interviews

We held our first set of interviews with patients already using PKB to better understand our users’ pain points, to see what they felt was missing from the test result display and how it could be improved.

Patients said the most vital information to them was their latest set of test results. Some even wake up in the middle of the night just to check them. These test results inform the patient on how their disease is doing. In some cases, also indicate whether they should change their lifestyle in some way or seek further medical attention. Some patients felt that PKB could improve the way this information was presented, to allow them to see their latest test results more clearly and give them access to a condensed list of recent results.


There were a few initial requirements from these sessions.

  1. All of the patients wanted the latest test results to stand out more clearly

  2. All of the patients wanted to see a list of the most recent results with values

  3. Some of the patients wanted an indication of whether the results were up or down from the previous value

  4. The patients liked having a search button to find test results


Second set of interviews

Taking the findings from the first set of interviews, we built some high-fidelity wireframes to display test results in a tabular view addressing the various points raised in the initial set of interviews. We then ran another series of interviews to gather patients' feedback on this new style of displaying test results.


The following feedback was collected from these sessions:

  • (Still in progress)


Further wireframe testing

We wanted to test the very foundations of our new concepts, that is the structure and visual hierarchy, so we ran the tests on some black-and-white variations to see which layout people prefer.

Three versions used for wireframe testing


Preference testing - Showing whether results are in or out of range

We ran these tests with 20 people in each round between 44-64 all from the UK.

The first test we tested three variations with only minor differences. What we wanted to know is which layout helps certain information stand out more clearly, in this case this information was the out-of-range test result. In the test, we showed people all three variations next to each other in random order and then asked “which design helps you spot your out-of-range test result more clearly?”

  • Version 1 had the labels on the right side of the result.

  • Version 2 had the in-range and out-of-range label tokens below the test result value.

  • Version 3 was just simply written out and the label was not featured so prominently.

The outcome was that the overall majority (87%) favoured Version 1 (the one with the label token to the right as opposed to below the value (13%) or written out without a token (0%).


Feedback snippets from preference testing

  • because I read left to right and it's quite natural for me to look for information in that way

  • the placement and alignment of the in-range and out-of-range bar means that the out-of-range stands out clearly

  • I found it much clearer to see at a glance


First-click testing are in or out of range

After the initial preference testing, we turned to another method called first-click testing. This technique, as the name suggests, is usually used to determine where people click first when given a task to complete and how quickly they do so. We ran these tests with 20 people in each round between 44-64 all from the UK.


As first-click testing measures the time it takes for someone to make a decision; our primary intention was to see which version helped people spot the out of range test result more quickly. The users were instructed by the task “You want to find out more about your out-of-range test result. Where would you click to get that information?”. After participants clicked the out-of-range test result they also answered three questions to get further confirmation on what we see from the data:

  1. “On a scale of 1 to 5 how easy was it to find the out of range test result on the screen?”

  2. "On a scale of 1 to 5 how clear did you find the presentation of your test results information?”

  3. “If you could change one thing about the design to make it better for you what would it be and why?”

The results suggest that the version with the label token to the right-hand side again performed significantly better than the other version where the label token was below the value. Participants found and clicked on the out-of-range test result much more quickly when the label token was on the right-hand side and also favoured it more, 67% said it was extremely clear and easy to use as opposed to 47% on the other version.


On the version where the label token was to the right, all participants clicked on the out-of-range result, whilst on the other version, lots of participants clicked on other areas of the user interface including the search button. This seems to further confirm that the time to click numbers suggest that people spotted the out-of-range label token much more quickly when it was to the right.


As this was a black-and-white version lots of people commented on the fact that they would prefer the out-of-range label token with a different colour as opposed to the in-range label tokens.


Feedback snippets from first-click testing

  • You could use a different colour for out of range although not red which could cause alarm.

  • Colour. Make the out of range red and in range green or something

  • A contrast between "IN RANGE" and "OUT OF RANGE" icons/labels, so that the latter stand out more clearly

Outcomes

Initial set of interviews


We took the feedback from the initial set of interviews and looked at how we could fulfil these patients' requirements in the near future. We agreed on the following outcomes:

  • Adding the test result value to the notifications panel

  • Making the latest test results stand out on graphs by:

    • Removing all numbers from charts except the latest result

    • Showing the value only for the latest result in a detailed information pop-up style

    • Changing the detailed information pop-up so that the value comes first and then the date and range and additional info

    • Changing markers from filled to with an outline to make the result stand out

    • Only showing the range line when hovering over a result

Adding result values to notification panel

Highlighting latest result

Showing whether results are in or out of range (Preference & click testing)

  • We will use the more prominent labels on the right side of the result

Second set of interviews (Still in progress)