Can We Understand Blood Test Results without Numbers? Ask Participant 23, Nightingale

About a year ago, I handed in my dissertation. I was examining new visualisation of blood test results amongst UK patients. The aim was to make the design of blood test results patient-oriented whilst staying truthful to the structure and the data. I was not the first one who desired a change. In 2010, WIRED magazine conducted a study to add clarity to blood test results. In 2011, the science journalist and healthcare innovator Thomas Goetz published a Ted Talk on the matter. 

Blood tests are one of the most common types of medical tests. They have a wide range of uses, such as diagnosing a condition, assessing organ health, and more. The test results are highly valuable – they provide up to three times more information than other patient data sources such as medical history, physical examination, and symptoms. Traditionally, the test results are presented exclusively with numbers alongside the text of the acceptable range. This information requires an interpretation by a medical professional, as patients typically lack the training to decipher the findings on their own. Blood test results provide meaningful information to four sets of users: the doctor who ordered the tests, the nurse who draws the blood, the laboratory worker who performs the analysis, and the patient who needs information. A change in the test results display should benefit all users. Doctors, for example, can more quickly discover irregularities in the results of an individual patient.  I focused my study on the patient as the user. I was eager to research visual representations without numbers at all. I hypothesised that all patients with essential health and numerical literacy could engage with their results if those data were visualised. My project was supervised by the inspiring Dr.Simone Gumtau from the University of Portsmouth. 

I believe that all design projects should follow the Design Thinking approach. Personally, I follow the “5-stages-model” as taught on IxDF; This model is based on five non-sequential phases built into an iterative process: empathise, define, ideate, prototype, and test. I interviewed a few non-medical, UK-based patients. to see if I could improve the patients’ engagement and motivate decision-making processes. I conducted two qualitative interviews with each participant. In the first interview, I aimed to gather information regarding the participant’s current experience with their blood test results, the traditional result sheet. The key was understanding the “whole picture” of needs and reasons for conducting the test, in order to get a broader understanding of the problem I wished to conquer.

After the interview, I synthesised and analysed the data and built a hierarchy map.

 Using the granular blocks of the map, I designed a personalised prototype for each participant, representing numerical data by colours, graphs, and text. I presented this prototype at the  second interview, which was conducted in order  to evaluate my concept. 

During my studies, I had a professional crush on Alberto Cairo. It started while reading his book The Truthful Art, which gave me countless aha moments. I decided to model my data visualisations based on Alberto Cairo’s qualities for great communicative visuals. First, I wanted to reduce the patients’ feelings of overwhelmingness. For example, one patient was timid about sharing her results at the interview’s onset due to a lack of confidence regarding how to read the results. So, I divided the  text and data into categories and eliminated information that might be considered unnecessary for the non-medical patient. (this should be examined by a medical practitioner).

I learned that patients struggle with understanding the meaning of the test and its relation to their symptoms. Therefore, I designed the new visual to present the results using causality order: the relationship between the cause of conducting the blood test and the effect of the results. As such, I displayed the results in a nested treemap chart. The bigger the rectangle, the more likely the test result is related to the cause of symptoms or diagnosis. 

I love the functionality of a treemap because it offers both hierarchical displays and part-to-whole relationships. With it, I generated a non-linear display. There is no right/wrong method of reading the content. Also, I limited my palette and used the traffic light colour scheme to communicate effectively without numbers, as colours offer fast attribution of the relationship between value and the acceptable range. To make the traffic light palette more colour-blind friendly, I tried to avoid the green-orange problem by using light, medium and dark shades- where the red is warm, and the green is cool. 

Presented here are the medical category and test titles without numerical values. The home screen (depicted above) uses negative white space as a buffer between the elements. 

The user can “drill down” into each category by clicking on it. 

The second level offers a detailed overview of the tests it contains. The colour-coded background represents the acceptable range of the test result. Using fewer words reduces the page length (ideally with no scrolling) and allows the text to “breathe.”

The third level offers a detailed review of the test results. This level is the only place I present the result’s numerical value because the patient shows deep interest in the resulting drilling to this level. My research discovered that numbers could deliver a sense of assurance and authenticity. I display the numeric value of the test result on a one-axis bar chart without units. The axis has a definite numerical indication of start and finish. I chose this representation because it simplifies the ability to read the data; it delivers complete access to the information without requiring a high level of numerical literacy.

Once I had a working prototype, I conducted the second interview that acted as user testing with participants to gather feedback on the redesign. I will focus on Participant 23’s reactions as she was the first to receive the numberless prototype.

She was content browsing the prototype because she realised most of her test results were within the acceptable range. Her natural next step was to seek a possible explanation for her symptoms. She was also eager to view the abnormal results. 

She reviewed the dietary recommendation section, as her doctor prescribed and understood the causality that led to this prescription. She was able to take action within minutes of exploring the prototype. I was so happy about that! 

Overall, the participant reported having a pleasant experience exploring the prototype. The click-ability of opening and closing categories was intuitive. She stated that she is pleased to see numbers and graphs only in relevant places, giving her a sense of validity. In conclusion, insights were gained, and actions were taken (for example, regarding vitamin D insufficiency).

To conclude, my proposed solution achieved two goals. First, it made the data approachable, so participants felt confident interacting with their results. Secondly, the visuals were insightful enough to encourage action according to stipulated diagnoses. All my participants in the study were made aware of their diagnoses and conditions through engaging with the personalised prototype, which also led to action initiations. 

Future roadmaps, only one “round” of iteration was conducted in my studies, the design thinking approach feels incomplete. Amending the prototype with the participant’s request was not part of the scope of the dissertation and is missing. Also, Interviewing more users with their results might cast light on different subjects. Ultimately, examining the visuals with a medical expert in the field could clarify the critical sections of blood test results regarding the decision-making and action-taking processes of the participant. These recommendations require further examination and might affect the visual and the outcome. But that would be a different study, or perhaps a PhD.

In addition to blood test results, other medical information could benefit from a change. I wish we could shift the power to the patient by applying data visualisation rules and a design thinking approach to all medical readings. It could potentially impact patients’ lives significantly.