Prostate cancer decision aid - a retrospective
In February 2016, my team and I were briefed on a client project for a well known cancer research organisation - who’s name can not be disclosed in order to protect their intellectual property.
During this project we worked through a creative agency - Light Creative.
I was grateful to be placed on a project that would make a difference in people’s lives. It also was a direct link to future employment opportunities that I had in mind - things were looking good.
I was placed in a team with two researchers, being confident in my creative ability, so I decided it was the perfect opportunity to build my research skills.
A well known Australian cancer research organisation is seeking to better communicate the treatment options for low-risk, localised prostate cancer.
They want to facilitate conversations and shared decision making between doctor, patient and support people.
To fulfill this objective, we have designed a prototype for a website that communicates treatments and the associated side effects in a visual and accessible manner.
Our solution will assist men recently diagnosed with low-risk, localised prostate cancer, their family and friends, in their decision making journey and will also facilitate conversations amongst each other and with medical professionals.
Duration 17 days
Techniques and tools
- Competitor research
- User research - surveys, user interviews
- Data synthesis
- Empathy maps
- Persona creation
- Design studio
- Information architecture design
- Wireframe and prototype design
- Iterate and test
- Interactive prototype
- Further testing
In our kick off meeting, we were handed an interactive prototype and asked to incorporate a user-friendly decision aid and comparison table within it.
We were given no access to the client itself, nor the target users of the website. We were bright-eyed and eager make it work so we were confident resourcefulness.
The content of the prototype we were working with was a medically written booklet that aimed to guide people in their decision making for treatment options for four key treatments.
The content was described as set in stone.
As new user experience professionals, we aim to ensure any design solution we work on provides a consistent and holistic user experience. We could quickly see potential pain points within the design so set to validate our thinking through guerrilla usability testing.
We asked users two things:
- Find information about treatment options. Imagine that you or a family member or close friend were recently diagnosed with low-risk prostate cancer. How would you use this website to choose a treatment option that would be best for you or your family/friend?
- Discuss treatment options with your doctor. How would you discuss your treatment options with your doctor?
The results found:
- When users were asked how they would use the website to choose a treatment option, 10/10 users went to ‘information’ instead of ‘My Decision’ as the first point of call.
- Users questioned naming conventions and grouping of content - the information architecture wasn’t intuitive, it confused users.
- 10/10 users couldn’t navigate to the required relevant pages based on the tasks given.
- 10/10 users found the filtering system in ‘My Decision’ confusing. They questioned the use of the drop down menus and the location and purpose of the search function.
- 4/10 users commented that they would follow the advice of a doctor rather than a website.
- Users didn’t seem to understand how this page could facilitate a conversation with their doctor.
As difficult as it was, we fed the findings back to our client and recommended that we revisit the information architecture and asked to view the content. We believed that reviewing the content would provide more context and help shape a more seamless user experience.
Thankfully, our client granted us the license to move ahead with what we felt was necessary to ensure the website’s success.
It was at this stage of our project that we committed to split our resourcing when required to get the maximum output. In the following days I concentrated on solving the issue with the information architecture whilst my team mates reviewed our research and created personas and user journeys.
Following the kick off meeting, we had been doing our best to get to know the experiences of men and their family and friends in dealing with prostate cancer diagnosis. We cast our net wide, however our networks unfortunately (or more fortunately) lacked the numbers we needed in terms of prostate cancer experience.
With no ability to access users from the organisation, we opted to broaden the scope to people with experiences in dealing with any cancer diagnosis.
Our research approach was:
- Interview people who had/have cancer
- Interview people who’s family or friend had/have cancer
- Quantify interview findings through surveying
- Review websites for people with prostate cancer
- Analyse treatment option tables and decision aids
Within the first two days of the project we had interviewed 5 people that had been impacted by cancer, through their own diagnosis or supported family through a cancer diagnosis.
Key themes through the interviews under covered the raw emotion associated with a cancer diagnosis; the question of mortality, why me and shear worry of the unknown. Secondly, the speed in which treatment is selected and scheduled and perhaps the most concerning theme, that surgery seemed to be the only option for treatment at the time.
Quantifying findings through surveying
As these themes presented themselves unanimously, we quickly set to quantify them by means of surveying. We broadcasted the survey through our networks in hopes to source confirmation as well as responses from our target user group.
Over the space of a week we had 63 responses - 10 of which were males, the age of our respondents was evenly spread from 26-65.
- 80% of the respondents were family or carers for those with cancer
- 7 people had had cancer themselves.
- Shock, anxiousness and worry were the most reported feelings at the time of diagnosis.
- The most reported treatment options were surgery, chemotherapy and radiotherapy.
- The most important information for users post diagnosis was success rate, quality of life and recovery rate.
Personas and user journeys
We grouped together the values and behaviours of the people surveyed and interview to create two personas - the patient and the support person.
During our interviews and surveys, we found that older men were inclined to internalise their feelings and follow the advice of a doctor. With this in mind and coupled with the insight of support people being more likely to look for information online, presented us with our primary persona - the carer.
We used our personas to create a user journey.
Understanding competitors and best practice
During this process we also conducted competitor and comparative analysis on information based websites that presented a treatment option or decision aid. We assessed both their features and overall heuristics.
The best competitor website was Option Grid, however it’s usability showed us many areas for improvement.
The best comparative website as Diabetes Decision Aid - it communicated complex information simply and visually. Whilst areas of the interactive design could be improved upon, we felt this design could greatly influence us in our design studio.
As part of our data synthesis we mapped the business goals against the user goals, our purpose was clear.
We used the insights and ideas discovered in our our research phase to create a minimal viable product. We mapped our must haves, should have, could have and won’t have of features.
Our MVP was to create an elegant and familiar design that nurtured users through their decision making process. Our design would provide clarity and reassurance at a time that was quiet possibly the most stressful time in someone’s life.
With these needs mapped we moved into design studio… create a many designs for one screen as you can in 3 minutes, pitch ideas and critique for 1 minute, refine for 5 minutes.
During our design studio, we committed to design mobile first to ensure that our design would not restrict us down the track. We quickly created a paper prototype that allowed us to test our thinking in a short period of time.
Overall our design was on track.
Following our client meeting, we conducted a series of open card sorts - 5 face-to-face and 17 online. We used the table of contents within the booklet to form the content on the cards.
In just a few days we had our results. People had grouped the information by;
- Information about Prostate Cancer
- Treatment Options
- Information about making a decision
- Sex stuff
‘Sex stuff’ stemmed from the topics on the cards - erectile dysfunction, fertility, change of penile sensations etc - all topics that within context fitted under surgery.
We used all the naming conventions as a tag references that could be used within the search function of the website to allow people to refine the results.
Usability testing and iteration
We conducted a total of 24 tests over the period of three days. Each day bringing a new iteration to the design and new insight on how the experience could be streamlined and improved.
We conducted with usability testing with a range of people aged 30 - 85.
We found that the design of our decision aid was very conventional, people understood that they were inputting information that would provide a tailored output. Our comparison table whilst providing a point of difference, allowed people to visually digest the treatment options in one place. The conventional and visual design meant it was accessible and easy to learn.
The pain points we set to remedy throughout our iterations were:
- Wording of questions was ambiguous and at times confusing.
- Difference of the comparison table and the output of design aid was not clear
- Users preferred to discuss treatment with a doctor
- Often users ignored or questioned the function of the ‘personal notes’
- Trust issues around the credibility of the site.
By our final prototype we had managed to remedy the majority pain points before we handed it over to the client. The final stage of our prototype can be found below:
- Research experience, values and behaviours of target users
- Understand devices of target users - computer, tablet or phone.
- Test feasibility of personalisation and account needs.
- Seek approval of cancer research organisation to proceed with current success.
- Usability test with target users.
- Design in high fidelity
- Refine decision aid wording and ranking convention (important/not important)
- Add more statistics in a visual way to decision aid
- Add mouse over hints for medical terminology
- Create awareness and credibility of site.
- Educate doctors and patients on shared decision making for treatment
- Client expectations are not only managed at the start of a project, they need to be managed throughout the entire project.
- Divide and conquer - splitting resources allows you to cover more ground.
- Awareness leads to success - having daily stand ups and check ins allows teams to communicate their experiences and challenges. I found it invaluable to stand back and question our approach and effectiveness each day to ensure we were on track and getting the best from each other.
- You don’t always have to play to your strengths - sometimes not playing to your strengths allows you to develop yourself and build new strengths.
- Have faith - at many times during the project I questioned our progress and stressed over our ability to deliver the extend of what we needed to in time. You have to maintain faith in your process, your team and your ability to do what it takes.