NHS.UK Connecting to Services Sexual Health Service Finder
Timeframe: 9 months (September 2017 - May 2018)
Role: User Researcher
Research and design team: Mark Branigan (Senior User Researcher), Simon Wilson (Senior Interaction & Service Designer), Emma James (Content Designer), Joe McGrath (Lead Product Manager), Eva M Barabas (Senior Software Engineer)
My primary responsibilities: User recruitment, creating research studies, moderating user research sessions, analysis of quantitative and qualitative data, faciltation of insights playback sessions, accessibility.
Outcome:
1547% increase of click-through conversion of test-finding journeys
67% increase in people finding a suitable service
211% increase of people under the age of 25 ordering a Chlamydia testing kit in the Yorkshire region (traffic coming from NHS.UK to freetest.me)
Design pattern from this digital service was later reused and scaled to other sexual health NHS services.
Problem:
People that were using previous online NHS services in Yorkshire had problems finding appropriate facilities and treatment options in their area for their specific sexual health problems. People needed an easy-to-use and transparent way to find testing options and treatments for their sexual health conditions. Our team at NHS.UK Connecting to Services embarked on creating a digital service in which people could easily and quickly find the most relevant treatment.
“How can we make it easier for people to find and receive treatment for their specific sexual health condition?”
“How can we increase the number of at-risk people who go on to find suitable treatment?”
During the discovery phase, the goal was for our team to understand the problem that needs to be solved. In this case, it was understanding our users and their behavior, both offline and online, within the context of finding help for their sexual health condition. During this research process we embarked on exploring the following areas:
Motivations and what users are trying to achieve.
Any constraints we may face making changes to how the service is run.
Underlying government policy regarding sexual health services.
Opportunities to improve the service and accompanying online and offline touchpoints.
At the start of the discovery phase, we were not presented with a pre-defined solution or told to build a specific thing. Instead, we were given space to explore the problem area and make evidence-based decisions through iteration from the inception of the project. Our exploration of statistical data of sexual health in the UK lead us to find that chlamydia was the most common STI in the Yorkshire area. Through interview sessions with frontline NHS workers in the Leeds area, we found that most health facilities were inundated with inquiries from people wanting to find treatment options for chlamydia. We also ran a banner intercept survey on the live legacy sexual health website and found that it was underperforming in terms of solving problems the users had in terms of finding suitable treatment.
Using this data, we believed that the current service was underperforming and that building a chlamydia service finder that could be scaled to other sexual health service finders would be the most appropriate option.
Being agile means starting with a good understanding of the needs of real users. Not by making a long list of imagined requirements.
Testing assumptions in discovery.
We tested our assumptions discovery by focusing on 4 main areas of exploration and validating or invalidating our assumptions to ensure that we were making evidence-based decisions and solving real problems with the service we were building. Utilizing a mixed-methods approach, we focused on 4 main areas of exploration:
Clinical facts and statistics (desk research of statistics and clinical policy)
Constraints and organizational policy (policy research and interviews with key stakeholders)
User beliefs and attitudes (surveys, user telephone interviews, and interviews with front-line workers)
User information-seeking behavior (surveys, user interviews, and digital service data)
During the discovery phase of this project, we created an assumptions board.
It was important to keep our assumptions visible and ensure we were making decisions based on research evidence.
Main findings from Discovery
Among the findings in our discovery phase we found that:
The main barriers to getting tested were embarrassment, stigma, and lack of knowledge about the accessibility of the test and the testing procedure. Many did not know that STI results do not have to go on your health record and that discreet testing options were an option.
NHS.UK receives the most clicks from chlamydia-related search terms (just over 40%) and sexual health search terms (29%).
STI diagnosis rates in 15-24 years olds are twice as high in men and 7 times as high in women compared to those ages 25-49 (PHE, annual data tables 2016).
Data about services at physical NHS locations exists but is scattered & inconsistently organized.
Data about services at a location differs between online/offline offerings and private/public ones.
'Free' screening test sites lead to paid-for kits as different testing services have localized funding.
Irrelevant of age, chlamydia cases are rising (datasets from 2015).
15-24-year-olds are the largest proportion of people diagnosed with Chlamydia at 62% (2009).
Increasing testing/screening has proven to reduce cases of chlamydia.
Using these insights helped us construct hypothesis statements as we moved into design in Alpha.
Alpha
In alpha, we had a chance to explore new approaches and take the best concepts and findings from our discovery into ideation for this phase. We spent alpha building prototypes and testing different ideas for the problems we learned about previously. To kick things off our senior designer lead a design sprint, in which each member of our team independently sketched ideas for a digital sexual health finder for people that have chlamydia. After each team member independently sketched their design concept, we then dot voted on various functionalities and design elements that we felt were most useful for users. After dot voting we had a post-sprint debrief discussion regarding which design elements we wanted to synthesize and bring into a higher fidelity design concept.
Design sketches for a digital sexual health service finder for people who may have chlamydia.
Silent team dot-voting and reflection
After our design sprint, we then created an early-stage high-fidelity prototype in Heroku. Through user research sessions with the prototype stimulus, our goal at this stage was to find out if people that have or previously had chlamydia found what we are building useful and usable. We used the prototype as a prompt for introspection, activity, and discussion over a series of usability sessions. Our goal at this stage was to observe user behavior with the design prototype, gain attitudinal feedback, and use this data against hypothesis statements and make future design iterations.
While working on this project I was regularly moderating lab sessions with research participants.
After several lab sessions and discussions with NHS frontline workers, we found a range of usability problems with the service, it was also clear that we needed a more holistic understanding of the offline and online touchpoints and triggers user experience when searching for information and treatment for their condition.
Using the insights we had collected, we constructed an end-to-end service design map to visualize the real-world process people go through when searching for sexual health information and finding treatment for their condition. The value of this design artefact was to better visualise information as a team for alignment and to find pain points and areas of confusion in the user journey that our digital service could better facilitate.
service design map for people seeking treatment for chlamydia
After synthesizing lab session data and visualizing the offline and online touchpoints people go through when finding information and treatment for their condition, it was clear that our early-stage prototype was not solving key user needs. Several themes started to emerge at this point.
Wording for chlamydia symptoms needed to be brief, clear, and specifically worded for men, women, and genderqueer people.
People seeking treatment need transparency regarding discretion as it pertains to their health records.
People demand discretion around accessing public health facilities for their sexual health concerns to avoid embarrassment (e.g. not seeing someone they know in a waiting room).
People overwhelmingly did not know that a discreet, unlabeled testing kit could be mailed out to them where they can test for chlamydia in the privacy of their own homes.
People wanted anonymity if they need to notify a previous sexual partner if they tested positive for chlamydia.
Treatment options were a concern but secondary in terms of privacy, speed, and discretion.
Using these insights we iterated the design of the sexual health service and treatment finder to focus on the core areas of:
Clearly worded unambiguous language around chlamydia symptoms for specific genders.
Transparency regarding STI test results not going on health records.
Clear language regarding privacy in the testing process.
Clear guidance around the options for STI testing based on area.
Addition of an anonymous notification messaging option so people can notify previous sexual partners of their results.
Insights from research for future design iteration of the service
During this Alpha research phase, we also carried out field research in the form of field visits to sexual health clinics, pharmacies that gave out testing kits, as well as ordered STI testing kits to better understand the service design.
Field research gathering insights of offline testing and treatment service design
We continued to make design iterations focusing on the previously mentioned user needs. I constructed a research plan in which we brought the new prototype into lab research with 15 users that had not had a history of any sexual health concerns and were not familiar with finding testing and treatment options. In this lab, we observed few usability issues and reception was overwhelmingly positive regarding content design, privacy, guidance around location-based services, and anonymity in notifying a previous sexual partner.
Moderation of research lab with people without a history of sexual health concerns
Focus on accessibility in alpha
In order to ensure the service we were building is accessible and easy to use for everyone, I lead on accessibility efforts in which I tested our service internally with testing tools such as Jaws screen reader, dragon voice recognition, and NVDA to find any opportunity for accessibility enhancements before researching with users with access needs.
Running an internal accessibility audit with JAWS, NVDA, and other accessibility tools before starting research with people with access needs.
Over the course of a month, I lead research planning and facilitation for 2 separate accessibility lab sessions with a total of 14 users with access needs ranging from motor-skill impairments, visual impairments, low-digital skills, autism, dyslexia, wheelchair access needs, and hearing impairments.
Feedback regarding the content design was positive. Users found the content “informative”, “useful”, “straight forward”, “jargon-free”, and “non-judgemental”. The service also tested well with assistive digital technologies.
However, users found the guidance and information on the results page for in-person testing and treatment to be lacking.
People with accessibility requirements voiced the need for clear guidance when taken on an offline journey
We used these insights to make content and accessibility enhancements regarding offline guidance as it pertains to users with accessibility needs.
Beta
As the service matured and we were confident that the service was solving legitimate user needs through the data we had collected, we moved into experimentation with the live service, utilizing A/B testing of different design patterns to track user conversion. I also set up a Hotjar feedback study and ran various usability pop-up studies in the Leeds area.
I was regularly hosting pop-up usability sessions at student centers around the Yorkshire area.
I worked directly with our data scientist to measure a range of KPIs in Adobe Analytics, Hotjar, and Qualtrics.
Our main areas of feedback and experimentation were:
If people found information that was helpful (Qualtrics)
Service funnel conversion to ordering a sexual health testing kit (Hotjar & Adobe Analytics)
Service usability (pop-up usability sessions)
After measuring a month’s worth of traffic in Adobe Analytics, we saw that the click-through rate for at-risk people going on to book a test-findings journey increased to 28% for the treatment vs 1.7% in the baseline live service (117,347 unique visitors split-tested)
Over the same month, we measured if users had found a service that was suitable for them. Of the 117,347 unique visitors, 16,714 answered our short Qualtrics survey. From the split test that was run, we found that 87% of people found a suitable service for the condition within the treatment compared to only 52% in the control (legacy service).
Total research for this project
Impact of this project:
1547% increase of click-through conversion of test-finding journeys
67% increase in people finding a suitable service
211% increase of people under the age of 25 ordering a Chlamydia testing kit in the Yorkshire region (traffic coming from NHS.UK to freetest.me)
Design pattern from this digital service was later reused and scaled to other sexual health NHS services.