LVNDR Health · Mobile App
Summary
Healthcare systems lack LGBTQIA+ data, and that gap feeds bias and inequality. I designed an adaptive questionnaire that collects this data before each consultation, without losing the trust of the people answering it.
The pilot app scored 83 on the System Usability Scale, against an average of 68.
Introducing
You feel sick, what do you do? You're struggling with anxiety, where do you go? A doctor, right? Maybe a therapist in the second scenario. At least, this is what I would do.
However, when you're part of a marginalised group, like many in the LGBTQIA+ community, the data reveals a problematic reality. People who need care the most often avoid it.
Below are some key findings from research LVNDR ran with 800+ LGBTQIA+ testers.
Some quotes from LVNDR's research
Being continuously misgendered is so invalidating and demotivating.
I need to speak to someone who understands me, who doesn't judge me when I tell them I have multiple partners.
Being able to speak to a doctor from home makes me feel safer.
I still have to explain to my doctor that I need a cervical screening even though I'm a trans man.
The bigger picture
Lack of trust and fear of discrimination keep LGBTQIA+ people away from healthcare. Healthcare systems end up lacking LGBTQIA+ health data, which increases data bias towards the community and, with it, health inequalities*.
*With AI, this bias can become bigger and even more excluding. I spoke about this at the 2024 UX Healthcare Conference in London.
A common scenario
Let's have a look at a very common scenario. An illustrative story, based on what we heard in our research. Use the arrows to follow it.
1 / 7
2 / 7
3 / 7
4 / 7
5 / 7
6 / 7
7 / 7
The system
In traditional care, patients often have to repeat their story at every appointment, revisiting painful memories, while clinicians struggle to find their history. The questionnaire answered both. Patients share their information once, before the consultation, and clinicians arrive knowing who they're talking to.
We started with over 80 questions. Working with clinicians and the product team, I organised and prioritised them, grouped them by relevance and marked each one as essential or optional.
The questions covered demographics, medical history, sexual history, mental wellbeing, risk and vulnerability, and drugs and alcohol. Some were asked along the journey, exactly where they were needed. For example, we asked for preferred pronouns during onboarding, so we could address patients correctly in every communication from then on. Sex assigned at birth came later, only when we needed it for lab tests.
Some questions were essential to complete the PrEP pathway. Others we wanted answered to start building the largest LGBTQIA+ population health database, to help reduce inequalities in healthcare.
I then designed a dynamic flow where each answer shapes the questions that follow. For example, users with a negative or no HIV test result continue the questionnaire, while those who test positive are directed to a different service.
This keeps cognitive load low and gives every user a questionnaire journey tailored to them.
Defining UX/UI principles
An 80+ question questionnaire about sexual and mental health is a lot to ask. So I had to find different strategies to help people complete as many questions as they could, in the least frustrating way possible. Through rounds of research, we understood the blockers and found ways around them.
We organised questions into clear modules on the home screen, each showing how many questions it has, so users can move through them easily.
Insight
People won't engage with a questionnaire if they don't know how long it will take.
Working with the clinical lead, we added space for extra details on some questions, which also gave us richer and more personal data.
Insight
Rigid answer choices don't fit everyone's health situation, especially inside a biased healthcare system.
We added an “I don't remember” option, so nobody gets stuck or picks an inaccurate answer just to finish a section.
Insight
Not everyone remembers when they last had sex, especially with an active sex life and multiple partners.
To build trust, I added statements about data privacy in context.
Insight
Users hesitated to answer certain questions because of concerns about the security of their personal information.
How might we encourage users to complete the questionnaire before their consultation?
The goal was to respect their time and avoid intrusive notifications, which could also raise safety and privacy concerns.
First, we placed must-answer questions within key parts of the app, to make them relevant.
Then, to find the best times for engagement, we used a motivational diagram. It helped us build a logic of in-app reminders and notifications, sent when users were most likely to respond.
The motivational diagram. This work was carried out together with an external designer specialised in behavioural psychology.
Reaching people when they're most motivated comes from behavioural science. Timing the reminders around those moments made people more likely to complete the questionnaire, without compromising their experience or privacy.
Design with care
Asking 80+ questions about sex, drugs and mental health to people who don't trust healthcare is a responsibility. Gender and sex assigned at birth were asked as two separate questions, so no one had to choose between who they are and what their body needs clinically. And every video consultation started by checking the patient was in a safe space.
Before launch, I led usability sessions focused on clarity, inclusivity, language and trust.
Testing also showed that some questions, especially in the mental wellbeing group, could be triggering. I still remember one participant telling us: “This question is tough. What if I answer that I had unprotected sex and that means I'm at risk of HIV? Then I'm just left feeling anxious.”
So where a question could cause anxiety, we added support resources and helpline numbers right next to it. Other questions we removed from the app completely and moved into the consultation, especially those about abuse or other situations of vulnerability. Some data is too sensitive to collect through a form, and needs a person on the other side.
Outcomes
Feedback from our pilot users shows the impact of a remote service that is more accessible, less stressful, and starts to close the data gap.
For LVNDR, it meant we could demonstrate that we were able to gather more and better-quality data. Nearly 100% of our 100 beta testers completed the questionnaire before their consultation, and the service was much more efficient than mainstream in-person care.
What I learnt
I've always supported LGBTQIA+ causes, but working at LVNDR showed me how much I still had to learn. Most of the team were part of the community, and their firsthand experience of inequalities in healthcare added immense value to our work.
As we move into an AI-driven world, hiring people from marginalised communities is one of the most effective ways to reduce data bias, something I spoke about at the UX Healthcare Conference in London. And in the end, our most valuable feedback came directly from our users.
Next up
