Case Study: Asking the Right Question About Sex on an HRA

Case Study-In my previous job, one of the types of products I worked on was health risk assessments (HRAs). These tools are used by health plans, large employers, and other groups interested in managing the health of a population to assess the prevalence of certain health risks and guide decisions about interventions to offer. Typically an HRA is a long questionnaire about all types of health risks and behaviors, with some kind of feedback to the user at the end about the areas where improvement is most needed. As you might imagine, writing the questionnaire for these tools takes a lot of time and involves careful thought to make sure the questions included are meaningful and produce useful data.

A specific issue that my team and I struggled with was around asking the user’s sex. Usually this question is positioned toward the front of an HRA and then used to trigger skip logic so that men see some questions and women see others. For example, only men will be asked about prostate exams, and only women will be asked about pregnancies. These questions are important because a key driver for health risks is making sure people are having the recommended screenings and tests, and many of these are different based on biological sex. For most people who take an HRA, the question about sex and the related ones that follow are easy to answer.

But for a small group of HRA respondents, things get tricky. In this group we have transgender people, who may be able to confidently assert themselves as male or female but then receive misleading results on their health risk with the following questions. A transgender man will not be up-to-date with prostate exams, but this is not a health risk for him as he does not have a prostate. Receiving feedback that prostate health is a risk for him is off-putting and doesn’t help him make future health decisions (and frankly, doesn’t provide useful data to the economic buyer of the HRA, either).

Then there are cis-gendered people who also struggle with sex-based follow-up questions. These might include women who have had hysterectomies (a question sometimes, but not always, asked in the HRA to further tailor the question set) or people whose doctors have, for whatever reason, given them personal guidance that doesn’t align with the standard national medical guidelines. These people will also receive feedback from the HRA that doesn’t accurately gauge their health risks.

Finally, there is a small group of people who don’t identify as either male or female or are gender non-conforming. Adding a third sex option such as “other” might make that question easier for these people to answer, but it doesn’t provide the right information for the HRA tool to ask follow-up questions that get at true health risks associated with ownership of uteri, prostates, and other sex-specific organs.

Here’s where we back up and reconsider what we really want to know. Do we care if a person identifies as male or female? Not really, not for the purposes of the HRA. We care about what physical parts a person has so that we can see if they are going through the recommended medical maintenance for those particular parts. Whether the parts reside within someone who was born male but is now female or someone who was born male and remains such doesn’t matter in terms of the actual health behaviors we are trying to evaluate and encourage.

So really, what we want to ask is:

  • Do you have a uterus?
  • Do you have a prostate?

If the answer to either of these questions are yes, then we would ask the follow-up questions.

I haven’t yet seen an HRA that takes this parts-based approach to the sex question, but I am curious to see if it works to alleviate the discomfort that some HRA respondents experience around these items. I think that it could help improve the quality of the data received. More importantly, from the perspective of someone who cares deeply about creating user-friendly tools that actually help change behavior, it might help the makers of HRAs avoid alienating a small but important group of people for whom the traditional question construction just doesn’t work.

Personally, I am against asking whether a person is transgender as a solution to the sex question issue in HRAs. Although more broadly there is a need to better understand this population, the HRA has not traditionally been used for this purpose. To me it feels intrusive and disrespectful to ask whether a person is transgender unless the HRA is providing follow-up specific to issues transgender people might face such as increased risk of violence by others or behavioral health issues (not something I’ve ever seen an HRA do) OR it is being used to collect data that will provide a better understanding of population health in the transgender community. In the latter case, it would be nice to explain to users that this is a goal of the HRA so the purpose of the question is understood. Until and unless HRAs start to offer feedback and guidance around transgender health rather than parts-based health, I think the parts-based questions makes more sense for that type of tool. Do you agree?

Writing about this topic took a path I didn’t intend. My initial interest was more around instrument design and how there isn’t always a good link between the way we’ve historically asked questions and the data we seek. But the path this thinking took illustrates how important it is for people working in health and behavior change, whether they are working directly with people or through technology development, to be socially aware. This issue is not JUST about question design. It’s about how to treat a group of people in a way that is sensitive and respectful, while also achieving the goals of the tool you’re creating. It’s a difficult challenge. I know that many of the people who work to design HRAs and other tools want to help solve it, including me.

Additional Resources

The Williams Institute at the UCLA School of Law has created a comprehensive report around survey assessment of transgender and gender non-conforming people that will help survey designers ask questions about biological sex and gender identity in a sensitive way while creating a better understanding of this population. I recommend checking it out.

Another great resource if you are interested in the health and wellbeing of transgender people is MyTransHealth, a new startup designed to connect trans people to physicians who can care for their particular health needs. You can also contribute to their Kickstarter.

World Professional Association for Transgender Health: Lots of info and links.