May 8, 2013 at 7:20 pm EDT | by Daniel O'Neill
Crunching the numbers
Kathleen Sebelius, HHS, gay news, Washington Blade

HHS Secretary Kathleen Sebelius has made good on her promise to collect data on the health of LGBT Americans. (Washington Blade file photo by Michael Key)

The momentum is decidedly on our side. A gay male professional athlete has come out. America is overwhelmingly in support of same-sex marriage, and the march for legislative approval of our relationships is well underway. And yet, progress is still lagging behind in one of the most important areas of our lives: LGBT health.

But the tide has certainly turned, as some notable advances have recently been made to curb the many health disparities our community faces, including disproportionate rates of certain cancers, smoking, substance abuse, HIV/AIDS, mental health problems and barriers to accessing care.

Marking a critical first step in addressing these shortfalls has been the collection of LGBT health data, which began in earnest on Jan. 1 through the National Health Interview Survey (NHIS) – viewed as the principle source of information on the health of Americans for more than 50 years.

With the inclusion of sexual identity questions in the NHIS, Health and Human Services Secretary Kathleen Sebelius has finally made good on her promise to collect data on the health of LGBT Americans, which will continue through Dec. 31 as part of a supplemental section, the Adult Sexual Identity Questionnaire.

Although inclusion of sexual identity questions in future versions of the NHIS is still not guaranteed, the ripple effect of this initial supplement is expected to drive other federal agencies, state departments of health and researchers, all of which look to the NHIS for guidance on structuring their public health surveys, to follow suit.

In summer 2011, while heads of many LGBT organizations were content that future data collection was a done deal, other advocates remained skeptical when Sebelius spoke of first needing to “market-test” sexual identity questions prior to their inclusion in the survey. Many felt this delay was unnecessary, since sufficient studies had already determined appropriate ways to collect such information. As such, they questioned whether the marginal gain in data accuracy from further question refinement would offset the loss of another two years without data on LGBT people’s lives.

On Oct. 24, 1997, then-HHS Secretary, Donna Shalala, faced an analogous situation, when she mandated “the inclusion of information on race and ethnicity in all HHS-sponsored data collections systems.” Within a week the HHS-wide policy went into effect, covering all federal public health surveys, sans the unnecessary field-testing now required for sexual identity questions. Sebelius could – and some argue should – have done the same; but other forces were at play this time.

In particular, due to concern over the sensitive nature of the sexual identity questions potentially causing reporting bias when collected, many worried that prohibitively expensive requirements, such as the use of audio computer assisted self-interviewing technology to administer such questions, would ultimately prevent their inclusion in the NHIS. Fortunately, the “market-testing” phase proved such requirements unnecessary; but this study could have easily gone the other way. Instead a decision was made to use flashcards to administer the sexual identity questions in lieu of interviewers asking respondents directly.

The main question, posed to respondents at least 18 years old, has two versions depending on whether the individual identifies as male or female.

Which of the following best represents how you think of yourself?

(Lesbian or)* Gay
Straight, that is, not (lesbian or)* gay
Something else
I don’t know the answer

Those answering “Something else” or “Refused,” are prompted by further clarifying questions. In its current form, only by answering “Something else” is a question about gender identity even offered. Unfortunately, this structure confusingly entwines sexual orientation and gender, two distinct dimensions of human identity.

The data will be first available to the public through a quarterly release of the raw data in September 2013, covering January through March 2013. The complete 2013 dataset will be released in late June 2014. For those interested, the survey questions and future data can be accessed through the website:

Asking these simple sexual identity questions signifies a monumental step forward for the health of LGBT Americans that should be celebrated.

However, our work is not done. We must continue to push for refined data collection in future iterations of the NHIS and other surveys as part of their core datasets.

  • DHHS had been collecting data on race for decades before 1994 and the requirments and categories had gone through repeated changes. What changed was the push for uniform standards across the federal government and the need to improve tracking of data collected as part of research grants. It’s not as if someone said collected the data and they miraculously appeared.

    The push for adding sexual orientation and gender identity questions to federal surveys came via the Institute of Medicine report on LGBT health, which was published in 2011. At that time, there were no standards for asking about sexual orientation or sexual identity. The only examples were from a small number of federal surveys, including one in which the sexual orientation item an option used by only a few states..Adding new items to surveys is expensive and time consuming process–the 2 year period is not unusual. Surveys are planned long in advance. The testing of the items involves asking questions multiple ways and debriefing participants in how they answer the question. It’s not as though this was done by someone while watching an old movie on television. In fact there is a paper available on line describing how this was done for NHIS: [[URL REMOVED]]

    I’m guessing that much of the info here came second hand rather than representing much in the way of research. Regardless, 10 minutes on Google would have enabled someone to fact check this enough to know that the writer didn’t know his subject.

    • Rich:

      I appreciate your more nuanced perspective on what happened in the 1990s. My intention was not to suggest data on race and ethnicity was not collected prior to that time; but rather, as you indicate, that there was no mandated, standardized way of collecting this data until Sec. Shalala made the change. I strongly believe that a basic, standardized LGBT demographic dataset should similarly be mandated as a core part of all federal public health surveys. It should not be seen as something exotic, with need for excessive question investigation and scrutiny, nor should it be collected as a supplement on a one-off basis – as is now the case.

      I’m also well aware of the process that has gone into developing these questions (and that it is still ongoing); however, my point was that this extensive process was not entirely necessary in light of the desperate need to start collecting data on LGBT people and for fear that the ACASI matter might derail the process altogether (costing orders of magnitude more than interviewers simply asking the questions) before any data were ever collected.

      After all, as I’ve argued, with the current question scheme – even after 2 years of market testing – data on gender identity will likely be significantly underreported, since transgender individuals must respond “something else” for a question about sexual orientation in order to be asked about their gender identity. Most transgender people I know confidently identify with a particular sexual orientation; so when they answer “gay/lesbian” or “straight,” the fact that they are transgender will not be captured. I mention this to highlight that although question refinement is still needed, it has not prevented us from finally proceeding this year with asking sexual identity questions. As such, I believe we could have started asking questions, especially on sexual orientation, much sooner.

      I’m all about doing what’s necessary to ensure that the most accurate data is collected; however, there certainly comes a point when this marginal gain in accuracy from further question refinement does not outweigh the cost of not collecting any data. I felt we had reached that point; but perhaps, that’s where we differ in opinion. Had questions been asked earlier we’d have data by now. Instead it will be more than a year from now before we have access to the complete raw dataset – with no summary report even guaranteed. …Another year, with no data on the health disparities impacting LGBT people’s lives.

      I’m also quite aware of the IOM report, as I used to work there, I’ve read the report and was present when it was released – having spoken to members of the committee that pulled it together. I’m also well aware of its recommendations in driving this effort. In fact, the report actually states what I have argued: “Data on sexual and gender minorities should be included in the battery of demographic information that is collected in federally funded surveys, in the same way that race and ethnicity data are collected.” So I’m unclear why you even bring it up.

      Finally, I have been following this subject for a number of years and spoke briefly with Secretary Sebelius in 2011 on her efforts to include LGBT health data in the NHIS. I’ve also had conversations with a number of LGBT health data experts who have been closely involved in this process and voiced to me their concerns. I also spent much time on the phone with officials at the CDC to determine the current status of the NHIS survey and details around ongoing collection and future release of the data. So your “guess” is incorrect; in fact, much thoughtful research went into this piece.

      If you have something additional to add to enlighten us further, please offer it up; as you seem to have some inside perspective on the matter.

      Ultimately, I’m very happy LGBT health data collection has begun; but it’s essential that we continue to push for its expanded inclusion in the NHIS and other surveys as a core demographic variable – per the IOM recommendations and our own common sense.

© Copyright Brown, Naff, Pitts Omnimedia, Inc. 2018. All rights reserved.
Blade Blast

Get the latest LGBTQ news to your inbox every Thursday!