I recall when a colleague told me in 2014 that some transgender women advocates were unhappy with iPrEx. I was surprised and disappointed. Since the study was conceived in 2004, I had struggled to include trans women for so many reasons and against so many objections.
We wanted transgender women in the study because we knew that the HIV burden was high among trans women, who have a prevalence of infection that far exceeds the general population (by 49 fold) and also exceeds the HIV prevalence among gay men. Yet, we were told by our sponsor (NIH) that we would not have enough trans women to make any separate conclusions and we should strive to study a well defined group of people. Such is a common in clinical trials striving for “purity.”
By the time the study started, I was happy that the protocol still allowed inclusion of trans women. The study inclusion criteria had been defined to be “male at birth.” I would now say “assigned male sex at birth.” We could include people regardless of their current gender identity.
Some transgender women still did not feel included. I recall a meeting in Sao Paolo Brazil in 2007 where a woman sitting in the back asked why the study protocol was called “Chemoprophylaxis for HIV Prevention in Men” (The protocol title used by the sponsor). After all, our inclusion criteria allowed women to join provided they were assigned male sex at birth. In a response which later became emblematic of my naïveté, I responded that whatever the study title, we welcomed the participation of transgender women.
Her anger exploded. She retorted that such language was far from welcoming. How could she possibly feel welcome if her gender was denied? And how could she trust us if we were so naive. She announced that she wanted to withdraw her support from the study and from our community advisory board.
Shortly thereafter, the head of study recruitment at the site informed me that without her support, he doubted that he could recruit more than a few participants in the trial. I was puzzled. After all this was just one person who was upset, and I had experienced many upset people in the course of PrEP studies in Cambodia, Peru, Nigeria, and Cameroon. Furthermore, no matter how much we reached out to trans people, the majority of our participants would be gay men, who have a high HIV infection rate in Brazil. I thought that transgender women might understand why I thought they might be men, at least in part, even if only a little.
The Paulista master of recruitment (patiently) explained to me that although trans women were relatively few in number in Brazil, compared with gay men, they were very prominent for their leadership on cultural issues, including gender and sex. Gay men looked to trans women for their courage and their beauty, and many wished they could be more feminine. This particular angry trans women was a leader among trans women and gay men, and slighting her had been catastrophic, however unintentional.
By that same afternoon, a “gender basics” workshop had been convened for all investigators and staff, with mandatory participation. I was a student in the front row. We heard from transgender women, gay men, activists of all types, and recruiters. We heard about how painful it was to be assigned a sex, and never feel it, and to be bullied and sent out of homes, schools, and work whenever one’s own gender was expressed. So many end up in sex work because it affirms female gender identify and it provides income when other employment is not available to “such people.” We also heard about anatomy and why so many transgender women keep their penises, despite having profoundly ambivalent feelings about them, so that they can satisfy their male clients who like having sex with women and who sometimes, usually in secret, like to be penetrated. Being transgender is hard, and investigators (like me) who cannot use the right words make it so much harder. I felt ashamed… and learned something.
Within months, the we had changed the official title of the protocol in Brazil to include trans women. We had adapted the informed consent to be gender affirming. The trans leaders came back to help enroll the iPrEx study successfully. Hooray!
Over the following 7 years, through the conduct of the iPrEx trial and the following open label extension, I spent time with transgender women. So many trans women recruited for our study, provided gender balance for staff volleyball games, entertained thousands with shows, and were so very charming on evenings out on the town. I came to imagine that I had come to terms with transgender women.
So I was surprised to hear that people were complaining in 2014. The complaint was that the primary iPrEx report (NEJM 2010) was a “case study” of how investigators lumped transgender women with men who have sex with men, with no acknowledgement of how the groups were different. Furthermore, the abstract of the paper noted that transgender women (who had sex with men) were included in the study, yet they were not reflected in the title of the paper. (The title had been edited for brevity with my consent). Perhaps more importantly, only 29 women (1% of the cohort) were noted to have been included, and there was no separate analysis of their experience in the study. The experience of another 13% of the cohort who identified as “trans” was completely obscured.
The news of disgruntled trans women came as we were writing up the iPrEx Open Label Extension (or OLE). For this reason, we included transgender women in the analysis of PrEP uptake and adherence. The manuscript was on a rapid timeline, as the journal (Lancet Infectious Diseases 2014) had generously agreed to publish at the same time the information was presented at AIDS 2014, the largest AIDS conference in the world to be held that year in Melbourne, Australia. The project was the first to describe experience with PrEP in a cohort who knew what they were getting (no placebos), so the data were hot. Amazingly, a complex and comprehensive analysis was completed by our heroic statistician, David Glidden, and multiple reviews and massive revisions were conducted in a short 4 week period – this process often requires 12 months. We were notified that the paper was going online as a published article on the same day we were to make the information available to the world at the conference.
To my horror, on the Saturday before the conference, I noticed that something had gone very wrong with the analysis of trans women. We had noted (in several places in the draft manuscript) that trans-identified women had had comparable retention and PrEP uptake, and their adherence to medication (measured by drug concentrations in blood) had been high. Yet, in a last final review of my slides days before the conference, I noticed in one of the data tables that transgender women were depicted to be only 5% of the OLE cohort included in the analysis. This could not be correct because transgender women had been about 14% of the randomized cohort, so their retention could not have been comparable to gay men if only 5% were present in the second phase of the study. My heart sank and I felt nauseated. Sleepless, not because of the jet lag, but because I had made some mistake in the analysis and I did not know what.
On that Saturday, our data analyst (Megha Mehrotra) stopped whatever young women do on a long Summer Saturday in San Francisco, and devoted herself to finding out what had gone wrong. Several hours later, I learned that the group of trans women had been mistakenly analyzed, in that trans women at only some study sites had been included, and not others. There had been differences in the survey instruments were coded because of differences in the words used for trans in different cultures and different societies. (The iPrEx study involved 11 sites in 6 countries on 4 continents). In fact, the correct analysis would show that Trans women had stayed in the study (good news!), and had opted to take PrEP at comparable rates (good news!), but had significantly lower adherence to PrEP (and not higher as we had proudly proclaimed in the paper about to go to press).
How do I stop the presses on a Saturday night in Melbourne half way around the world from London where the publication was about to be released on Tuesday morning? I wrote emails to everyone I knew at the journal and spoke to several of their contractors on the telephone – everyone was polite but there was no way to help. No sleep for me Saturday or Sunday night. Finally, monday afternoon London time, which was Monday morning Melbourne time, I received a call from the editor saying “it is ok, we will make the changes, the corrected version will come out on time.” Whew!
I resolved to make sure that the trans story in iPrEx be told. I made a video about trans women’s experience in iPrEx with Shannon Weber of UCSF HIVE in October of 2014. The protocol statistician and I met with leaders of the UCSF Transgender Center for Excellence in 2014. These leaders, Joanne Keatley, Jae Sevelius, and Madeline (Maddie) Deutsch, had made some complaints about lack of transparency about the trans experience in iPrEx. At that point, I was willing to admit that we needed to do much more to tell the trans story, and that I was not competent to do it without help. I apologized for how long it had been since our original publication in 2010. We committed to write a paper together, under Maddie’s leadership.
Developing this paper took more than a year, partly because everyone involved had other projects, and mostly because the work was difficult. We struggled to define who among the iPrEx participants should be included as transgender women, as only 29 identified as women, and so many more identified as “men” (when given only two choices) yet also identified as “trans.” Still others identified as neither “women,” nor “trans,” yet were taking hormones for feminization. More challenges arose from the curiosity and integrity of the trans specialists, who insisted on getting every detail right; because there was so little published information, they wanted to verify every point. The notion of “feminizing hormones” turned out to be complex, and we discovered late in the process that many gay men were using a hormone that blocks testosterone, not for feminization, but as a treatment for baldness. Such balding men did not identify as trans or women and had no desire to look more feminine, so they should not be in an analysis of transgender women. This led to yet another delay while all of the analyses were redone. Other mistakes were discovered and refinements were made, and time was consumed.
I think the transgender leaders invited me to give a plenary at the Transgender Summit in April 2015 partly to make sure that I stayed engaged in the analysis and that I stayed attentive to making sure that everything was done correctly.
I went to the Transgender Summit in Oakland from April 16 and 17, 2015, with the paper not yet submitted, yet almost ready. The conference proved to be a delight. I learned so much about feminizing hormones, reproductive capacities, and the struggles that people have. I met and talked with transgender women and men. The people-watching was stunning, and exhausting, as so many amazing people challenged my conventional notions, in very attractive ways.
I was proud to be included. I celebrated my “transition” toward cultural competence with regard to gender identities and perceptions, and I was forgiven for still having a long way to go. One trans man explained: my first puberty took 7 years, as did my second puberty. I realized that my 7 years working with transgender people, and not ever really getting it right, was a good start to a better life.
We submitted a completed manuscript about trans women’s experience in iPrEx for peer review in April 2015, and it was published in September 2015. The paper about trans women came 5 years after the original iPrEx report – so so very late, yet so important for me, and for us.
The slides for my talk are linked here: Bob Grant Trans Plenary 2015 trimmed final.
Key lessons learned for HIV prevention researchers was recently published.