I don't even know anymore

(TL;DR at the bottom!) here's the full paper, by the way. it's called "Transgender and Nonbinary Deaths Investigated by the State Medical Examiner in the Portland, Oregon, Metro Area and Their Concordance With Vital Records, 2011-2021". it's a study about discrepancies in death certificate data and medical examiner records, with the goal of helping people make death certificates more accurate in the future.

it's important to remember that this data is:

local to the Portland metro area

from January 1st, 2011 through September 30th, 2021

a very, very small sample size (only 47 people!)

like to give you an idea of how small a sample size it is, the article says this:

Estimates of the transgender population in the United States vary; a recent estimate is 0.6% of the adult population (or around 1.4 million); for Oregon, the estimate is nearly 20 000 individuals. This article focuses on the Portland, Oregon, metro area (Clackamas, Multnomah, and Washington Counties), which has an estimated total population of 1.8 million residents; 0.6% of the adult population is an estimated 9500 transgender individuals.

also it seems this study only included information about violent deaths, or deaths that were investigated for potentially being violent, because that's what was available.

Oregon was one of the first states to participate in the National Violent Death Reporting System (NVDRS), which contains detailed information on violent deaths, including homicide, suicide, and overdose. The NVDRS has a variable to collect gender identity (for violent deaths only) via a checkbox for transgender status but does not have specific gender categories. This difference is important, because research has demonstrated an increase in violence among nonbinary transgender persons versus other transgender individuals.

In Oregon, the Office of the State Medical Examiner investigates certain types of deaths (Oregon Revised Statute 146.090) in its jurisdiction, including by possible suicidal, homicidal, and accidental means.

Note that “natural” deaths are not under the purview of the ME system.

and very important to note that these are only deaths where investigators knew and recorded that the dead individual was transgender. there are likely many other trans people in the system whose deaths were not included in this study for that reason.

Moreover, if death investigators did not ask about the decedent's gender identity, then the data would not be included in the case narrative, thus never making it into the NVDRS pipeline. Furthermore, there is a multiple year lag time with NVDRS data, and generally data at the county level are not easily accessible.

and the study goes even further in saying that the MEs probably wouldn't have recorded a patient's trans identity unless it was relevant to their death

Gender identity is collected in the narrative portion of the report if it is perceived as forensically germane to the death investigation. However, sex at birth on a decedent's official documentation may differ from the self-identified or preferred gender of the decedent. The 2015 US Transgender Survey reported that only 11% of respondents had all their identifications with the name and gender they preferred, while more than two-thirds had none of their identifications with their preferred name and gender. Thus, evidence collected by the investigator regarding gender identity and transgender or nonbinary status may be inconsistent.

(the bolding here is mine)

I want to make sure everyone here sees what I'm seeing: we really have no idea how many people in their death records are trans. we have no fucking idea, because people straight-up do not record that shit, and to pretend that this study is a great and balanced example of every single trans person who died in the Portland metro area would be very weird.

we can't know how many trans men were only listed as female, or how many trans women were listed as male, or how many nonbinary people were listed as either male or female. that is an unknowable thing for us!

this is one reason why erasure is a kind of violence: even when people are trying to study it, the lack of information being recorded means that we do not have a way to do that.

even if someone combed through every single death record, looked at everyone's social media on the wayback machine, talked to people that they knew, they would have at best a vague guess of what some of those individuals' identities might have been.

that's why the researchers in this study put it out there, to say hey, maybe we should make it easier for trans people to be remembered the way that they want to be:

Typically, the funeral home director adds the final information about the decedent's sex in the official death certificate, based on informant interview (legal next of kin in Oregon). However, the next of kin may not be aware of or disagree with the decedent's preferred gender or sex. This phenomenon can lead to what has been termed nonconsensual de-transitioning—a type of gender bias after death.

In this study, we used narrative information from the Clackamas, Multnomah, and Washington County Medical Examiner case investigations between 2011 and 2021 to determine and describe transgender status of decedents. To further investigate the concordance between transgender status and the official death certificate sex, we matched data between the ME system and the death certificate. We expected poor concordance because gender information is not a required element of the death certificate. However, reporting on this discrepancy is an important step in determining whether individuals are being nonconsensually detransitioned.

(bolding still mine)

got that? that this study was written by some very nice people who want trans people to get to have a legacy that they choose? and who expected the data to be transphobic, but put it in writing it and published it in hopes that maybe it will help trans people in the future?

okay cool. moving on.

here are the death stats:

We identified 51 deaths from the ME narrative and/or suicide form in the Portland metro area between January 1, 2011, and September 30, 2021, in transgender persons (Table 2). More deaths were identified between 2020 and 2021 (n = 27) than all the other years combined (n = 24). The most frequent gender reported was transgender female (36/51, 71%), followed by transgender male (11/51, 21%). Suicide was the most frequent manner of death (36/51, 69%), followed by accident (7/51, 14%) and natural death (7/51, 14%).

so 51 deaths: 36 trans women, 11 trans men, and 4 nonbinary people.

(the above paragraph I quoted doesn't mention that the remaining four were nonbinary, but the rest of the study does; you can check if you don't trust me)

(also fuck, it's bleak that so much of their data came from just 2020 and 2021. I'm hoping that that means an increase in people being able to be identified as transgender and not more trans deaths, but this is yet another thing I can't extrapolate from a sample size this small)

Table 3 presents the results of matching the ME reports to vital statistics death certificate data (n = 47). Excluding the death certificates marked with “unknown” sex (n = 4), there were 47 deaths to compare. The κ coefficient for this table was −0.0657, indicating poor concordance.

the study says that three trans women and one trans man had death certificates that were marked as "unknown" when it came to sex. the researchers excluded these from the study, dropping our number of people to 47.

"why exclude those people from the study?" you may be asking, and there is an answer for you:

We excluded 4 deaths marked as unknown sex in vital records for a total of 47 deaths to compare, because unknown in vital records is reserved for cases when sex cannot be determined after verification with medical records, inspection of the body, or other sources and does not correspond with any ME category.

(I'm not a scientist but as far is I understand it, because they're specifically looking for cases where ME data and death certificates match, the ME not having a category for "unknown" would mess with their study.)

here's table 3, and now we can see that on their death certificates, 13/33 trans women were gendered correctly, 5/10 trans men were gendered correctly, and 0/4 nonbinary people were.

also good to note that in Oregon, they do have X as an option on their death certificates, though as of 2022 it seems like maybe nobody was using it:

The state of Oregon did add a new category of “X” to the death certificate in April 2018 to indicate nonbinary. “X” is inclusive of individuals whose gender identity is not exclusively male or female; however, “X” is not descriptive in terms of the individual's transgender status. Use of this category “should” match the sex on other legal documents, especially the birth record. Each epidemiologist checked his or her jurisdiction's vital records for “X” and did not find any records (J. Walters, MPH, M. Mew, MPH, and K. Repp, PhD, MPH, unpublished data, 2022). It is unclear whether “X” is not used because it is unclear, uncommon, or not considered respectful of a person's gender identity.

and, the study mentions that while people can also have X put on their legal documents in Oregon (which could make it more likely for X to end up on their death certificate), there are still things that make it hard for a lot of people to change their gender information:

Changing a driver's license or ID card requires making an appointment at the Department of Motor Vehicles office, requesting a new card, meeting the requirements for issuance of a new card including proof of address, getting a picture taken, and paying a $40 to $70 fee. Obtaining a new birth certificate with a corrected sex requires an application signed by the applicant in front of a notary, a completed order form, a $35 fee, and a photocopy of the applicant's ID. Because of these lengthy processes, it is likely that only a small proportion of transgender persons have all their official identification matching their preferred gender. A person may go through life relatively unencumbered by mismatched identification but, upon his or her death, the inconsistencies cause gaps in data that obscure death disparities among gender minorities.

so what's the result of all of this?

To our knowledge, this is the first report to assess concordance of sex on death certificates and gender identified through death investigations for transgender individuals. Our findings demonstrate the significant implications for transgender individuals in a death system in which they are systematically unable to be recognized as their chosen identity after death in nearly every part of the death process.

shit sucks. for all of us. that's the result.

I don't know that I agree with all of the writers' suggested options for trying to fix this problem (but I'm not a medical examiner so I'm no expert), but the results of this study really are: in the Portland metro area from 2011-2021, it seems like trans people had a pretty high chance of being misgendered on their death certificates.

now.

if you're going to say, "yeah, and trans women have it worse than trans men! they're definitely less likely to be gendered properly!" I will say sure, in this study of 47 entire people from an area with an estimated 9500+ trans people, sure, trans women had it worse than trans men.

and nonbinary people had it worst of all in this study, with absolutely none of them being gendered correctly upon death! which is deeply sad for them, though I was honestly glad to find out Oregon even has an X option on death certificates, as I'm betting a lot of states don't.

but if you read this post, and look at this study, and for some reason decide that you're going to use it to try to argue with other queer people on the internet instead of like...idk, telling your next of kin to help you preserve your legacy when you die? calling your representatives about whatever censorship bullshit the government is up to? organizing in your community? or even like, watching Speed Racer for the 12th time and playing minecraft? then I cannot help you.

there are better ways for all of us to spend our time than starting shit (or even continuing shit!) with people who largely believe in our beliefs. if you disagree with other queer people, that's awesome because all people disagree on things all the time! but in an era where trans people have to keep a Do Not Travel list of states that are unsafe for us to even visit, we need to learn to stop doing the facebook boomer mom thing™ and actually read the data when we see posts like this.

like ideally we shouldn't start our days ready to pull out, "yeah but I'm more likely to have my memory desecrated by the government!" as a gotcha to make ourselves feel better, especially not when referencing a study of 47 people from Portland who had potentially-violent deaths over the course of a fucking decade.

TL;DR: This study was about 47 trans people from the Portland metro area who died in potentially violent ways from 2011-2021. The point of the study was to point out that it sucks that trans people can't always have the legacy that they want, or have their legal documents match their identity.

The results show us some solid examples of people being misgendered on death certificates, but teach us absolutely fuck all about which specific people are more likely to experience this, because trans people across the board are misgendered and erased so constantly that we can never know what the actual numbers are. This is something the study itself pointed out multiple times.