Right. So this is bothering me enough that I'm going to get into it.

***Content warning for entire thread***
I'm going to be talking about terminology and concepts surrounding sexual assault, forced/coerced intercourse, and some gender identity aspects. https://twitter.com/tcshields/status/1315677128013348866

I'm not an expert on the most appropriate use of these terms, legal or otherwise. I've relied on some sources to frame the issue from a medical perspective. This is one recommended to me by a forensic nurse: https://theconversation.com/whats-the-difference-between-sexual-abuse-sexual-assault-sexual-harassment-and-rape-88218

In my work as a reference medical librarian, I've had search requests on this and overlapping topics. I'm going to say up front that these aren't easy searches - because of terminology, as I'll get into, but also because of the emotional work involved. Again, content warning.

There's a lot of terminology around this topic, but similar to that article I'm going to focus on 3 of them as they apply to searching the biomedical literature: sexual harassment, sexual assault, and (to a lesser extent) rape. So let's look at Medical Subject Headings (MeSH).

Here's the tree for "Sexual Behavior"[Mesh] :
Sexual Behavior
- Coitus+
- Courtship
- Extramarital Relations
- HIV Serosorting
- Masturbation
- Safe Sex
- Sex Work
- Sexual Abstinence
- Sexual Harassment
- Sexuality +
- Unsafe Sex

That's where you'll find "Sexual Harassment"[Mesh], defined as:

"A form of discrimination in the workplace which violates the Civil Rights Act of 1964. Offensive sexual conduct may include unwelcome advances, comments, touching, ...

... questions about marital status and sex practices, etc. Both men and women may be aggressors or victims. (Slee and Slee, Health Care Terms, 2d ed, p.404). While civil rights legislation deals with sexual harassment in the workplace, the behavior is not restricted to this; ...

... it may take place outside the work environment: in schools and colleges, athletics, and other social milieus and activities."

Sexual harassment as a MeSH term was introduced in 1993. The Anita Hill / Clarence Thomas sexual harassment proceedings happened in 1991.

Wondering where sexual assault is in MeSH? Well... there's not a term, and it's not available as an entry term for a related concept, either. So how does that literature get indexed? (I'll come back to how PubMed handles the automatic term mapping of the term.)

Looks like mostly terms like "Sexual Harassment"[Mesh] and / or "Sex Offenses"[Mesh]. A fair bit get "Crime Victims"[Mesh].

For examples, see: PMIDs 32762635, 32712578, 32374684, 31764526

That's... not great. But I guess it could be worse?

Let's take a look at "Sex Offenses"[Mesh], which is under "Crime"[Mesh]::

Sex Offenses
- Child Abuse, Sexual
- Human Trafficking
- Rape

Something to notice: entry terms include sexual violence and sexual abuse, but not sexual assault. Ooof, that previous indexing...

Which brings me to "Rape"[Mesh], defined as "sexual intercourse without consent of the victim." And here's how MeSH handles sexual intercourse, which prefers the term "Coitus"[Mesh]:

Coitus
"The sexual union of a male and a female, a term used for human only."

The FBI defines rape as "The penetration, no matter how slight, of the vagina or anus with any body part or object, or oral penetration by a sex organ of another person, without the consent of the victim." https://www.justice.gov/archives/opa/blog/updated-definition-rape

As that article I cited upthread notes, the FBI updated the definition in 2012 to be more gender neutral. MeSH has a way to go to catch up to the US Department of Justice. There's a lot to unpack in the MeSH for that term, but I think that should be a whole other thread.

Going back to sexual assault... the index terms aren't great, but what about Automatic Term Mapping? Say you do a search, untagged. Well.

[ sexual assault ]

Does this (see screeshot):

First, it doesn't treat as a phrase because "sexual assault" is not an entry term for any relevant MeSH. It does "Sexual Behavior"[Mesh] with a dash of "Sexuality"[Mesh], and associated synonyms. It also does some synonyms for assault.

So... let's just call that problematic.

That's all the nuances of the terms and indexing and such.

Here's why it matters: Most folks (and I bet even some #medlibs) aren't going to delve into MeSH, play around with terminology, do the background on the legal side, etc. especially for this topic. For lots of reasons.

Unless you are going to do a comprehensive search on par with a systematic review one that might stand up to peer review and use a combination of MeSH and tagged terms, you are not going to get great results relying on ATM and "Best Match".

Add to that, I'd bet that some folks aren't going to want to spend a lot of time looking at the results and seeing those terms over and over in playing around with the search. As I said before, there's a lot of emotional work here, even if you are only glancing at results.

There's more than this - gender identity for one, the terminology of victim versus survivor for another - surrounding the medical literature on sexual assault, but I'm going to stop here. Hopefully it has been enlightening.

The takeaway: you can't rely on MeSH to do your work.