The part nobody says out loud

Separation is one of the markers that turns up most consistently around the worst outcomes, and a course that told you to leave without saying so would be doing you harm.

Here is what was actually measured, because the short version you will have heard is bigger than the evidence behind it. The largest study of its kind compared women who were killed by a partner with women who were being abused by a partner and were not killed. Everybody in it was living with harm. Among them, having separated from him after living together carried roughly three and a half times the odds of being in the group that was killed, and having ever left or asked him to leave, roughly three times. Where he was also highly controlling, separation carried about nine times.

Now read what that compares. It compares abused women who had left with abused women who had not. It does not compare leaving with being safe, because nobody in that study was safe. It cannot tell you that staying is the lower risk, and it was never built to.

So this is not an argument for staying. Nobody here is telling you to stay, and staying carries its own risk, which does not go away by being familiar. It is an argument for one thing only, which is that if you are going to go, the preparation happens before, and it happens with somebody who does this work, and not on the night.

It is also worth knowing which markers weigh heaviest, because almost all of them are about him rather than about anything you decide. A review pooling seventeen studies put access to a gun, pressure on the neck before, forced sex, threats with a weapon, controlling behaviour, and threats to harm you at the top of the list. Whether you have left is on that list. It is not the top of it.

If you have already left, or he knows you are thinking about it, that is not a reason to panic. It is the reason to make contact with a service this week rather than next month.