Preventing domestic homicide

Prompted by a piece of research (which I now cannot track down - help!**) which showed (I think) that many people carrying out domestic homicides were not on any kind of high risk register, I have just written to the FoI departments of all the UK domestic police services with the following request: 

I am now carrying out a small piece of research into the preventability of domestic homicide. The numbers of women being killed by former or current partners seems to be a constant average of two per week. Based on some scant data, I have formed the hypothesis that many of the men committing these very serious crimes are not known to the police before hand – or on any kind of ‘watch’ list / risk register. And I need your help to find out if there is any basis to my hypothesis. Consequently, I would be most grateful if you could answer the following questions:

1. How many domestic homicides has your force dealt with over the last five years? A total figure is fine, or if you wish, break it down year by year. In this context, I am defining a ‘domestic homicide’ as being when a person is found guilty of killing a partner/former – murder or manslaughter. I am excluding other tragic cases of where parents kill themselves and/or their children etc.

2. How many of these persons (the ones convicted of killing) had any kind of prior criminal record ranging from low level crime to more serious acts of violence, cautions to imprisonment etc?

3. And how many of these people were on any kind of watch list / high risk assessment register / or similar?

Please note that if you cannot answer one of these questions (due to a FoI exclusion etc), please do still answer the other questions that you can. I would greatly appreciate avoiding the email tennis of you writing back to me to say that since Q2 (say) cannot be answered, you cannot answer Q1 & Q3 unless I want to alter my inquiry etc. Many thanks. And please forgive the round robin email: it is just so much quicker for me.

Any feedback on refining these questions would be much appreciated, if you think that would help me investigate the hypothesis more effectively. 

Thank you for all your help.

I will keep you posted on the responses.

Any feedback to me? 

**UPDATE 1321 | 250414: My plea has been answered by Peter Neyroud (thank you!). Here is a link to power point slides (NB they will download on clicking) of research by CC Sara Thornton: Does prior history of domestic violence predict domestic murder or other serious assaults? Very helpful piece of research which deserves to be read and acted upon.