A useful figure on the rate of homicide by "patients" in the UK

From P34 of National Confidential Inquiry into Suicide and Homicide by People with Mental Illness Annual Report July 2009
http://www.medicine.manchester.ac.uk/psychiatry/research/suicide/prevention/nci/inquiryannualreports/AnnualReportJuly2009.pdf

"During 1997-2005, 510 people convicted of homicides (10% of all those convicted) were identified as patient homicides, i.e. the person had been in contact with mental health services in the 12 months prior to the offence."

The figure is for convictions of homicide and I am unsure if that covers manslaughter, diminished responsibility and not guilty by reason of insanity. It also only covers people who have been in touch with services within the year.

P 33 has information on the number of people in psychotic or abnormal states however I'm too lazy today to analyse the data on the page.

There is a figure I'm hunting to find a reference for
"95% of homicides are committed by people who have not been diagnosed with a mental health problem"

I can't find the source and the high quality sources indicates the figure would be higher.

The chance of being killed by someone experiencing the symptoms of a mental illness is also about 10%. (http://www.mind.org.uk/help/research_and_policy/dangerousness_and_mental_health_the_facts#_edn5)

As with all these sort of things the percentages are meaningless in relation to the risk. The risk of being murdered is tiny though there is a high degree of public paranoia and hysteria about the risk of death.