Dozens of women murdered after police rely on ‘deeply flawed’ domestic violence tool

[IwantToRetire]

For years, police knew Dash was failing. In 2016, the College of Policing said Dash was unfit for front-line use.

Standing Together, a domestic violence charity, found that domestic homicide reviews showed victims had been killed after scoring an average of eight ticks, not the 14 needed for urgent support.

In 2019, Manchester University researchers concluded “officer risk predictions based on Dash are little better than random”. In 2022, academics from Manchester and Seville found Dash “performs poorly at identifying high-risk victims”, wrongly classifying more than 96 per cent as standard or medium risk.

But although problems with Dash were widely known among researchers in the criminal justice sector, bereaved families whose loved ones were wrongly graded “low” or “medium” – or had “high risk” status delayed – are only now discovering that it may explain why no one intervened.

Full article here https://www.telegraph.co.uk/news/2025/08/16/flawed-domestic-violence-tool-dash-checklist/

And here https://archive.is/K81Mn

Any risk assessment tool should be used as a guide with significant limitations. Professional judgement is a lot more valuable but has the supposed downside that it cannot be quantified or accurately measured.

Fucking hell. Those poor women.

I don’t understand why they didn’t review the tool after it became clear it was not working. I read an article on this and you could answer “yes” to “has he ever tried to strangle or choke you” and still be graded as low risk. Thats clearly ridiculous and it’s cost women their lives.

Oh that’s awful! That’s so sad.

Cynical, but my instinct is to ask whether it was calibrated to save money. What were the exact terms of the contract? What amendments were made over the years to the criteria supposedly catalysing urgent intervention? Why weren't these criteria adjusted downwards (fewer ticks needed) - and were they ever adjusted upwards (more ticks needed)? It makes me think of the omission of the PC of sex from the list of characteristics subject to "hate" crime, partially on the basis that including this would put unsustainable pressure on the police.

Police Chiefs' Council gave a direct order three years ago, that all police must stop using the system.That order is still being disobeyed by the majority. So what were the sanctions? None.

The police hate women equally as much as the male population.

I can’t get on board with that.

Equally concerning, or maybe more so, is to find the charity who designed this didn't do anything and are still promoting it.

https://safelives.org.uk/resources-for-professionals/dash-resources/

Bumping as

Typical technocratic solution to a problem that requires a tool for measuring risk to be used carefully i.e. with a pinch of salt.

I think at it's core, DASH would be a useful tool - but it has become a tick box exercise, where they go "less than 14 ticks, you're fine love". Even within the guidance of the checklist it says if things are escalating, or there has been a lot of police call outs then they should progress the case.

To me, there would be some things that should count 'double/treble' e.g. strangling.

You have to wonder why there's such a chasm of ignorance between widely known expert knowledge and what police forces and family courts understand of domestic abuse in all its manifestations.

Do they simply not understand plain English? Or is there willful disbelief involved?

The point is, as I said upthread, the tool was created by so called DV support experts.

So in using it the police thought the "experts" had got it right.

So what is so terrible, and tragic, is that SafeLives are continuing to promote is as a tool, and apparently haven't made any changes.

That is the issue.

Way back in 2015 I tried to get my city to replace it, or at least supplement it with other tools, but they weren't keen. It's such an odd instrument that I did a very modest amount of digging into how it was developed and was horrified. Basically it's an ad hoc product of dodgy analysis of a flawed sample of DV perpetrators. It should never have got beyond pilot status.

At the time I was actually less bothered about the way it was used in risk assessment than about the fact that it was also being used to infer support needs. It took much longer than it should have done to get the council to recognise that risk didn't correlate with the impact of abuse or the support a victim would need to recover.

Do people not use their brains anymore?

Its this huge gap between someone having to be aware of an issue (ie the police) as part of their job, but not having to understand it.

And someone who thinks they are an expert saying that's okay I'll make a quick go to assessment.

This is one of the main flaws of any system like this. Whether a check list on a piece of paper, or an online form or an app.

Maybe there should be some sort of committment that even if you use a form or tech to gather information, an actual human being has to make an assessment.

I suspect there are many other standardised assessments that just do not do justice to the individuals evaluated in this way.

The trouble is that most of the time, mostt people make worse assessments than a properly researched and validated standard tool.

I reviewed some research on development of a standardised medical assessment tool for making decisions about referrals from primary care to a particular medical speciality. Most GPs made worse referral decisions when relying on their clinical judgement (false positives and false negatives).

The only exception was a small group of highly experienced GPs, whose outside-criteria referrals had a decent probability of being justified by subsequent investigation. The researchers were keen to try and uncover the combination of unrecorded cues these doctors were picking up on...

Britain's main domestic abuse risk-assessment questionnaire, DASH, is set to be updated by its lead author and leading charity SafeLives amid evidence its limitations may have contributed to deadly failures.
https://uk.news.yahoo.com/uks-domestic-abuse-risk-assessment-123011722.html

We are aware of the limitations of DASH when used by frontline officers. The College of Policing reviewed the use of DASH in 2015; this involved in-depth research which found inconsistency in the way that the DASH tool was applied by first responders and information recorded and the need for a greater focus on controlling and coercive behaviour. An extensive international literature review did not identify a better alternative.

The College of Policing sought to address these shortcomings with the creation of DARA. For instance, DARA uses open questions which have been shown to better get to the root of what is happening, particularly in higher risk cases involving coercive and controlling behaviour. DARA remains the recommended tool for frontline officers conducting primary risk assessments, and many forces use this tool, sometimes in conjunction with DASH as a secondary risk assessment tool.

While there is the potential for Artificial Intelligence-assisted predictive models to inform decisions around risk, either at the frontline or in subsequent offender management, the College does not believe existing models can yet replace the primary risk identification interview or wholly replace professional judgement-based assessment.

https://news.npcc.police.uk/releases/police-use-of-domestic-abuse-risk-assessment-tools

Home Office told domestic violence tool was endangering women years ago

At least 55 women were murdered after police relied on a flawed risk assessment that failed to identify them as being in immediate danger

Alongside a “quick and easy” questionnaire for first responders, Safe Lives proposed more “in-depth” tools for specialists.

The recommendations were set out in a report titled ‘Evolving The Risk Led Approach’, which was submitted to the Home Office and Ministry of Justice in late 2022. The then-chief executive met Conservative ministers in February 2023 to discuss the findings.

The recommendation to replace Dash would have been interited by Ms Phillips when she took on the safeguarding brief after Labour’s victory last July. Safe Lives told The Telegraph that it has continued to lobby the Government “ever since”.

But the warnings appear to have gone unheeded, and Safe Lives chose not to publish the report for fear of damaging public confidence. It only became public after a Freedom of Information request to the Home Office by this newspaper.

https://www.telegraph.co.uk/news/2025/11/15/home-office-told-broken-domestic-violence-hotline-endangering-women/

Can be read at https://archive.is/zmIWT