By Catarina Demony and Sam Tabahriti
LONDON (Reuters) -A month before she was stabbed to death by her ex-boyfriend, Bethany Fields walked into a police station in northern England to report his abusive and controlling behaviour. He had threatened to kill her, but she was not assessed as high risk.
Fields is among many domestic abuse victims failed by risk assessments based on a form known as DASH used by Britain's overstretched police forces, social workers and others for more than 15 years, according
to two academic studies, several women's charities and victims' relatives.
"To get that form right literally means the difference between life and death," said Bethany's mother, Pauline Jones.
The charity which co-developed the form, SafeLives, has now been tasked by the government with a project which its CEO Ellen Miller described as examining the way to see "how a review could work through, how a bigger rewrite could happen".
"We know so much more now, that it needs to evolve, it needs to change," Miller said, confirming a project that is yet to be publicly announced.
DASH had saved many lives, she said, adding that it should be rewritten rather than replaced and that it was up to police to use it properly.
"The problem is not the DASH. The problem is police officers' values and behaviours," she said.
A report by the Independent Office for Police Conduct (IOPC) into Fields' murder, published online in April this year, said the officer conducting the assessment was inexperienced to undertake such a sensitive assignment and lacked supervision; it also questioned the DASH form.
Those filling in the DASH (Domestic Abuse, Stalking and Honour-Based Violence Assessment) tick: "yes", "no", or "don't know" - sometimes labelled "other" - in response to questions about possible abuse.
The IOPC said the third option reduced the chances of officers pinning down a potential yes. That in turn reduces the chances of getting enough ticks for the assessment of high risk that triggers a referral for extra support.
Marsha Scott, chief executive of Scottish Women's Aid, said the DASH tool was "deeply flawed".
Researchers from the Universities of Manchester and Seville found the DASH "performs poorly at identifying high-risk victims", with 96.3% of such cases being wrongly assessed as standard (low) or medium risk in their study of 350,000 incidents logged by an unnamed major UK police force.
SafeLives did not respond to a request for comment on the statistic, published in Madrid-based journal Psychosocial Intervention in 2022.
The British system is not the only one to come under scrutiny. Spain's interior ministry said in January it had updated its gender violence management system to make it more effective; in June, it said British officials had visited to find out more.
Britain's interior ministry did not respond to a request for comment for this article. It has said reducing violence against women and girls is a priority and that it plans to publish a new strategy on the issue.
MISTAKES
In the year ending March 2024, there were 108 domestic homicides, of which 83 of the victims were female.
Reuters was not able to establish how many of those deaths followed DASH risk assessments or what those had concluded but found other deaths in which an assessment had been undertaken.
Seventeen weeks pregnant, Fawziyah Javed was pushed to her death from the top of the Arthur's Seat hill in Edinburgh in 2021. Her mother, Yasmin Javed, told Reuters her daughter had reported her husband's violent and controlling behaviour to the police.
Officers told her she was medium risk and despite the risk factor of being pregnant and a visit from police six days before she died, she was never referred for high risk support, her mother said, adding: "These mistakes are costing lives."
Contacted for comment, police said only that the IOPC is still investigating the force's response.
Both Javed's husband and Fields' ex-partner were convicted of killing them.
When asked about fatalities after DASH assessments, Miller said: "I think the DASH could always be better," adding that it was up to police to safeguard people from the risk of death in their recruitment and training.
Her hope, she said, was that a "refreshed DASH" would be used adequately in every case.
OTHER OPTIONS
The interior ministry has asked other charities to help SafeLives "bring together insights" into the systems for assessing and managing risk in the country, a letter from the ministry to one of them who asked to remain anonymous showed.
Several charities contacted by Reuters said they feared the project would not go far enough.
Ngozi Fulani, CEO of Sistah Space, a charity supporting victims from African and Caribbean heritage, said the DASH may not be effective for many in those communities, due to mistrust of police linked to institutional racism.
"They chose to leave us out, that speaks for itself," she said, referring to the fact they have so far not been asked to contribute to the project.
The Spanish police and some organisations are using new technology to assess future risk: Berlin-based startup Frontline has a machine learning risk assessment and British data scientist and former police officer Tori Olphin has created an algorithmic model to predict future harm, for example.
The College of Policing, a professional body for police in England and Wales, has developed its own tool, DARA, with Cardiff University Professor Amanda Robinson, to address what she says are some of DASH's flaws, particularly around coercive control, which only became a criminal offence there in 2015.
A police representative said they would assess all the available tools.
"Police officers must be supported with the right training and tools to identify offences and protect victims," said Assistant Chief Constable Claire Bell, Deputy Director of the National Centre for Violence Against Women and Girls and Public Protection.
(Reporting by Catarina Demony and Sam Tabahriti; Editing by Philippa Fletcher)
