Five years after more than 11,000 unprocessed rape kits - some dating back to the 1980s - were discovered in a police warehouse in Detroit, cities across the US are working to eliminate their own huge backlogs of untested evidence. Thousands of sexual assault cases are beginning to be resolved.
When Detroit prosecutors and state police toured a large storage warehouse in 2009, they made a startling discovery - more a huge cache of untested rape kits, each representing a report to the police and a lengthy hospital visit to collect evidence.
Armed with a large grant by the National Institute of Justice, prosecutors and Detroit police have now completed testing of 2,000 of those kits and are in the process of testing another 8,000.
The testing, as of October, had produced more than 750 DNA matches to a national database managed by the FBI known as Codis.
Investigations of these matches continue, but so far the Wayne County prosecutor's office - which includes Detroit - has produced warrants for 23 alleged rapists and convicted 14 of them, with three awaiting trial.
The office, run by the county's top prosecutor, Kym Worthy, has also identified 188 serial rapists from the processed kits who have committed crimes in 27 other states.
Detroit is not alone, however. Amid a reinvigorated call to test the estimated hundreds of thousands of rape kits in police storage across America, other US cities are also seeing dramatic results - a high number of previously unidentified serial rapists and dozens of unsolved cases going to prosecution.
In Cleveland, which has submitted all of its 4,300 kits for testing, police have opened more than 1,800 investigations, with more than 1,000 still in progress. The local prosecutor's officer has indicted 231 people, a third of whom had at least one previous rape conviction.
More than 50 people have been arrested so far from ongoing investigations after officials in Houston completed testing 6,600 kits.
Memphis, which is only part way through its more than 12,000 backlogged kits, has opened 243 investigations, indicting 36 people.
And a new effort in Colorado saw 24 Codis matches from the first 150 kits tested, according to local media.
Such results suggest other cities and states may have thousands of new cases sitting on their shelves.
No 'rational reason'
A push for testing backlogged rape kits began with New York City, which cleared 17,000 kits between 1999 and 2003.
But successful rape kit processing in New York, and later Los Angeles, did not spark a nationwide movement, despite improved DNA tests and millions of offender profiles to search against in Codis.
"You have capacity and ability, and yet jurisdictions were not sending their rape kits in for testing," says Sarah Tofte of the Joyful Heart Foundation, which advocates for rape kit backlog testing.
Tofte says many police departments don't see the value in testing old kits when resources are often tight.
She says backlog results are "very sobering" for cities that considered these cases not worth testing.
"It's really shown many jurisdictions how many initial assessments they made were not the correct ones.
"It always surprises me a little bit when they're always so sure exactly what is in that rape kit backlog," she says.
Many police departments across the country do not test kits in non-stranger rape cases where the victim and suspect were known to each other and consent - not identifying the suspect - is the main issue.
Rape kits were also not sent for testing when victims who made the report decided not to pursue prosecution or could not be found again.
But some of these cases show very little investigation on the part of police.
"It wasn't like there was a rational reason to not pursue," says Dr Rebecca Campbell, a researcher at Michigan State University who studied the results and case files from 1,600 Detroit kits.
Looking back at those police files, Campbell says, there was a "widespread" attitude of not believing the victim, and only doing minimum investigation on the cases, many just including the initial police report.
Tofte says police and prosecutors across the US have "struggled to adopt a really comprehensive approach to rape investigations", including failing to do a thorough investigation on the suspect instead of simply scrutinising the victim's report.
In one case from the Detroit backlog, a 12-year-old boy had told police in 2005 he had been sexually assaulted by a man who offered him money to clean out his basement.
The boy gave a detailed description of the man who attacked him and pointed out the man's house to police as he went to the hospital.
But the case never went forward, closed as "victim refused to prosecute" after a few attempts at contacting the boy and his family on the phone failed.
The kit was not sent for testing until seven years later. The DNA matched Billy Bingham, a convicted sex offender living at the address the boy had pointed out.
The boy, now a young man, testified at the trial, and Bingham was sentenced to at least 40 years in prison.
Even if attitudes about the value of rape kits change, however, the money may not be there.
Sgt Amy Millis, who runs Dallas's sex assault crimes unit, says her city had a desire to test their backlog but had been stalled until now because of funding.
"This is an extremely expensive endeavour," Millis says, citing not only the costs of testing the kits but "additional investigators, additional court personnel, prosecutors … advocates and counsellors to help people who are reopening old wounds".
As Millis suggests, a successful match is only the beginning.
"The case still needs to be fully investigated like it should have been fully investigated when this first happened," Wayne County prosecutor Worthy says. "You put a case together the old-fashioned way."
Part of that process is contacting the victim - in person - and telling them there is new information in their case. Investigators need to speak to the victim to make sure any DNA sample found is not from a consensual partner. Then they ask if they want to move forward with a prosecution.
Sometimes, prosecutors make "John Doe" indictments when they find DNA matches across multiple cases but the identity of the person is not known, allowing them to pursue the cases if they enter Codis separately.
Cleveland has made dozens of these indictments based off their own backlog. Twenty-two of Memphis' 36 indictments are John Does.
"For me the worst case scenario is we run the profile of a number of these kits and we find out we got a serial rapist … but he hasn't been caught ever," says Tammy Kemp, a Dallas prosecutor who will be in charge of any Dallas backlog cases.
"So we'll have a number of women who may have been raped by the same person, but we still don't know what his identity is."
In the wake of these results, laws about how sexual assault kits are treated are changing.
As of next year, Michigan will join California, Texas and Illinois as states that require testing of every single sexual assault kit, no matter the circumstances.
Results from Cleveland's backlog
As of mid-November, 2,985 tests out of a total of 4,400 have been completed
- Total investigations assigned: 1,859
- In progress: 1,162
- Sent to prosecution: 252
- Cases previously resolved in court: 286
- DNA was from consensual partner: 42
- Suspect's death: 25
- Insufficient evidence: 48
- Statute of limitations expired: 12
- John Doe or unknown male: 89
- Dismissed by court or prosecutor: 14
- Guilty by jury: 16
- Hung jury: 1
- Not guilty: 7
- Guilty plea accepted: 44
Funding for rape kit backlog testing has been stuck in Congress for over a year, but last month New York City prosecutor Cyrus Vance Jr announced a new $35m (£22m) fund - taken from a forfeiture payment in a financial case - for counting and testing kits.
"I think it's going to be interesting to see what comes out of the woodwork - when they are offered a very concrete opportunity for help," Tofte says.
But new funds for rape kit testing or mandating new kits to be tested will not improve failures of investigation, she adds, saying the backlog represents cases where police do not "really comprehensively investigate the sexual assault cases that are reported to them".
"You're not going to be able to resolve that by just requiring them to test everything."
Worthy believes cities that are just starting to come to grips with the size of their backlog and whether to test all kits should not shy away from the challenge.
"Put your issue out there and own it." she says. "[Don't] make excuses saying it didn't happen on my watch or these are old … because those can still yield very valuable evidence… Own it, and bring justice to these victims."
Texas and the statute of limitations
Because of the limited funds, many cities have prioritised cases at risk of passing out of the statute of limitations, but Dallas is starting with the most recent "old" cases - from 2011 - in part because Texas laws will not prevent them from "losing" any cases after 1996. The belief is that victims will simply be easier to find and cases easier to investigate the more recent the crime.
But Dallas is not yet testing rape kits outside the statute of limitations for cases with DNA evidence.
Whether to test rape kits too old to be brought to trial is a challenging decision. Tofte and Worthy are among those who believe it is worth it. Positive DNA matches on such cases may help in prosecution of other cases, especially if the suspect in a current sexual assault case can now be linked to a previous crime, even if they will never be charged separately.
Positive matches can also be used as factors in parole board hearings if the suspect is already in prison.
But given the use of additional resources, there is also the risk that police find out who a suspect might be but cannot do anything about it. Or regardless of a statute of limitations, a DNA profile is not associated with a specific person.