Some cities experiencing high numbers of unsubstantiated rape reports have also been troubled by the statistics and changed reporting practices to avoid wiping reports off the books.

Sgt. Larry Scirotto said that when he took over the Pittsburgh Police Department's sex offense unit two years ago, many cases in which the victim recanted or didn't want to move forward were being marked unfounded, meaning the incident did not happen.

Scirotto said that was the wrong approach in a city that for years was, along with Baltimore, leading the nation in the percentage of police reports labeled unfounded.

He changed the procedures so those cases would be tagged "cleared by exception," a designation that keeps the incident counted among the city's crime totals but allows detectives to focus resources on other cases. The case can be revisited if the victim decides – whether through counseling or a change of heart – to pursue charges later.

"When you classify a crime as unfounded, it says you didn't believe the victim, or that we determined a crime didn't occur to begin with," Scirotto said. "I'm not concerned with the statistics. I'm concerned about being able to prosecute at a later time."

Interviews with advocates and victims, and a review of reports requested under a Public Information Act request, reveal an attitude of distrust by police toward victims and a reflex to dismiss rape reports both in the field and after investigating.

Lauren — who did not want her last name used to protect her privacy — believed that she might have been sexually assaulted last year after a night of drinking with friends. She drove to Mercy Hospital to get examined and reluctantly agreed to speak to a detective.

The initial officer listened and was comforting, she said, but a detective from the sex crimes unit immediately started an interrogation.

"He was lecturing me on the justice system and was giving me lectures about drinking," she said. "He was also questioning me about, 'Do you have a boyfriend? Does your boyfriend know about this? Sometimes people make up this stuff because they made a mistake. Just because you didn't remember everything, doesn't mean you didn't want something to happen.'"

No report was taken to document Lauren's concerns. Years ago, that might have happened, too, in Louisville, another city that once rivaled Baltimore's rape statistics. But now, Louisville investigators say, such a case would result in a police report and a classification of "unsubstantiated," which allows detectives to focus resources elsewhere.

Sgt. Andy Abbott said the Kentucky department has recently been using the in-house classification of "unsubstantiated" to keep cases open but put them on the back burner when necessary. Because of that change and others, the percentage of Louisville's unfounded cases has dropped and the number of reported rapes has increased by 17 percent.

"Unsubstantiated means there's a possibility a crime may have occurred, but we don't have enough to prove one way or the other," Abbott said.

Police: Investigations improve

For years, Baltimore's percentage of rape allegations labeled as unfounded was in line with the national average. But by 1998, the rate had doubled from its longtime average. Concern arose in 2003, and the Baltimore Police Department undertook an audit that found it had under-reported rapes by 15 percent.

By the next year, when Commissioner Leonard D. Hamm took over the department, the percentage of unfounded reports had doubled again and increased to a high of 37.6 in 2006. It hasn't fallen below 29 percent since, even as reported rapes continue to fall. Police could not explain the increase.

A review of FBI data from across the country shows significant disparities in rates of unfounded rapes. Though most have a percentage in the single digits, some cities, including New York and Cleveland, report zero – a number that experts say is just as eyebrow-raising as Baltimore's high rate. The FBI does little to monitor the accuracy of reporting.

"Why is it that women in the greater Baltimore area are more disposed to lying about sexual assault than anyplace else in America?" asked Branson of Turn Around. "Is it in the water? What exactly would make us the ones most likely to tell a story about being sexually assaulted?"

Uzarowski, who spent 35 years with the Police Department, said sex crimes investigations have come a long way. When he started, there was no specialized unit.

Officers scrutinize data more closely and with greater urgency, he said. A commander retrieves sex offense calls and checks to make sure they have been assigned to detectives. If not, "that's where we backtrack and find out whether there was no one on the scene, or was not anything suspicious, or a vacant building or a false call," Uzarowski said.

He and Lt. Jon Foster both said they were pleased at the results of a recent audit of last year's numbers.