Can Keller Conviction Stand Without Physical Evidence?

When a child is sexually abused, there isn't necessarily physical evidence to prove any abuse happened. But whether the lack of any physical evidence of abuse is significant in the case of Fran Keller, convicted of sexually assaulting a child left in her care back in 1991, is a central issue in an appeal of that case currently pending in a Travis County court.

Fran Keller was sentenced to 48 years in prison for allegedly sexually abusing a 3-year-old girl, Christy Chaviers, who in 1991 was an infrequent drop-in at the home-based daycare Fran and her husband Dan Keller ran out of their Oak Hill home. (Dan was also convicted and sentenced to 48 years. The current appeal has been filed on Fran's behalf, but will impact Dan's conviction as well.) After a day in care at the Kellers' home that summer Chrsity told her mother, Suzanne Stratton, that Dan Keller spanked her. That allegation quickly morphed into an allegation of sexual abuse – and then again, into wilder allegations of abuse perpetrated by the middle-aged couple on Chaviers and two other children who were also drop-in clients and whose parents were friendly with Stratton. By the fall, the allegations turned fantastical: The Kellers had taken the kids on plane rides to Mexico, subjected them to satanic bone-replacing rituals, killed animals and babies – allegations that led Christy's therapist, then Donna David Campbell, to conclude Christy had been a victim of "ritual abuse."

The Kellers were among hundreds of daycare workers across the country accused in the Eighties and early Nineties as being part of satanic cults that abused children taken by their parents to daycare facilities. In 2008, the Chronicle began a reinvestigation of the case against the Kellers and found that there was, in fact, scant evidence that any children had been abused by the Kellers, let alone any evidence of satanic ritual abuse.

Indeed, the only physical evidence that suggested Christy had been abused came in the form of testimony from a then-novice Brackenridge emergency room doctor, Michael Mouw, who examined Christy in August 1991. In the fall of 1992, Mouw testified in court that in examining Christy he found deformities to her hymen and to her posterior fourchette – a tender fold of skin on the rear of the vagina near the anus. Those deformities – described as possible lacerations to the hymen and a tear of the fourchette – could be signs of sexual abuse, he said. That determination was confirmed by pediatrician Beth Nauert, a local expert in child sexual abuse who expounded on Mouw's findings in her trial testimony. Notably, Nauert examined Christy some two weeks after Mouw and found no signs of any deformities.

When he was contact by us for what eventually became our 2009 cover story on the case ("Believing the Children," March 27, 2009), Mouw said that not long after he testified at the Kellers' 1992 hearing he realized that what he thought were injuries were in fact "normal variants" of female genitalia. "At the time, in good faith, I saw something [I thought] was abnormal" about Christy's hymen, he told us for that story. But with more experience and education, that is not what he now concludes.

Mouw's about-face, meaning he provided during the 1992 trial what he now knows to be erroneous medical testimony, is among the claims for relief included in an exhaustive writ filed in January on Fran Keller's behalf by Austin defense attorney Keith Hampton. Put simply, Mouw's more erudite consideration of the case has not only eliminated the only piece of physical evidence to suggest that Kellers abused Christy, but also the only concrete evidence that any crime ever happened.

Mouw's changed opinion took center stage at a hearing Tuesday afternoon before retired District Judge Wilford Flowers, who as an elected judge presided over the Kellers' original trial. Under questioning by Hampton, Mouw testified that he had only been practicing medicine for some four years before examining Christy and had little experience, if any formal education at all, in conducting sexual abuse examinations of children by the time she was brought to the Brackenridge ER by her mother. It was a "chart I picked up randomly in the middle of a busy ER shift at Brackenridge," he said. That night, and with an institutional bias toward nabbing bad guys, Mouw examined Christy and found what he characterized as injuries to her genitalia.

It wasn't until several years after that, and after he'd testified about the alleged injuries in court, that he attended a presentation put on by Nauert that detailed normal variations of female genitalia. During that presentation, he said, Nauert showed a slide of normal genitalia that looked exactly like what he'd seen when he'd examined Christy. "Kind of a light bulb went off and I was like, 'that's just like what I saw'" in Christy, he recalled Tuesday afternoon. Is there any doubt in your mind today, that what you saw in 1991 wasn't an injury, but was in fact a normal variant, Hampton asked of Mouw.

"No doubt," he said.

Nonetheless, prosecutor Scott Taliaferro dug into Mouw, asking about what he'd told us in 2009 – seemingly suggesting that Mouw's regret and subsequent action to correct his testimony was prompted by our article, which Taliaferro characterized as not reflecting every bit of testimony from the Kellers' trial and including information that had not been provided to jurors in the case. Mouw choked up when talking about how he regretted not pressing the issue sooner – it was Nauert's presentation that first prompted his doubts, which were compounded by a subsequent article written by Gary Cartwright for Texas Monthly, and then later reinforced by his reading of our article, he said. Under subsequent questioning by Hampton, Mouw recalled that after the Nauert presentation he did reach out to Austin Police with his concerns (likely to then-Detective Larry Oliver, who was the department's lead investigator in the Keller case; Oliver is now retired). In response, Mouw testified that he was told he "didn't need to worry about it because there was stuff he couldn't tell me about...but he was convinced they were guilty."

Nauert took the stand after Mouw at trial, serving as an expert witness to say what Mouw said he found back in 1991 would be consistent with sexual abuse – and she again took the stand following Mouw at this week's hearing to say that at the time she examined Christy back in 1991 there were no signs of trauma on the girl's genitalia. But that doesn't mean she wasn't abused, Nauert said.

Nonetheless – apparent in their questioning and in her responses – the state, and Nauert, concede that there were no injuries to Christy's hymen – and thus, that Mouw's correction to his initial findings in that regard are sound. But prosecutor Elizabeth Phillips questioned Nauert about whether a "fissure" or "tear" on Christy's posterior fourchette could be caused by something other than abuse. It could be, Nauert said, but would be unlikely. A tear to that area would be painful, she said. Under questioning by Hampton, Nauert conceded that Mouw's exam report did not indicate that Christy complained of any pain in that area.

In other words, the state appears inclined to agree that Christy's hymen was normal – and that Mouw was indeed mistaken about what he thought he saw in 1991 – but that an injury to the fourchette would still serve as evidence of sexual abuse. But Nauert said she found no indication of injury to that area when he examined Christy, and Mouw testified during the hearing that he was likely wrong about seeing any injury to that area as well.

Mouw testified that he simply was not trained to examine children for sexual abuse and that the science of doing so has evolved significantly in the intervening 22 years. Moreover, he said he did not have access to a culposcope in order to examine Christy back then, a tool that likely would make moot contemporary questions about what he saw, or didn't see, during that initial examination. A culposcope is a magnifying device that helps doctors to examine the vagina – and takes magnified pictures of examined female genitalia. Under questioning by the state, Nauert said that while she once used the culposcope in her examinations, she doesn't anymore. She sees 200-250 children each year for "evaluation for potential sexual abuse," she said, but apparently doesn't use the culposcope to take documentary pictures of what she finds. Instead, she said she uses her own eyes, or is aided by an otoscope (commonly used for looking in a person's ears) to magnify what she's looking at.

Regardless whether Nauert uses a culposcope, Mouw testified that, as a new doctor unfamiliar with such exams, had he used one back in 1991, the questions being asked now would be readily answered by photographic evidence. "The main issue for me on this case [is if I] had a culposcope [we] would've had photographs" of Christy's genitalia, which would be "much more reliable forensic" evidence, he said. "It would've aided me."

After a nearly three-hour hearing, the case was left pending.