El Paso 'shaken baby' conviction latest capital case challenged under TX's junk science writ

Texas' junk science writ continues to impact high-profile capital cases, this time causing re-evaluation of faddish "shaken-baby" science, which in recent years has come under fire. An El Paso judge recommended a new trial for Rigoberto Avila, declaring scientific testimony against him in his case "false and misleading." Now the Court of Criminal Appeals must decide his fate. Reported the Texas Tribune:

“The new scientific evidence creates a compelling case for Mr. Avila’s innocence, and a judge has now found that the verdict against him rests on false and misleading testimony,” Avila’s attorneys, Cathryn Crawford and Rob Owen, said in an emailed statement. "After spending 17 years on death row – and facing four serious execution dates – for a crime he did not commit, Mr. Avila is anxious to present the reliable scientific evidence to a jury.” 

In August, podcast co-host Mandy Marzullo and I discussed how the junk-science writ has impacted death-penalty cases, and this story adds more context to that discussion.

The combination of Texas' Forensic Science Commission's work - along with the state's early adoption of the junk-science writ - has put the state at the bleeding edge of efforts to challenge faulty forensics. We have both a dedicated body charged with critiquing bad forensics and a legal means to challenge them that doesn't exist in other states. (Texas' near-term leadership on forensic reform was cemented after the Trump Administration nixed national USDOJ initiatives to update and improve modern forensics.)

In Texas, capital cases are the one sliver of indigent defendants whose appeals are all paid for by the state, meaning those defendants have access to attorneys to file a state habeas corpus writ. So it makes sense that many of the most high-profile, early uses of the junk-science writ would come in death-penalty cases. By contrast, plea the case to life without parole, and a defendant accused of the same crime with the same evidence would have no access to an attorney at the habeas corpus stage.

That's why the junk-science writ is being used disproportionately in capital cases. It's not that faulty evidence wasn't used to secure other convictions. It's that those other folks don't have access to attorneys or experts to challenge the false evidence.

The Daubert and Frye standards used by the courts have utterly failed to keep junk science out of Texas or American courtrooms on the front end. In the long run, Grits' hope is that the junk-science writ provides a back-door means to challenge shoddy forensics, and that, once discredited, courts will stop using that evidence going forward. That's what happened with scent lineups (using police dogs), and to me it looks like the most likely path for eliminating bad forensics in an era when judicial gatekeepers have utterly and profoundly failed us on the front end.

Arson, shaken baby forensics called into question

A couple of recent academic papers related to areas of disputed forensic science deserve readers' attention. Here are links to the papers and the abstracts:

• A Perfect Storm Brewing for Fire Investigators in Court

The genesis of this piece comes from a trend the authors have observed in three separate but related areas, which we believe are converging into a perfect storm for fire investigators. These are: 1) the ongoing movement by courts across the nation to scrutinize more closely the reliability of expert testimony, 2) a growing apprehension about wrongful convictions stemming from faulty forensic evidence and problems in fire investigations, culminating in the revolutionary report published by the National Academy of Sciences, and; 3) the continuing development of industry standards that are raising the bar for fire investigators. Part I describes each of these forces, and then Part II demonstrates how together they are creating a mounting pressure on fire investigation experts to defend their qualifications and the reliability of their opinions in court, particularly insofar as analyzing the fire scene and interpreting fire patterns is concerned.

• Flawed Convictions: 'Shaken Baby Syndrome' and the Inertia of Injustice: Introduction

The emergence of "Shaken Baby Syndrome" presents an object lesson in the dangers that lie at the intersection of science and criminal law. As often occurs in the context of scientific knowledge, understandings of SBS have evolved. We now know that the diagnostic triad — the three neurological symptoms once equated with guilt — does not itself prove beyond a reasonable doubt that an infant was abused nor that the last person with the baby was responsible for the baby’s condition. Nevertheless, our legal system has failed to absorb this new consensus. As a result, innocent parents and caregivers remain incarcerated and, perhaps more perplexingly, triad-based prosecutions continue even to this day.

This is the CONTENTS and INTRODUCTION to "Flawed Convictions: 'Shaken Baby Syndrome' and the Inertia of Injustice" (Oxford University Press, April 2014). "Flawed Convictions" surveys the scientific, cultural, and legal history of SBS from inception to formal dissolution, exposing extraordinary failings in the criminal justice system’s treatment of what is, in essence, a medical diagnosis of murder. The story of SBS highlights fundamental inadequacies in the legal response to science-dependent prosecution. "Flawed Convictions" proposes a restructuring of the law that confronts the uncertainty of scientific knowledge.

The piece on fire investigations includes an extensive discussion of debates in Texas around arson science. Via CrimProf Blog.

Dallas DA pledges to review convictions based on junk science

Assuming he wins a third term as Dallas District Attorney, which without opposition in the Democratic primary appears increasingly likely, Craig Watkins said recently he wants to expand the work of his Conviction Integrity Unit - which spearheaded his office's review of old DNA cases, leading to numerous exonerations - to include arson and shaken baby cases. Reported the Dallas Morning News ("Craig Watkins says he still has big plans for DA's office," Nov. 14):

Dallas County District Attorney Craig Watkins has gained a national reputation for spearheading prisoner exonerations.

As he prepares to seek a third term, Watkins said Thursday he wants to expand on that role and add a few others. ...

Watkins, a Democrat who was first elected in 2006, gained attention for using DNA tests to overturn convictions, and he said his office has a few more such cases pending.

When prosecutors finish with those next year, Watkins said, he wants his team to take another look at people convicted of arson and those accused of shaking their babies to death. Watkins said he has concerns about the science used in the prosecution of both types of cases.

“The science has changed. We need to revisit it,” Watkins said without elaborating.

With the passage of SB 344 by Whitmire/Turner, people convicted based on junk science now have a clear path to pursue habeas corpus writs to challenge their convictions, with old arson and shaken-baby cases high on the list of bad science likely to be challenged. It will be welcome news if Watkins takes leadership and gets out in front of those issues the way he did on DNA testing. The main difference will be that, until the Legislature changed the law in 2011 (SB 122 by Ellis), DAs could prevent DNA testing in old cases if they chose, just as Williamson County DA John Bradley thwarted testing in the Michael Morton case for many years simply by objecting. By contrast, the passage of SB 344 means junk science cases can now get back into court via habeas writs on their own, so Watkins and other District Attorneys will be forced to revisit them whether they want to or not. Still, it's refreshing to see a DA willing to seek out false convictions in junk science cases instead of reflexively fighting tooth and nail to keep them intact.

Shaky science, un-recorded confession form basis of El Paso habeas writ

Jordan Smith at the Austin Chronicle has an excellent new story out ("Science goes to court," Nov. 1) regarding a habeas writ challenging a child-murder conviction out of El Paso. The defendant is seeking relief based on Texas' new law, authored this year by state Sen. John Whitmire and Rep. Sylvester Turner, allowing relief when new scientific developments contradict expert testimony at trial. (In this interest of full disclosure, this was legislation your correspondent helped promote on behalf of my employers at the Innocence Project of Texas.)

As is likely to be the case in much of the litigation surrounding this new habeas provision, the debate centers on whether the defendant's writ relies on new science or evidence that should have been available at the time of trial - in this case biomechanical modeling that has debunked many of the underlying premises of "shaken baby syndrome" and other expert testimony about child trauma. Wrote Smith:

Biomechanics has long been relied on for injury prevention and repair – the development of air bags in vehicles, of helmets to protect football players' heads, of replacement hips and knees – but only recently has biomechanics been incorporated into a criminal law context, to describe with scientific certainty the force needed to cause specific injury. Biomechanical engineering is pivotal to the reconsideration of cases of so-called "shaken baby syndrome," or SBS, and has demonstrated that simple shaking by an adult cannot create the force necessary to kill a child.

"Many if not most non-physicians assume that physicians, skilled in the art of medicine, must have particular knowledge of injury mechanisms. This assumption is wrong," John Plunkett, a veteran pathologist and expert in the diagnosis of child injuries who has been a vocal critic of SBS, wrote in an affidavit filed along with a new appeal in Avila's case. For the most part, physicians, such as Raschbaum, "need not know or apply injury mechanics if they are responsible for diagnosis and treatment," Plunkett continued. "However, if a physician ventures from diagnosis and treatment to speculation of the ultimate force, stress, or energy required to cause injury, he/she must understand mechanics ... and perform or refer to the appropriate experiments."

At the time of [Rigoberto] Avila's trial, and first appeals, biomechanics was not being applied to criminal cases involving child abdominal injuries, a leading cause of death in children, Plunkett and other experts say. Now, Plunkett wrote, such an analysis prior to deciding whether an injury is criminal or not would be "mandatory."

The argument that Avila's conviction was based on faulty, pseudo-scientific conclusions is at the heart of a new appeal filed in September. The appeal cites passage this spring by state lawmakers of Texas' first-in-the-nation law to allow for appeals based upon relevant and newly ascertainable, or evolved, science that contradicts evidence used at trial. The new statute took effect Sept. 1, and is an acknowledgment that criminal law – rigid and, importantly, wedded to finality – must evolve to keep pace with scientific advances.

The principal dispute, Smith reported, surrounds whether biomechanical research regarding child trauma is new science or merely old science re-packaged:

El Paso District Attorney Jaime Esparza does not agree that Avila's case falls under the new law. In a motion filed in October on Esparza's behalf, prosecutor Tom Darnold argues that there is nothing new about the science involved in the case that would warrant review. "Avila acknowledges that the relevant scientific knowledge, that is, the physics of impacts, dates back to Newton and has not undergone significant change since Avila filed his first writ application" 10 years ago, reads a motion to dismiss the case. And even "Reimann notes [in his affidavit] that his analysis of these types of issues is 'based on introductory physics normally presented in a general physics course required for biological, health-science, and pre-med students,'" Darnold noted. As such, the "materials Avila has submitted ... defeat his claim that the scientific knowledge, or the method on which that relevant scientific knowledge is based, has changed since the time he filed his first writ application." And the fact that biomechanics was understood and practiced at the time Avila was tried means only that his lawyers didn't exercise "reasonable diligence" in seeking out that knowledge to present at trial. The plain language of the new law, he notes, "does not authorize the consideration of the merits of a subsequent writ ... based on scientific knowledge that was previously available but not commonly used, or based on scientific knowledge that was previously available but simply was not sought out by the doctors or attorneys in the case."

While it's true that biomechanics takes basic Newtonian physics – the law of bodies in motion – and applies it to living tissues, there is nothing at all static about the science involved, says Peter Stephens, a retired forensic pathologist. And when he hears people dismiss the science as not new or say something like, "'We've moved beyond Newtonian physics now,' most of us just roll our eyes." The fact of the matter, says Stephens, is that using biomechanics as a diagnostic tool for childhood injury is fairly cutting-edge – particularly when it comes to assessing abdominal injuries, like that which ultimately killed Nicholas.

The whole "shaken baby syndrome" (SBS) fiasco - where a field of expert testimony arose in the 1970s to prosecute cases without the underlying science having been tested (see background here) - was one of the brands of junk science under consideration when this new legislation was passed. Grits has no direct knowledge about when new biomechanical modeling that debunked SBS was applied to other types of injuries like the ones in this case. But the idea that the claim should be denied because the physics of impacts "dates back to Newton" is a laughable stance that ignores the history of SBS and the belated rise of biomechanical modeling that has only recently countered it.

Indeed, later this month in Dallas a "World Conference on Infant Head Trauma" will convene to evaluate the evolving status of science in this area. According to the event's website, "The World Congress on Infant Head Trauma brings together international speakers to find common ground and debate controversial topics in pediatric forensic pathology. This unique congress focuses primarily on the pathology and etiology of head trauma (and its mimics) and not the overarching issue of child abuse." This is still very much an evolving field.

In the wake of the new law, not to mention the National Academy of Sciences report in 2009 calling into question an array of non-scientific forensics, we're going to see more habeas writs focused on the question of "what did science know and when did it know it?" That's as it should be.

The case also highlights the drawbacks of failing to record custodial police interrogations. (Regular readers will recall that requiring such interrogations to be recorded is the final, un-enacted recommendation of Texas' Timothy Cole Advisory Panel on Wrongful Convictions.) Here's how Smith described the circumstances of the disputed confession:

Around 11pm that night, as Raschbaum worked on Nicholas [the child victim] at the hospital, Avila was taken to the El Paso PD to give a statement to Det. Tony Tabullo, then a 24-year veteran of the force. Avila was already a suspect in the case, and around 11:30pm Tabullo advised him of his rights before beginning the interview. At roughly 2:10am, Avila signed a statement wherein he detailed that he had been watching TV when Dylan told him that Nicholas wasn't breathing; per Tabullo's instructions, Avila read and placed his initials before and after each paragraph of the statement, and next to the time and date – 23 separate places in all – before signing off on the document.

What happened next is disputed. Accord­ing to Tabullo, after he took Avila's statement he found out that Polaroid photos of Nicholas' body had been brought to the police station. Those photos, he testified, revealed a large area of bruising across Nicholas' chest (the bruised area was seven inches by three inches, according to the autopsy report) that resembled a shoe print. Armed with this information, Tabullo said, he returned to Avila and confronted him with what he considered evidence that Avila had stomped on the baby's chest. Tabullo asked Avila if he could see his shoe; Avila complied. "Well, I asked him," after inspecting Avila's sneaker, "do you want to tell me the truth?" Tabullo testified. "He said yes. He shed a few tears and started telling me the truth."

The truth, according to a second statement Tabullo said Avila offered, hours later, at 5:46am, was that while Avila was watching the basketball game at Macias' apartment he got up and went to the bedroom where Nicholas was alone. "I saw him laying on the floor," reads the second statement. "I don't know what came over me, but I walked over to him and stamped on him with my right foot." Avila then allegedly picked Nicholas up and brought him into the living room where Dylan shook Nicholas to try to "wake him up" and then hit him on the chest with a magazine. Avila called 911. Why had he done it? According to the second statement, it was because he was "jealous" that Macias paid so much attention to the toddler.

According to Avila, Tabullo fabricated that second statement. He said that when Tabullo returned to the interview room where he'd been left after making his first statement, Tabullo confronted him with the photos of the alleged shoe-print bruise. Avila said he knew nothing about the injury, he testified in court. Tabullo told Avila he would have to wait a while longer in an interview room. Tired, Avila asked if he could leave; no, not until Tabullo was "done with you," he testified. Avila was tired and was falling asleep as they talked, he recalled. "I said, 'Is it okay if I go to sleep,' and he says, 'Yes,'" Avila testified. "He goes, 'I just need to make some changes on the statement.' He says, 'If you want ... I'll wake you up when I'm done.'" And that, said Avila, is what he did. When Tabullo woke him up later, Avila said, he was told only that he needed to sign his statement again, and that he could then leave. Avila says he did as he was told – without ever reading the new statement. He was then arrested.

Although there is certainly reason to suspect the validity of the second statement – it was conducted by Tabullo alone and was not recorded, and unlike the first statement, Avila did not initial any of the paragraphs, though Tabullo said that's because Avila told him that he trusted the detective and didn't need to go through that – the damage was done.

How much better would it have been for everyone if Avila's interrogation had been recorded? Without a recording no one besides the two men involved can know for sure what happened in that interrogation room. One of them is lying and it's impossible to know who. Maybe in 2015 the Texas Legislature will finish the tasks assigned it by the Tim Cole panel and finally require recording of custodial interrogations. Otherwise, there's little doubt similar situations will continue to arise, calling convictions into question and muddying the waters for appellate courts trying to suss out the mess after the fact.

Capital case overturned because ineffective counsel failed to challenge flawed forensics

The Court of Criminal Appeals yesterday granted a habeas writ from Manuel Velez, a capital defendant out of Brownsville whose case was the subject of an investigative report by Pam Colloff at Texas Monthly earlier this year. They found Velez's defense counsel were ineffective, improperly failing to investigate the case or to call rebuttal experts to challenge questionable scientific evidence establishing the timing of abuse to the victim, 11-month old Angel Moreno. Here's Colloff's description of the disputed evidence:

Velez’s attorneys won the right to hold an evidentiary hearing, which took place last December. During the weeklong proceeding, not one of the seven medical experts who testified supported the state’s theory of the case. The state’s own expert witness, Dr. Norma Farley—who conducted the autopsy back in 2005—backed away from much of her original trial testimony, in which she stated that all of Angel’s injuries had taken place within two weeks of his death. Some, she allowed, could be 20 to 36 days old. Two forensic pathologists and a neurological surgeon testified that Angel’s skull fractures were more than two weeks old, and perhaps months old. Forensic pathologist Dr. Janice Ophoven explained that not much trauma was needed to push Angel into a critical state on October 31, 2005, if his brain was already swollen from head trauma. A rapid increase in head circumference is evidence of “increased intracranial pressure,” she said, which made the boy “a time bomb.”

The CCA's ruling suggested that the victim's mother - who Colloff noted received a lenient plea deal in exchange for her testimony against Velez, though his lawyers did not disclose that to the jury - may have been the real culprit: "Family members and neighbors also testified at the habeas hearing that they witnessed the victim's mother neglecting and abusing him and his siblings in the months and weeks before his death. during tht time they also observed that he displayed symptoms, such as lethargy, that were consistent with head trauma."

The high court ruled that the record supported the district judge's conclusion that "there is a reasonable probability that, but for defense counsel's failure to investigate and present evidence, the outcome [of the case] would have been different." The case now heads back to the Cameron County DA who must decide whether and how to proceed.

Texas habeas law needs updating to accomodating changing science

I ran across a Chicago Sun Times editorial from last month which recited a litany of examples of cases where a variety of different forensic errors led to false convictions, honing in particularly on the dubious "science" behind so-called shaken baby cases ("Disproving proven theories that led to injustice," Sept. 2). The article opened:

Too often, cutting-edge scientific theories send people to prison only to turn out to have more holes than Swiss cheese.

There was the medical theory — later discredited — that helped land about 30 adults from Bakersfield, Calif., in prison for sexually abusing children in the 1990s.

There were the arson theories that helped send Cameron Todd Willingham to a Texas execution chamber in 2004, only to be rejected later by other experts relying on advances in fire science.

There was the theory of identifiable unique shoe-wear patterns that helped send at least eight men to prison for life in the United States and Canada in the 1980s before scientists concluded the whole idea was fiction.

So when a case involving the theory of “shaken-baby syndrome” returns to the DuPage County courthouse this fall, the theory should be put under a microscope. Only proof that puts all doubt to rest will suffice.

A decade or so ago, many medical experts thought that a “triad” of symptoms in a dead infant — bleeding behind the retinas, brain swelling and bleeding on the brain surface — could be caused only by major trauma, such as a car wreck, or by violent shaking.

If those symptoms were found in a baby or young child who died — and no major trauma occurred — it followed that the baby died by violent shaking at the hands of a caregiver. No witness to the shaking was really necessary. Thousands of Americans have been sent to prison based on medical shaken-baby evidence.

But here’s where the doubt comes in. In the intervening years, some medical experts have traced the “triad” of symptoms to other causes. Even Norman Guthkelch, the Evanston pediatric neurosurgeon who was among the first to describe shaken-baby syndrome, has his doubts about how the diagnosis is used in court. Sometimes, he says, illness can cause the same symptoms.

Other scientists say the symptoms could even be caused by ordinary falls.

Regular readers know the National Academy of Sciences issued a report in 2009 calling into question the scientific basis of quite a few forensic disciplines - particularly those which rely on subjective comparisons by technicians as opposed to testable and demonstrable science. But in general, forensic errors are a less frequent cause of false convictions than, say, eyewitness testimony, IMO mainly because forensics are typically less likely to be the sole evidence against a defendant. That's in part because, in most cases where forensics are submitted into evidence, they play a corroborative role. Typically accusations are made by police and prosecutors early in the process, while forensics are frequently not analyzed unless closer to when a case is ready for trial. And since only about three percent or so of felony cases go to trial, it's not uncommon for convictions to be secured without forensics ever being tested or vetted by defense experts.

But in cases where forensic science forms the primary basis for criminal accusations - especially comparative disciplines like shoe-priint evidence or theories espoused by just a handful of physicians, as with "shaken baby syndrome" - there's a greater risk of false convictions based on erroneous forensics.

Making matters worse, after the Texas Court of Criminal Appeals last year approved Ex Parte Robbins (discussed here), as Judge Cathy Cochran lamented in her dissent, there is now more than ever in Texas  a "fundamental disconnect between the worlds of science and of law," particularly in post-conviction proceedings where junk science is most frequently challenged. Cochran expanded on that theme earlier this year in another case, opining that:

This disconnect between changing science and reliable verdicts that can stand the test of time has grown in recent years as the speed with which new science and revised scientific methodologies debunk what had formerly been thought of as reliable forensic science has increased. The potential problem of relying on today's science in a criminal trial (especially to determine an essential element such as criminal causation or the identity of the perpetrator) is that tomorrow's science sometimes changes and, based upon that changed science, the former verdict may look inaccurate, if not downright ludicrous. But the convicted person is still imprisoned. Given the facts viewed in the fullness of time, today's public may reasonably perceive that the criminal justice system is sometimes unjust and inaccurate. Finality of judgment is essential in criminal cases, but so is accuracy of the result--an accurate result that will stand the test of time and changes in scientific knowledge.

Sometimes, Cochran noted, the problem "is not that the science itself has evolved, but that it is alleged that the scientific testimony at the original trial was not fully informed and did not take into account all of the scientific evidence now available." In such instances, following Ex Parte Robbins, there is presently no basis for relief via habeas corpus writs available under Texas law, at least as viewed through the narrow lens of the majority on the Texas' Court of Criminal Appeals.

In the past several sessions the Texas Legislature has taken a keen interest in innocence legislation, passing landmark reforms to improve procedures for gathering eyewitness evidence, expand access to post-conviction DNA testing, require corroboration for jailhouse informants, and approve the most generous compensation package in the country for exonerated defendants. This gap Judge Cochran has identified in the Court of Criminal Appeals' jurisprudence on challenging flawed science through habeas writs is arguably the most pressing innocence-related issue the Texas Legislature needs to address when it meets again in 2013.

State Senate Criminal Justice Committee Chairman John Whitmire has pursued legislation for the past two sessions to provide a vehicle for reconsidering flawed forensic evidence in habeas appeals, which was a recommendation of the Timothy Cole Advisory Panel on preventing false convictions. But that legislation faced opposition from district attorneys who considered the issue not yet ripe - before Ex Parte Robbins, the CCA had hinted that such appeals wouldn't pass muster but had never ruled on an on-point case. After that case, however, the ball is now squarely in the Legislature's court. If cases like those described in the Sun Times editorial are to get a fair hearing in Texas, the Lege must assert itself to insist on it.

RELATED: Northwestern's Medill Innocence Project turns attention to shaken baby cases.

Christmas in March for Amarillo man freed due to flawed forensics in child death

In a story earlier featured on PBS Frontline, NPR, and ProPublica, which brings the latest news, Ernie Lopez returned home to Amarillo yesterday after the Texas Court of Criminal Appeals invalidated his conviction because his defense attorney did not call expert witnesses to invalidate the science underlying his conviction for child sexual assault. The CCA "didn't absolve Lopez, but said there was a 'reasonable probability' that he would have been acquitted had his attorneys countered the medical evidence arrayed against him." Says ProPubica:

Lopez's story reflects a broader controversy. Increasing numbers of doctors and other experts are questioning the reliability of the science used to prosecute cases of fatal child abuse and sexual assault. In Canada and the U.S. at least 23 people who were wrongly accused of killing children based on flawed or biased work by forensic pathologists have been cleared over the last 15 years.

The Texas court's ruling set the stage for Lopez to be released Friday on a $10,000 bond. His release order bars him from initiating contact with children, though he is allowed to spend time with his three children. It also requires him to wear an electronic monitoring anklet and to abide by a 10 p.m. to 6 a.m. curfew.

At his parents' home on the outskirts of Amarillo, Lopez celebrated the events and holidays he missed during his long confinement. The family even put up a Christmas tree.

Related: TX conviction overturned because of failure to challenge junk science in child death case.

Junk science v. Mom

With this blog's interest in junk science, Grits should point out two recent stories regarding women convicted of killing children based on questionable forensics: One so-called "shaken baby" case, which is a highly questionable medical diagnosis created for the sole purpose of prosecution, and another case here in Austin where an expert hired by the defense was so unprofessional he did more harm than good:

• "Shaken baby syndrome faces new questions in court," New York Times, Feb. 3
• "A Parliament of Experts: Did 'expert testimony' convict an innocent woman of murder?," Austin Chronicle, Feb. 4

I don't have time today to write these up, but wanted to pass the links along.

See related Grits posts:

• Science undermining 'shaken baby syndrome' cases
• Shaken baby diagnosis from biased medical examiner unsubstantiated, new autopsy say

Shaken baby diagnosis from biased medical examiner unsubstantiated, new autopsy says

This blog focuses frequently on the subject of dubious forensics like that used in many older arson cases and so-called "scent lineups" performed by dogs. The Houston Chronicle this week published a story on another branch of questionable forensic science: The diagnosis by medical examiners of so called "shaken-baby syndrome," which has resulted in several recent exonerations in Harris County after former associate medical examiner Patricia Moore's work on the subject was discredited. ("New results of autopsy spur plea," Jan. 14)

Most recently, Harris County was compelled to change the autopsy results by the same medical examiner in an 11-year old case after concluding there was no basis for a shaken-baby finding. Reported the Chron:

The original autopsy classified the baby's death as a homicide and was used by prosecutors as a key piece of evidence against Cynthia Cash, now 53, a former nurse convicted of fatal injury to a child after 4-month-old Abbey Clements died after being rushed to the hospital from Cash's home.

But the modified autopsy report made public in a new appeal calls the cause of death “undetermined” and found no evidence of “trauma” in the postmortem exam. Those changes came five years after local officials announced a review of problematic autopsies conducted by a former Harris County associate medical examiner, Dr. Patricia Moore. Moore, who declined requests for comment, left Harris County in 2002 but still works for Southeast Texas Forensic Center, a Conroe-based company that provides forensic work for six counties.

It is at least the fourth time Harris County officials have reclassified a child's autopsy that Moore originally labeled as a homicide. Two women have been cleared in other cases — including Brandy Briggs, who was jailed at 19 after rushing her baby to the hospital and who spent several years in a prison isolation cell before being freed in 2005. Dr. Luis Sanchez, head of the medical examiner's office, did not respond to Chronicle questions about Cash's case or whether he has finished an audit he promised to conduct after finding problems in the Briggs case.

It's particularly disturbing to learn that Moore continues to perform autopsies in Texas criminal cases, making one wonder what quality control mechanisms exist (or should be implemented) to eject incompetent or biased forensic workers from the field.

Over at The Agitator, Radley Balko commented on the case, noting that "In 2004, a statistical analysis showed Moore diagnosed shaken baby syndrome (already a controversial diagnosis) in infant deaths at a rate several times higher than the national average. Roger Koppl and I noted her case in recommending statistical analysis as one way of checking the integrity of state forensic specialists."

Related Grits posts:

• Science undermining 'shaken baby syndrome' cases
  
• NAS Report: Many forensic disciplines prone to error
• Justice system commonly relies on shoddy forensic science
• USDOJ should embrace, not fight science-based forensics

Science undermining 'shaken baby syndrome' cases

So-called "shaken baby syndrome" is another area where it increasingly appears that flawed forensic testimony helped secure numerous false convictions over the years, particularly among parents and child care workers: According to an essay by Maurice Possley at The Crime Report:

A soon-to-be-published analysis of shaken baby cases and recent developments in the medical community by University of Maine School of Law professor Deborah Tuerkheimer presents persuasive evidence and raises troubling questions about whether many of these convictions were of innocent people who were found guilty on the basis of faulty science. The analysis is scheduled to be published in September by Washington University Law Review.

Tuerkheimer, who is joining the DePaul University College of Law faculty on July 1, points to new research in the United States and abroad showing that a variety of circumstances, including something as seemingly innocuous as falls from a short height, can cause fatal head injuries that appear very similar to injuries routinely diagnosed as SBS.

If research shows that the physical conditions that once automatically resulted in a prosecution could actually have been the result of an accident, the implications are enormous.

“Given the scientific developments…we may surmise that a sizeable portion of the universe of defendants convicted of SBS-based crimes is, in all likelihood, factually innocent,” Tuerkheimer writes, adding that a far greater number of defendants among the group were likely convicted on legally insufficient evidence.

“While we cannot know how many convictions are ‘unsafe’ without systematic case review, a comparison of the problematic category of SBS convictions to DNA and other mass exonerations to date reveals that this injustice is commensurate with any yet seen in the criminal justice arena,” Tuerkheimer writes.

Keith Findley, a clinical professor of law and co-director of the Wisconsin Innocence Project, who headed Audrey Edmunds legal team, said, “The system is sending people to prison based on findings of beyond a reasonable doubt when in many of the cases the only evidence is medical evidence on which many medical experts…have a substantial doubt.”

He added, “This is not about being opposed to child abuse prosecutions. No critic of SBS theory wants anyone to get away with child abuse, but when the diagnosis becomes the entire basis for the prosecution, that’s problematic.”

Read the full story for more detail on why such testimony has come under fire. Convictions are still being obtained based on this forensic theory even though “there is no consensus among medical professionals as to whether the symptoms that have traditionally been attributed to SBS are necessarily indicative of intentional shaking.”

Like arson, this is a crime for which people can be convicted based solely on "expert" forensic testimony. But the underlying tenets of "shaken baby syndrome," which have been relied upon in court for many years, have now been widely disputed by credible, scientific research:

Dr. Bruce Gross, a Fellow of the American College of Forensic Examiners, writing earlier this year in The Forensic Examiner, noted that studies have called into question the SBS triad as the result of only violent shaking. “The prevailing notion is that the injuries ‘characteristic’ of SBS are equivalent to those seen in a 35 mph automobile accident in which the infant victim was unrestrained, or a fall from a two –story building. Yet, research (including biomechanical analysis) has shown that, although fortunately not the norm, infants and toddlers can and do die from falls as short as 1-4 feet.”

Gross added, “In brief, biomechanical research suggests that basing the diagnosis of SBS only on the presence of the triad of symptoms [retinal hemorrhage, bleeding in the brain and brain swelling] lacks scientific certainty.”

Last year I got to hear a presentation about some of the biomechanical simulations that appear to undermine the traditional shaken baby diagnosis. Bottom line: the symptoms that supposedly characterize SBS could also have been explained by accidents or birth defects, but doctors instead testified it could only have happened through malicious shaking. These conclusions were based on speculative theories that, by their original authors' own admission, relied on circumstantial guesswork rather than experimentation and proof. Today more experimentation has been done and the original SBS theorizing would never hold up to peer-reviewed scrutiny. But the power of precedent continues to give such theories a foothold in the courtroom.

When you think about it, the death of a baby is already such a terrible event, how much more would it compound the tragedy to then falsely accuse a parent or child care worker of homicide? But to judge by the latest research, that appears to be exactly what's happened in many of these cases since the diagnosis was first popularized in the '70s.

The shaken baby cases provide further evidence, if more were needed, that innocent people can be convicted in many more circumstances than just those where DNA is available to definitively exonerate a defendant.