Once a national model, Georgia’s child death reviews now an empty exercise

By the time the local child fatality review committee took up her case, Marnee Kay Downey had been dead one day longer than she lived.

The committee, as mandated by state law, compiled the details of the 8-month-old’s life and death: Chronically malnourished, she weighed less than 10 pounds. Child-protection workers had removed her siblings from their home. And her death, on Oct. 10, 2012, was a homicide: an intentionally administered overdose of a potent painkiller intended for terminally ill cancer patients.

But the committee remained silent on perhaps the most pertinent question: Why was Marnee even with her parents? Her mother took illicit drugs during pregnancy, the child was born with drugs in her system, and her parents celebrated the conclusion of a child-neglect investigation by getting high.

The review committee’s report mentioned none of that. Nor did it suggest anything the state Division of Family and Children Services or other agencies might have to done to protect Marnee. In fact, it recommended nothing at all.

What happened after Marnee Downey died illustrates the breakdown in a critical component of Georgia’s child-protection system: the review process that is supposed to dig deep into why a child died and search for ways to prevent more deaths.

Almost a quarter-century ago, Georgia became the first state to create committees for each county to examine every death of a child. Today, all other states follow a similar practice.

But Georgia’s review process, an investigation by The Atlanta Journal-Constitution found, has become an empty exercise.

The newspaper examined reports on 464 deaths that county committees submitted in 2012 to the state Child Fatality Review Panel, which oversees their work. Officials redacted significant portions of the reports: the names of deceased children and their parents, dates of birth and death, even the time of day a child died.

Still, enough data remained to show that, in most cases, the county committees’ work was superficial and slow.

The committees have a broad mandate to study failures by government agencies and to seek changes that would correct mistakes.

But not a single report among the 464 delved into mistakes by DFCS or other agencies. Just two reports requested additional investigation by police or other authorities.

In cases for which the Journal-Constitution could establish a complete timeline, the committees took more than twice as long to complete reviews as the law allows. A few counties, such as DeKalb, left some cases unreviewed for a year and a half or longer and didn’t complete reports on 2012 deaths until November or December 2013. In a recent publication, the statewide panel said timely reviews “build momentum from the tragic event to effect change in the community.” But one-fourth of reviews took longer than the child’s lifetime.

Almost 500 deaths — slightly more than half of the 940 recorded among Georgia children age 17 or younger in 2012 — received no review. Many were attributed to causes that rarely raise suspicions, such as disease or premature birth. But the Journal-Constitution identified 56 deaths in 38 counties that seemed to warrant investigation. Among them were 10 homicides, eight suicides, and seven drownings. Twelve of the 56 children came from families that DFCS had previously investigated.

Perhaps most significant, almost three-fourths of all reports contained no recommendations for preventing other deaths — one of the primary purposes for the reviews. Even when reports included suggestions, they usually offered what amounted to after-the-fact, self-evident parenting advice.

When an unsupervised 2-year-old wandered away and drowned in a pond, the Atkinson County review committee said parents should “watch kids at all times near open water.” When a woman ran over her 4-year-old granddaughter in the driveway, Muscogee County’s committee suggested “buying cars with back-up cameras.” When a baby died from a methadone overdose, the Fulton County committee advised simply, “Parent education.”

Jones County’s committee reviewed the death of an 8-month-old whose family already was the subject of a DFCS investigation. His mother, an unemployed high-school dropout, had a history of marijuana and cocaine abuse, according to state records. Three times DFCS had substantiated abuse or neglect allegations against her. On Oct. 25, 2012, she handed the baby a whole hot dog and left him alone. He choked to death.

The committee’s recommendation: “not to feed an infant-toddler a whole hot dog.”

The performance of the review committees — made up of prosecutors, judges, police, coroners, social workers and other officials and chaired by the local district attorney — stands in contrast to the mission the statewide panel articulated in a summary of its work from 2012.

“When a child dies, we have an obligation to ask ourselves, was there something that we — as caregivers, as a community, and as a society — could have done to prevent this tragic consequence?” the panel wrote. “Did we fail this child in some way?”

The chairman of the statewide panel acknowledges a gap exists between rhetoric and reality. Cobb County Superior Court Judge Tain Kell, who has led the panel since 2012, supports a bill in the General Assembly that would transfer the panel’s operations from the Office of the Child Advocate, a division of the governor’s office, to the Georgia Bureau of Investigation.

With few resources, Kell said in an interview, the state panel cannot properly monitor the county committees’ work.

“What we do,” he said, “is only as good as the data we get.”

Marnee Downey died in Haralson County, an hour’s drive west of Atlanta. Her death, like hundreds of others, was allowed to go by with little notice, even in the small community that failed to prevent it.

A life in brief

With hindsight, it’s clear that Marnee Downey’s life was following a short, tragic arc.

Her mother, Melanie Ann Prichard, had served time in an Alabama prison in 2007 and 2008 on a drug-possession charge. Prichard’s drug use caused state social workers to take two of her older children into protective custody. When Prichard gave birth to Marnee, according to court records, routine hospital tests found drugs in the blood of both mother and child.

Marnee’s father, Robert Derek Downey, was an insurance agent from a prominent family in Tallapoosa, a dozen miles from the Alabama state line. He and Prichard had been together, on and off, for almost 20 years — since he was 27 and she was 17. Downey also had a history of substance abuse and violent behavior, according to state records. After the baby died, Prichard told the police Downey had physically assaulted her and had planned to poison her former husband. Once, she said, Downey fired a gun at her, striking the cigarette pack she was picking up. But she never called the police on Downey or tried to leave him.

DFCS got involved in Marnee’s life within a few hours of her birth.

The agency substantiated a report of “newborn drug exposure,” a caseworker later wrote, but decided not to take Marnee into protective custody. Instead, the caseworker instructed Prichard and Downey to get counseling and other “family preservation” services. The only other condition was that the couple and the baby temporarily move in with Downey’s parents.

A few months later, the caseworker wrote that she “was able to see positive changes.” As Prichard “became stronger emotionally,” the caseworker wrote, she and Marnee exhibited an “obvious bond.”

DFCS closed the case on Aug. 6, 2012.

Downey and Prichard took Marnee home to Downey’s cedar-sided ranch house on a country road a couple of miles outside Tallapoosa.

The first evening, Prichard would later tell the police, she and Downey started smoking marijuana again.

In a snapshot from about that time, Marnee’s cheeks and upper arm are chubby with baby fat. Her blond hair is little more than peach fuzz. She gazes away from the camera, her mouth formed into a half-grin.

In the weeks that followed, though, Marnee developed a painful skin rash, which a doctor diagnosed as eczema. Meanwhile, her parents argued over how much weight she was losing.

“The baby is fat,” Prichard said Downey told her. “You’re feeding her too much.”

The evening of Oct. 9, the DFCS investigation had been closed barely two months. Prichard put Marnee to bed in a pink velour dress and, she told the police, smoked marijuana with Downey and took three Valium tablets before going to bed.

Prichard and Downey argued the next morning, and she said he struck her in the face. She said Downey went into Marnee’s room for a minute or two before he left for work about 7:30, but she didn’t check on the baby until just after 9. Prichard found Marnee in her crib, lying on her back, motionless.

Prichard said she grabbed Marnee and ran down the road to the nearest neighbor.

“Help me!” she screamed. “Something is wrong with Marnee!”

An ambulance took Marnee to a hospital in Bremen, 11 miles away. A doctor there pronounced her dead at 11 a.m.

At her autopsy, Marnee weighed nine pounds, eight ounces — about half the weight of a healthy girl her age and only three pounds more than her birth weight. The medical examiner said she was malnourished and dehydrated. But the cause of death, the autopsy determined, was intentional poisoning by Fentanyl, an opiate painkiller.

The drug, according to pediatric experts, would have slowed her breathing. As the amount of carbon dioxide in her bloodstream rose and oxygen levels decreased, she may have passed out. Finally, her heart would have stopped beating.

Marnee’s family buried her in Tallapoosa, beneath a heart-shaped tombstone inscribed, “We love you.”

Haralson County authorities arrested both parents on charges of murder, child cruelty and drug possession.

No longer a couple, Prichard, 38, and Downey, 48, each pleaded not guilty and, in court papers and statements to the police, blamed one another for Marnee’s death. Although Downey gave no statement to the police and did not directly answer Prichard’s allegations, his lawyers described her as “a chronic user of narcotics” and claimed she “likely obtained and administered the Fentanyl without (Downey’s) knowledge.”

But Prichard’s lawyer, Mac Pilgrim, said in an interview: “How it got there, who put it there, who could have put it there — those are questions for the jury.”

The trial is scheduled to begin Monday. District Attorney Jack Browning, who was elected after the review committee looked into Marnee’s death, said he could not comment on the case until the trial concludes.

Beginning the day Marnee died, members of the county’s child fatality review committee faced a tight deadline: 90 days to investigate and report on her death.

They wouldn’t even come close.

‘Everybody’s frustrated’

A scandal sparked the idea of child fatality review in Georgia.

In 1989, the Journal-Constitution reported that dozens of children’s deaths had been mislabeled as “natural” or “accidental” rather than as homicides caused by abuse or neglect. The General Assembly responded the next year by creating a network of review committees — one for each of Georgia’s 159 counties — and a statewide panel to oversee their work. The scope of that work had few limits; committees could seek new criminal investigations of deaths, propose legislation to protect children, or revisit cases that other officials had already disposed of.

State Rep. Mary Margaret Oliver, D-Decatur, sponsored the 1990 legislation. The intent, she said recently, was “to understand where the gaps in child protection were. What were the most prevalent reasons children were dying? What could we learn from that and, more importantly, what can we do about it?”

But many counties resisted — more than one-third filed no reports in 1995, for example — and even now, compliance is far from universal. Thirty-two counties where at least one death occurred in 2012 failed to report.

Kell, the Cobb judge who chairs the statewide panel, said counties should report on “each and every child fatality.” Otherwise, he said, “it hinders our ability to prioritize” and to spot trends in how children across the state are dying.

But the panel, too, has lagged in meeting its responsibilities.

When county officials don’t comply with the law, the panel can seek contempt of court sanctions. But it has never done so.

The panel’s staff is unable to make sure counties examine every death or even that their reports contain correct information. In many cases, the staff acts as nothing but a repository for the counties’ submissions.

“They are not able to perform some of the functions they used to perform,” Kell said. “I don’t know that we are able to review each and every one of these (reports).”

Training for county committee members also is inadequate, Kell said. Last week the panel’s website featured a notice of training scheduled for September — September of 2013.

“I think everybody’s frustrated,” Kell said.

So from county committee to county committee, the law is applied unequally.

“Some of them are really good; some of them don’t meet,” said Melissa Carter, a former state child advocate who is executive director of the Barton Child Law and Policy Center at Emory University’s law school. “Some are very rigorous; some don’t report.”

The Journal-Constitution found wild inconsistencies in the committees’ conclusions. Two deaths under almost identical circumstances may be considered the result of neglect or abuse in one county, an unavoidable accident in another.

For example, the committees reported on 153 children who died in 2012 while sleeping in unsafe situations — with an adult, for instance, or in a crib filled with pillows or stuffed toys. But the committees said just 71 of those deaths, or fewer than half, could be blamed directly or indirectly on actions by parents or other caregivers.

Troup County’s review committee said an infant there died of asphyxiation after his father put him down for a nap on an adult-sized bed. DFCS already was investigating the baby’s family over a burn suffered by a sibling. The day the baby rolled to the side of the bed and died, the committee said, the father was drunk while supervising all seven of his children.

Yet, the committee said, the father was not to blame for the baby’s death.

In Fulton County, 7-month-old Timothy Landers III died in a motel room overlooking I-285 in College Park. While the father came and went from the room, the mother fell asleep in bed with the baby. The father noticed the child wasn’t breathing, according to police reports, so he “picked up the baby and ‘shook him a little.’”

An autopsy found the boy died from swelling in the brain. He also hemorrhaged in each eye. Both can be signs of abuse, the medical examiner said.

College Park police arrested Timothy’s parents on murder charges. Both denied killing the child, and prosecutors decided not to take them to trial.

Fulton’s review committee filled out the standardized reporting form used statewide, but gave the same answer to 54 questions: “U/K,” for “unknown.” The committee said it didn’t know whether either parent had a criminal record; in fact, both did — the mother for a drug offense, the father for aggravated assault. Nor could the committee say whether the parents had used illegal drugs — even though police found marijuana in their motel room the night of Timothy’s death.

The review committee’s conclusion left open even the possibility that Timothy died of natural causes: “Unknown if external causes contributed to death.”

Unanswered questions

A month before Marnee Downey died, Haralson County’s child fatality review committee met for the first time in 2012. One case was on the agenda: the death of a 14-year-old who shot himself to death — six months earlier.

Even with that much time, the committee still could not answer numerous questions: What type of gun did he use? Where did he get it? Was it stored fully loaded or with ammunition nearby? To whom did the weapon belong?

The committee gave the same answer to each question: “Unknown.”

In a departure from the norm, the committee suggested prevention measures, including family crisis intervention. No record exists of whether the committee or any other organization carried out the recommendations.

The Haralson committee met next on June 13, 2013, nine months later, to review three deaths.

As required by law, the meeting was closed to the public. Representatives of seven agencies — among them, the district attorney’s office, law enforcement, the coroner’s office and DFCS — attended, but public records don’t identify them by name.

That day, the committee reviewed the cases of a 16-year-old who died from a seizure disorder and a 9-month-old who died while sleeping on a sofa. The teenager had been dead 13 months, the baby for seven. The committee recommended nothing in either case, noting that circumstances beyond its control prevented effective reviews: “Meeting was held too long after death.”

The other case was Marnee Downey’s.

The committee reported to the statewide review panel that Marnee had been a victim of previous neglect. It said her mother smoked cigarettes before and during pregnancy and that both parents had been abusive or negligent toward their children. It cited the autopsy’s finding that a drug overdose killed Marnee, but said the mother’s negligence — specifically, her failure to give the baby enough food — directly caused the death.

More notable, perhaps, may be what was missing from the report:

An entire community failed to recognize that the child was in distress. Sores from eczema covered Marnee’s skin. She had lost so much weight that her ribs stuck out from her sides. Yet, neither DFCS workers nor neighbors, neither doctors nor relatives could save her.

Near the end of its report, the review committee came to a fundamental question: Could the death have been prevented?

A DFCS worker who prepared the report checked the first option: “Yes, probably.”


The committee left that section blank.

Support real journalism. Support local journalism. Subscribe to The Atlanta Journal-Constitution today. See offers.

Your subscription to the Atlanta Journal-Constitution funds in-depth reporting and investigations that keep you informed. Thank you for supporting real journalism.