In Lake County, a disfigured 2-month-old whose mother did not want him is left alone in a motel room for 90 minutes, and is later found smothered. His family had been the subject of 38 prior investigations by the state’s child welfare agency.
“It is a general consensus,” a report said, “that [the mother] was involved in the death of her child.”
In Santa Rosa County, child welfare authorities allow a “chronic and severe” drug addict to bring her newborn home, though her two older children had been removed from her care for their safety. Eighteen days later, the mother takes an unprescribed Lortab painkiller and places her baby next to her in bed. The child is found dead.
And in Polk County, a mother leaves two toddlers alone in a “kiddie pool” — and returns to find her 1-year-old daughter face-down in the water. Her 2-year-old son later discloses he pushed his sister down while she was crying. He now suffers nightmares.
Never miss a local story.
The children, who all perished last year, are tragically bound by more than death: Even as the Florida Department of Children & Families has promised greater openness, the three fatalities, and dozens of others like them, have never been counted among the state’s victims of fatal abuse or neglect.
No state can protect every child who is born to troubled, violent or drug-addicted parents, and even youngsters for whom child protection administrators make all the right choices can sometimes fall victim to unforeseen circumstances. To ensure that state social service agencies learn from mistakes, the federal government requires that states count and investigate all child fatalities that result from abuse or neglect.
Regulators don’t, however, strenuously oversee how the counting and investigating occurs.
After the Miami Herald published a series examining the deaths of 477 children — and Florida’s failure to protect some of them from abusive or neglectful parents — the state promised a new era of openness and more rigor in the way it investigates child deaths.
But except for abiding by a new state law that required DCF to create a website listing all child fatalities, Florida has continued to undercount the number of children it fails.
“Nothing has changed,” said former Broward Sheriff’s Office Cmdr. James Harn, who supervised child abuse investigations before retiring when a new sheriff was elected last year. “Some day, somebody will say ‘let’s just stop the political wrangling.’ Here’s what you’ve got to do: Just tell the truth.”
For several years, BSO, which has investigated child deaths under contract with DCF, has recorded significantly more fatalities due to neglect or abuse than other counties, where DCF does its own investigations. One important reason for the disparity is that the sheriff’s office long has insisted that drownings and accidental suffocations — among the leading causes of child fatality — be counted, while DCF has, in recent years, declined to include the majority of those in its abuse and neglect tally.
As a result, said Harn, the statewide numbers “are cooked.”
“It’s not going to get fixed as long as they want to hide things,” Harn added.
A DCF spokeswoman in Tallahassee, Alexis Lambert, said the agency studies all child fatalities — not just the ones it verifies as resulting from abuse or neglect — to “improve and strengthen child welfare practice and services provided to vulnerable children and at-risk families statewide.”
She added: “The safety and well-being of Florida’s vulnerable children is DCF’s top priority. Understanding and assessing child fatalities is one way the department analyzes the issues facing families and develops strategies to meet the needs of struggling families and protect vulnerable children.”
Child deaths became an election issue in recent weeks as both Gov. Rick Scott, a Republican, and his Democratic challenger, former Gov. Charlie Crist, traded accusations over whose administration better protected children. In two debates, Scott has stated that child fatalities have declined dramatically since he took office in 2011. In 2009, he said in the first debate, 97 children with a DCF history died. “Last year,” he added, “we were down to 36 deaths.” The DCF child death website actually lists 45. The Miami Herald has counted 72, twice the governor’s figure.
“That’s still way too many,” he said. “We don’t want to have one death. But it’s a dramatic improvement from when Charlie was there.”
Scott repeated the assertion in a news release and in the second debate last Wednesday.
A careful study of thousands of pages of state documents makes clear that the number Scott cited and the one on the DCF website are both distortions of reality. They are contorted by years-long delays in completing investigations — thus keeping deaths off the books — by a decision to narrow the definition of what constitutes neglect, and by a determination to “unverify” some child deaths that had previously been “verified” as abuse or neglect.
Said Pamela Graham, a Florida State University social work professor who served on a Department of Health statewide death review committee for five years: “Numbers lie if you aren’t counting them.”
For many years, state child protection systems have been evaluated in large part by a standard measure: “verified” child deaths “with priors” — that is, the number of youngsters whose families had prior contact with the state. That is the number that Gov. Scott says is declining. But the decline in deaths with priors can be traced to the factors cited above.
Drownings and accidental suffocations differ from, say, beating or shooting deaths in an important way: “There are human decisions in how you categorize them,” said Richard Gelles, who is dean of the University of Pennsylvania’s School of Social Policy & Practice. There is simply no wiggle room as to whether a beating is a child abuse death. In contrast, a drowning can be a neglect death, or, as many more of them are now called, merely tragic accidents.
Verified child deaths did spike during the Crist administration, whose term ended in January 2011. And the numbers on DCF’s newly minted website show that overall child deaths — whether from abuse or illness or something else — have receded somewhat. But the more striking decline has been in the percentage of child deaths that are “verified” as neglect or abuse.
In 2009, 43 percent, or 206, of DCF’s 474 reported child deaths were verified. The percentage dropped to 34 percent in 2010, 31 percent in 2011 and 2012, and then to 10 percent in 2013. So far this year, 13 percent of DCF’s 348 reported deaths have been verified, the records show.
Changes in the standard for verification of child deaths have led to some unusual variations. For instance, Broward County — where BSO investigates child deaths under contract, independent of DCF — tallied 21 “verified” child deaths from abuse or neglect in 2013 out of 30 total fatalities, or more than two-thirds. In Miami-Dade, as of Thursday night, DCF had not verified a single 2013 death as being from abuse or neglect out of the 39 overall fatalities for that county listed on the agency’s website.
On Thursday, the June 21, 2013, killing of Ezra Raphael remained unclassified, though the boyfriend of Ezra’s mother, Claude Alexis, is in jail awaiting trial on a murder charge. Alexis told North Miami Beach police he whipped the child with a belt for spilling bathwater. After the Herald asked about the 39 Miami cases Friday morning, Ezra’s case was switched that same day to “verified.”
The July 20, 2013, death of Jayden Villegas-Morales remains unverified, though his father, Angel Villegas, is charged with manslaughter. Of Jayden’s death, DCF says only that the “2-year-old child was found unresponsive by his father.”
Lambert, DCF’s spokeswoman, said 15 of the 39 Miami-Dade child fatalities from last year remain under investigation by the department.
Aftermath of reforms
Following the Herald’s series on child deaths, Innocents Lost, lawmakers passed a sweeping reform bill. One of its provisions required DCF to maintain the website with details on every child death that is reported. The website depicts a dramatic decline in the verification of deaths in categories that are susceptible to manipulation: the drownings and accidental smotherings. In 2009, before the new neglect guidelines took effect, 76 drownings were recorded. Of those, 58 — or 76 percent — were verified as resulting from neglect, and, among those, 26, or 45 percent, came from families with a prior agency history.
So far this year, DCF has tallied 66 drowning deaths. Only nine of the 66 cases were “verified” as resulting from neglect — or 13 percent — and only one of those nine involved a child whose family had any DCF involvement in the prior five years.
Over five years, then, the share of drowning deaths that were verified as neglect dropped from 76 percent to 13 percent. And in the category that is used to judge Florida’s child-protective efforts against other states — deaths with priors — only one of the 66 drownings from this year is on course to make the list.
The percentage of unsafe sleep deaths verified by DCF also declined, though far less sharply. In 2009, DCF reported 97 such fatalities; among them, 31, or 30 percent, were verified. Of those, 21, or 68 percent were children with a prior family history within the previous five years. In 2013, DCF reported 103 unsafe sleep deaths, and 26 of them, or 25 percent, were verified, DCF’s records show. Among the 26 that were verified, fewer than half involved children with a prior agency history.
Lambert said investigating drowning or unsafe sleep deaths is particularly challenging. “Deaths resulting from drowning and co-sleeping require the most extensive analysis and investigation as these deaths can sometimes be tragic accidents,” she said.
Gelles, the social work dean, said the total number of drowning and unsafe sleep deaths is generally not susceptible to manipulation — but the precipitous decline in cases that are verified suggests that those numbers are “fudged.”
“They are no less dead,” he added.
The death of Nyla Hardy was counted. And then it wasn’t.
Nyla was born on April 17, 2013, to 22-year-old Alysha Rivers and 24-year-old Kendrick Hardy, two longtime drug abusers. Rivers, according to state records, “admitted to smoking two blunts [of marijuana] a day, and the father admitted to smoking five blunts a day.” A former foster child herself, Rivers had lost custody of an older child in 2007 due to marijuana abuse.
Two months before Nyla was born, DCF received a report of physical abuse, burns and environmental hazards involving older siblings, including one who was previously removed but later reunited with the parents. The allegations were ruled unfounded, and the agency closed its case when the mother agreed to accept help from the state.
When Nyla was 3 weeks old, her parents smoked marijuana and then went to bed. Hardy had rolled over onto the newborn once before, the mother said, but the couple continued to co-sleep with the child anyway. That night, “the mother stated she placed the baby in the middle of the bed, with her arm around her,” a report said. When she woke up the next morning, Nyla “was all blue; she was cold and hard.” She also had suffered “multiple hemorrhages...two contusions,” and was drenched in her own blood — so much so, a report said, that the local medical examiner initially said the death most likely “will turn into a homicide.”
In the end, the cause of death was ruled undetermined. In November 2013, the Herald obtained records on 40 child death investigations DCF submitted to a consultant, Casey Family Programs, for the consultant to analyze. At the time, Nyla Hardy’s death had been “verified” as a neglect death. But months later, when a formal death review was attached to DCF’s fatality website, the classification had been switched to unverified. It now will never be counted among the 2013 neglect-related deaths.
Even as it deemed the case an unverified death, the final death review noted: “Due to the extensive use of drugs by the parents, a child died while in their care.”
‘Verified’ to ‘unverified’
Charlize Terrell’s name has vanished from the 2013 count, as well. Her death also was among the 40 cases reviewed last year by the consultant, and a five-page “casework analysis” provided to the Herald noted an investigation of her May 4, 2013, death was “closed” and “verified.” Charlize was born addicted to her mother’s opiate drugs, and spent several weeks in a hospital detoxifying from them.
Charlize died 10 weeks after birth. Both of her parents “admitted to being under the influence of methadone and alcohol the night Charlize died,” a report said. And though the details of her death remained undetermined, “investigators believe that the mother likely rolled over on top of the infant after falling asleep under the influence.”
In a final death review, the “verified” finding was changed to unverified. Charlize’s death will not be counted.
The “casework analysis” of the May 31, 2013, death of Brooklyn Stewart prepared for the DCF consultant concluded that the 1-year-old girl’s drowning “should be” verified, records obtained by the Herald showed. Brooklyn was the toddler left unsupervised in a kiddie pool with her slightly older brother. “Neighbors reported that the children were often seen outside unattended,” the initial review said. It added that Brooklyn’s mother offered differing accounts of the drowning.
Brooklyn’s parents had been the subject of 11 DCF complaints, the most recent four months before Brooklyn’s death.
A bruise on Brooklyn’s head supported her older brother’s claim that he had pushed her in the pool, the report said.
Despite what was written initially, DCF still closed the case as unverified, without offering any explanation in the formal death review for the switch. And despite the formal finding that abuse or neglect did not cause Brooklyn’s death, her siblings were removed from their parents’ care. Brooklyn’s death did not count.
In another of the Casey Family Programs reviews obtained by the Herald, an unnamed DCF administrator wrote that the March 17, 2013, drowning of a 1-year-old Polk County boy should be verified, as the agency had faulted the family for failing to supervise the toddler — a lapse that led to the boy’s death. Yet that case, too, was never counted.
An unnamed DCF administrator in another case reviewed by the consultant chided death investigators for failing to verify the Feb. 28, 2013, smothering death of a 4-month-old Manatee County girl. The infant’s mom had been the subject of four prior child abuse hotline reports from 2011 through 2013, including two reports of violence between the parents.
“The child’s father,” the administrator wrote, “admitted he had been educated on co-sleeping dangers by the paternal grandmother, who works with infants and children, yet despite this knowledge, he chose to place his infant daughter face down on a blanket in his bed.”
Still, Tampa Bay’s child death supervisor at the time, Lisa Rivera, suggested — in clear conflict with the administrator cited in the Casey Families report — that the father may not have been aware of the dangers of co-sleeping. Though the infant’s mother had been warned in a prior DCF probe, Rivera wrote, “the mother admitted that she later failed to share that information with him.”
Rivera now is the top child death administrator statewide. She wrote a recent email in which she asked that a death in Broward from unsafe sleeping be changed from verified. “Increased risk factors, yes,” Rivera wrote, “verifiable maltreatment, no.” It is unclear whether the death was later discarded from the state’s tally.
And in a particularly unusual case, the agency chose not to verify a suffocation death only because investigators were unsure which parent smothered the 35-day-old baby in bed with them. “Case is being closed with not-substantiated findings of death, due to not being able to determine which caregiver was responsible for the rollover onto the child,” the report said.
Lambert said the department has limited options when a medical examiner’s findings are undetermined and police conclude there is insufficient evidence to make an arrest. In such cases, Lambert said, “DCF also does not have enough evidence to verify. However, we do have authority to remove surviving siblings” when evidence suggests they remain in danger.
Graham, the FSU professor who spent five years on the Department of Health’s statewide fatality committee, which attempts to glean patterns from child deaths, said she was struck by how often the same agency missteps repeated themselves. “You see the same issues over and over and over — and they are all correctable,” Graham said.
“The thing that disturbs me the most is that the energy that is put toward keeping the [“verified abuse”] numbers at a minimum could be better spent looking at the real issues, and preventing future deaths,” Graham said. “To me, that is the biggest crime: We are spending so much energy not looking at cases to make ourselves look good. We should be figuring out how to do this better.”
Florida child deaths by year
Total deaths: 474
Percentage verified: 43
Total deaths: 469
Percentage verified: 35
Total deaths: 428
Percentage verified: 32
Total deaths: 408
Percentage varified: 32
Total deaths: 432
Percentage verified: 10
2014 to date
Total deaths: 358
Percentage varified: 13
Source: DCF child death website