The killing of Joseph Bennett set in motion one of the most tragic and painful chapters in Tampa Bay history.
Four-year-old Joey was shot in the head with a rifle. His mother's boyfriend, Hank Earl Carr, embarked on a daylong rampage after police tried to question him, killing three lawmen before shooting himself in front of a hostage.
There were warning signs: In the months before the killings, investigators for the Department of Children & Families (DCF) were told that Carr had dropped Joey on his head for soiling his pants and beat the boy at a doctor's office in front of horrified patients in a waiting room.
But if there were lessons in Joey's May 1998 death, one would be hard-pressed to find them.
Sign Up and Save
Get six months of free digital access to the Miami Herald
Joey's case is among more than 80 for which the DCF has no written report.
Such death reviews as are done are often deficient, experts say.
Federal law requires the state to do a death review - a comprehensive internal analysis intended to prevent other children from dying - for every case in which a child dies of maltreatment. The state also is required to study cases in which children nearly die - though this is almost never done.
"This was a myriad of opportunities where we could have learned how to prevent these horrible deaths, " said Nancy Barshter, a member of the Statewide Child Abuse Death Review Team for three years, who has analyzed hundreds of Florida child fatalities. "We have not only lost the children, we've lost the chance to learn from their deaths."
Under former Secretary Kathleen Kearney, a team of troubleshooters, including the agency's chief doctor and quality assurance specialist, was sent to study particularly troubling child deaths immediately. But the Rapid Deployment Team was ordered to put none of its findings in writing.
Agency spokesmen said the policy is being reexamined by new Secretary Jerry Regier.
Bob Brooks, Regier's spokesman, said he does not know why so many child deaths have gone unstudied. "I'm not saying there are no reasons why. I'm saying I haven't found any."
Ideally, Brooks said, written reviews would be done on every death, including a look at what problems occured with the investigations, whether staff discipline is necessary, and what can be learned from previous mistakes.
"Like any organization, we are not perfect, " Brooks added. "This may be an area we need to look into."
Even when death reviews are done, the experts say, the investigation is often thin and underreported.
"There is little attempt at self-accountability, " Barshter said. "It's a tragedy how poorly some are done, if done at all. I call them 'drive-by investigations.' "