Craig Futterman, associate head of pediatric intensive care at Inova Hospital for Children. (By Frank Johnston – The Washington Post)

In 1994 in the District, the percentage of young children's deaths attributed to an "undetermined" cause was 12.4 percent of all deaths – five times the 2.4 percent nationally for the same category.

The D.C. police department is the de facto clearinghouse for all suspicious child death information in the District, including those children who die at Children's Hospital, regardless of where their injuries occurred. Likewise, the D.C. medical examiner's office performs the autopsies for all those children.

A special review of child deaths by the D.C. police homicide unit in late 1996 uncovered 92 cases in 1995 and 1996 – half were children younger than 1 – for which the medical examiner's office forwarded no final cause of death. In more than one-third of the deaths, 34, the preliminary findings stated that the cause of death was "undetermined."

Arden said that for the District to have more infants dying from unknown causes, which he described as a "wastebasket category," than from SIDS "begs for an explanation" and "raises the question of undetected smotherings."

Arden initiated a review of the 34 cases. In 14, records were missing or incomplete. Of the rest, 16 died from natural causes, one was correctly listed as undetermined, and three are now under active review as possible homicides, Arden said. He declined to identify those children.

Chaulette Willis was not on that list of 34, nor was Ebony Brown, whose case Arden reviewed at The Post's request and has now reopened to determine whether the original stated cause of death is correct.

Like Chaulette, Ebony was born prematurely and drug-addicted, medical records show, weighing but 3 pounds, 13 ounces. She didn't leave the hospital until she was 7 weeks old and soon was hospitalized again for a severe diaper rash and a viral infection. But by Sept. 15, 1994, when she was 2 months old and sent home from the hospital again, she was up to 6 pounds, 9 ounces.

Eight weeks later, Ebony was dead.

Her mother told police that she found Ebony unconscious about 8:45 a.m. that Saturday in November. A doctor at Children's Hospital pronounced her dead 45 minutes later. An autopsy found numerous healed and healing fractures, including four fractured ribs and a broken bone in her arm. She also had multiple scrapes on her tiny body and severe diaper rash. She weighed just seven pounds. In two months, she had gained less than a half pound.

The D.C. medical examiner's office ruled at that time that the cause of Ebony's death was undetermined. No criminal investigation.

Chaulette's parents and Ebony's mother could not be located for comment on their daughters' deaths.

Several medical specialists who reviewed The Post data, saw questionable patterns in other death categories.

In Prince George's County in 1994, 11 percent of all deaths of children younger than 5 were attributed to "other newborn respiratory diseases," another red-flag category, those national experts said. Only 2.4 percent of child deaths nationwide fell in that category. The county also accounted for most of the deaths attributed to that cause for all of Maryland in 1994 and 1995.

Without delving into its old cases individually, the Prince George's County Health Department said it could not comment on those newborns' deaths. The Maryland medical examiner's office declined to comment, referring questions to the county.

To get a picture of SIDS deaths, a category that has come under increased suspicion in recent years, The Post broadened its database to include causes of death for the 4,598 children younger than 5 from the District, Maryland and Virginia who died from 1992 through 1995.

In 35 rural Virginia counties and cities over a period of four years, there were 53 cases of SIDS deaths for children from age 28 days through 5 months, including some counties where every death of infants for that multi-year period was listed as a SIDS case and other counties where SIDS was the only cause listed for that age group in several of the years. Most of those counties are in rural areas, clustered in the Appalachian Mountains, the area between Petersburg and the North Carolina line, and along the Eastern Shore and the Chesapeake Bay.

Since 1993, Virginia law has required an autopsy be performed before a death can be attributed to SIDS, and that requirement, said Virginia's chief medical examiner, Marcella Fierro, leaves her confident of the SIDS findings in the clusters. "There are no child abuse deaths in those areas," Fierro said.

In addition to an autopsy – which likely would not distinguish a SIDS death from a smothering – the American Academy of Pediatrics recommends an extensive battery of testing and a detailed investigation, including visiting the scene at which a child was found, before making a SIDS determination. A site visit could reveal discrepancies in accounts of the incident or other evidence that would point to a different cause of death.

Both the National Center for Health Statistics and the World Health Organization have agreed that no death should be ruled as SIDS without the death investigation the pediatrics group laid out. Virginia, Fierro said, is "largely relying" on testing recommended by the Armed Forces Institute of Pathology, which does not require a site visit.

One cluster of deaths currently recorded as accidental involves children suffocated by adults who may have been drunk or using drugs and in their sleep roll over on a child. Such deaths occur throughout the area, with two such fatal incidents in the District alone within a three-week period this summer.

Ryan A. Rainey, an assistant U.S. attorney in the District who for a decade has specialized in child homicide cases here and elsewhere, said prosecutors are looking for a test case to try to show that suffocations amount to a criminal act when they result from an adult's use of illegal drugs or from alcohol abuse.

"What's the difference between suffocating a baby while on drugs or intoxicated and getting behind the wheel of a car while on drugs or intoxicated and killing someone?" Rainey asked.

Confronting the Unthinkable

Even when the unthinkable becomes undeniable, a child's killer may never face any consequences.

Two years ago, a 4-month-old District girl was found dead in bed, and her death preliminarily was attributed to SIDS. When toxicology tests were completed months later, the true cause of death was found to be alcohol poisoning. A District police detective was then sent back to the home, but she said there were too many people sharing it for her to determine who might have fed alcohol to the child. No one has been charged in the case.

A year ago, 14-month-old Devon McCallister, an otherwise healthy boy, an autopsy shows, was found unconscious by his father in his babysitter's apartment on 12th Street in Southeast Washington just after midnight. He was pronounced dead at D.C. General Hospital. An autopsy showed he died of morphine poisoning.

Throughout the day, according to police, Devon had been in the care only of the family and the sitter. No one has been charged in his case.

The autopsy of 8-month-old Alonzo Rajah, who died in 1996 in the District, revealed that he had died of a fractured skull and indicated that a AA battery found lodged in his throat may have been inserted after death. Alonzo's death was ruled a homicide, but his killer has not been found.

It was a second child's killing that raised questions anew about a District child's death that had been ruled an accident.

When Monica Wheeler's battered body was found in the bathroom of the apartment of her mother's boyfriend in September 1997, D.C. paramedics immediately suspected that the 4-year-old had been beaten to death. An autopsy confirmed that, and her death was ruled a homicide. The boyfriend, Michael Lorenzo Tubman, 33, was charged.

Her killing came almost exactly three years after her brother, Andre Wheeler, 2, was found drowned with bruises on his head and neck – also while in the care of Tubman, who at the time gave various accounts of what had happened. The police homicide investigation of Andre's death ended when the medical examiner ruled it accidental. But after his sister's killing, it was reclassified – as a homicide.

Tubman denies killing Andre. When he pleaded guilty to manslaughter in Monica's death, he asked for leniency, blaming Monica's death on his "flipping" out on drugs. Without comment, D.C. Superior Court Judge Mary Ellen Abrecht sent him to prison for the maximum term allowable: 10 to 30 years on the manslaughter charge and 40 months to 10 years on the child abuse count, to run consecutively.

Prosecutors have decided not to pursue a homicide trial in Andre's death because the medical examiner's conflicting rulings undermine the case.

Doctors, police and prosecutors struggle with the emotional toll child death cases exact.

Assistant U.S. Attorney June M. Jeffries, who handles many of the child death cases that reach court in the District, said her work can be "a rather surreal world" in which she may attend a child's autopsy, then be obliged to carry on with routine appointments.

"I feel that you should be screaming, 'This baby is dead.' And instead, we have to go to lunch."

To Bill Hammond, who conducts training sessions nationally for the U.S. Justice Department in how to investigate children's deaths, there is a sad, disturbing truth about child killings.

Too often, he said, there never is a suspicion raised, and a child's passing is forever recorded as the unfortunate consequence of disease or accident.

"The younger the child, the easier it is to murder a child, and get away with it," Hammond said.

Next: Why child killings go undetected