The initial report already had been published months past the deadline outlined in state law.
Pennsylvania – After reading the state’s first round of child abuse statistics, a Pennsylvania resident most likely walked away with a basic understanding of the children who died or nearly died last year.
Boys had died at a rate of exactly half of the listed girl victims, with 10 male victims and 20 female. Near-fatalities skewed toward boys, with 34 male victims to the 31 female victims.
The problem is that all of those interpretations were wrong — a result of inaccurate data.
The state Department of Human Services published its revised 2014 annual child abuse report on Thursday, after removing the information from its website in July to correct errors in the initial data. The initial report already had been published months past the deadline outlined in state law.
Two more boys died and two fewer girls. And near fatalities were an even split at 33 for each gender, according to the corrected data.
Those found responsible for the fatalities, on the other hand, were not an even breakdown between males and females. Instead, four more women were found to be suspects, and six more suspects were listed overall.
According to the report’s explanation for the errors: “Changes to the Child Fatality/Near Fatality Analysis, including updates to Figures, C, D, E, F, G, L and M, are due to using the CY-48 Child Abuse Investigation Form in place of the previously used CY-921 Fatality/Near Fatality Data Collection Tool. The CY-921 was not completed for all fatality/near fatality reports from 2014.”
Changes also were made to at least five additional charts.
Meaning, the investigation form, as the reported noted, was accurate, but “the ancillary methods for tracking and obtaining additional information on the fatality/near fatality contained the inaccurate information,” according to an email from Kait Gillis, spokeswoman for the state Department of Human Services.
In a few cases, the employee who wrote the summary incorrectly documented the child’s gender, resulting in the misinformation, she said.
A false narrative
The erroneous report featured incorrect data in charts meant to develop an illustration of child victims across the state during 2014. The data ranged from the number of children of a particular sex and age to descriptions of their suspects. The incorrect data resulted in skewed totals and county-by-county breakdowns.
Child advocates and media outlets noticed discrepancies after the state made changes to the initial report, prompting the department to remove the document altogether.
For Cathleen Palm, founder of the state’s Center for Children’s Justice, the changes were anything but insignificant, especially when it came to the child fatality and near-fatality analysis. The changes, she said in an email to PennLive, actually blew her away.
Palm said her review of the amended report revealed that “virtually all of DHS’ analysis about how many kids died/nearly died, the sex/age, life circumstances and who killed them is so different in the revised report.”
And, what if concerns hadn’t been raised and DHS hadn’t pulled and amended its report?
“We would have had a wholly false narrative about what the lethal and near-lethal toll of child abuse is in PA,” she said.
Cross-checking the data
Cathy Utz, the department’s deputy secretary for children, youth and families, led an internal team to review and correct the errors in the first report, Gillis said.
The team looked at each fatality and near fatality report separately. For each case, it cross-checked records, including the child abuse investigation form, internal tracking log, the child fatality/near fatality summaries and the county data collection tool.
The team conducted additional research to find the correct information if the records showed data inconsistencies, according to Gillis.
Additional research meant reviewing forms, checking statewide databases and “if necessary contact was made with the county or regional office to verify the data,” Gillis said.
The state then corrected the information and updated the annual child abuse report. DHS officials also apologized for the misinformation.
“I apologize for any inconvenience the inaccuracies in the report may have caused,” Gillis said in the email to PennLive following the reissuance.
Tina Phillips, director of training at Pennsylvania Family Support Alliance, said her organization is still reviewing the report but “is pleased that the Department took the time to ensure the accuracy of the entire report.”
Palm said that while the state took an important step in pulling the initial report, acknowledging the errors and correcting the data, it shouldn’t be praised too strongly for its apparent transparency.
“I think, though, that we can’t permit too much celebration of ‘transparency’ when today we still live in a world where data and so much of child protection is housed, analyzed, reported and then ‘corrected’ entirely by the same entity,” she said, adding that a push for real-time data is crucial to a better understanding of child abuse.
Palm suggested that child advocates carefully review the corrected report, which appears to have minimal shifts through the documents in areas like the number of suspects, and added that an even closer review should be made of the method by which information is collected statewide.
“We should not stop asking the more critical question of how, when and with what level of independent check and balance will data be more transparent and reconciled in order to permit the public to have confidence in [how] PA protects our children,” Palm said.
The Pennsylvania Family Support Alliance also is looking to future reports, where it will hone in on different reported areas, such as whether lowering the threshold of what is considered child abuse results in an increase in the state’s reported abuse substantiation rate.
The results of the report could trigger widespread changes in child welfare statewide.
“We hope that as much care is taken for the development of the next annual abuse report, which will be the first to reflect the changes to the Child Protective Services law,” Phillips said. “Its accuracy will be essential in evaluating the impact of the legislative changes.