CO- State Reviews 13 Child Deaths, Finds Systematic Errors

Policy Changes Planned To Protect At-Risk Children

4-16-2008 Colorado:

DENVER -- A state review of 13 recent child deaths released Tuesday found systematic problems like social service workers not interviewing suspected child abusers, not checking backgrounds including sex offenses and not entering the information into the statewide database.

Many of the findings mirror an 11-month CALL7 investigation showing flaws in Colorado’s child protection system.

The report recommends that the Colorado Department of Human Services in the next 90 days require more detailed documentation of interviews and make sure county departments do more timely documentation of findings in an abused child’s home.

In El Paso County, for example, two cases did not have the proper documentation in the state computer that allows workers to track families if they move from county to county.

"It is absolutely critical if we are going to prevent child abuse from occurring one county to another county to another county, that that documentation be there,” said CDHS executive director Karen Beye. “We, as a state system, need to do a much better job.”

The report recommends that the human service investigator interview the perpetrator and all witnesses. In the death of 3-month-old Luz Valdez, the suspect in the case was never interviewed, the CALL7 investigation found. Similar problems were cited in cases from around the state, like the death of a 4-year-old in Otero County last year.

“I think there should be a basic understanding of what are the core pieces that you have to do,” Beye said. “And yes, I think interviewing everyone in the family is one of those.”

In one case in Weld County, workers did not follow up on a referral that the child was in danger and another in the same area found the workers failed to investigate prior dangers in their risk assessment.

The report found that 70 percent of the 13 child deaths in 2007 that CDHS workers reviewed involved previous domestic violence and 50 percent of the families had histories of drug abuse.

Also, 46 percent involved Hispanic families, prompting Beye to say that there needs to be more emphasis on finding workers with Spanish speaking skills.

The report also recommended that county workers check national and state databases of registered sex offenders. While not linking the recommendation to a specific case, the move comes after Denver County workers failed to track a sex offender who was later named a suspect in a child's death. That fact was reported last year as part of the CALL7 investigation.

Denver workers checked the state database but not county records, which would have showed where the sex offender was living.

State officials said their recommendation to check the state and national databases are a start until they can get money to allow workers to check more complete and costly databases. However, county databases are available free of charge.

In the long term, the state will increase training for human service workers to make sure they know the right questions to ask.

“In order for the counties to adequately do their jobs, policy instruction, policy requirements must be clear, must be consistent and must be monitored,” Beye said.

Overall Beye couldn’t say which of the child deaths could have been prevented but conceded that proper procedures might have saved lives.

“I certainly look at them in terms of are there things that could have been done that might have prevented this child’s death, and in several of those cases the answer for me was ‘yes,’ ” Beye said. ..more.. by Arthur Kane and John Ferrugia , CALL7 Investigators


Colorado: Child Maltreatment Fatality Report 2007

Executive Summary
Colorado is one of 13 states with a state-supervised and county-administered system. In its most concrete definition, county administration is statewide social services programs at the local level. State supervision is indirect management as exercised through program development, practice standard development, workload standard development, model office design, rule promulgates, technical assistance, monitoring, program evaluation and performance improvement plans.

In response to the increase in child maltreatment fatalities where the victim and families were previously known to Child Protective Services (CPS) agencies, the Executive Director of the Colorado Department of Human Services ordered a review of child maltreatment fatalities focused on identifying any commonalities and making recommendations for improvements in the system based upon those findings. This review specifically examined 13-recent child maltreatment fatalities that occurred in Colorado where CPS had prior involvement in the last five years. In order to determine systemic issues, information from these 13 reviews was then combined with data regarding all child maltreatment fatalities occurring in Colorado over the past five years, as well as data at a national level and from research conducted within the child welfare field. Findings were categorized across four major areas and are summarized here by each category.

1. Child Characteristics
The majority of child maltreatment fatality victims in Colorado over the past five years tend to be Caucasian (ranging from 34% to 51%), with a large percentage claiming Hispanic ethnicity (ranging from 27% to 39%). While not significantly different, slightly more than half the victims were female. Lastly, approximately 40% of child maltreatment victims in Colorado were infants, with approximately 90% of the victims being under the age of 5.

2. Parent Characteristics
Parents of victims tend to have their own history of prior involvement with CPS. They also tend to be younger. For example, most parents were in their early 20’s at the time of birth and death of the victim.

3. Environmental/Situational Characteristics
Overall, almost 70% of the families in this fatality review had some history of identified Domestic Violence, while 54% had experienced Substance Abuse issues.

Birth order appeared to be associated with fatalities. For example, over the past five years a range from 43% to 51% of child maltreatment fatality victims were only children. Of those with siblings, it ranged from 59% to 93% of the time the victim was the youngest child in the family. Related to this, the number of children and adults in the household tended to be associated. For example, neglect (both fatal and non-fatal) tends to occur in families with more children, while abuse (again, fatal and non-fatal) may be more likely to occur in families with fewer children. Also, high family mobility is often associated with child maltreatment fatalities both nationally, and in Colorado. As Colorado is a county administered system, this means that every time a family moves to a new county, a new agency becomes responsible for service provision and safety management. Family compositional characteristics were also identified. Specifically, 30% of the 13 cases reviewed had children involved in referrals that occurred while they were living with different families or family members, 38.5% had multigenerational involvement with CPS, and 46.2% were single female head of household. Finally, additional family stressors were found to be involved in a substantial portion of the 13 fatalities reviewed, including Substance Abuse issues (53.8%), Domestic Violence (69.2%), and Child’s Medical Issues (38.5%).

4. Systemic Characteristics
Data integrity within the Statewide Automated Child Welfare Information System, known as Trails, as well as Colorado’s process for tracking child fatalities were found to be inconsistent. For example, the ACCESS database of child maltreatment fatalities kept by the Division of Child Welfare did not match the Trails system. In addition, over the past five years, an average of 37.5% of victims did not have a date of death recorded in Trails. Also, for those victims listed as fatalities both in Trails, and on the ACCESS database, over the last five years the dates did not match approximately 30% of the time.

Over the past decade, two Agency Letters – CW02-215I dated May 30, 2002 and CW00-25A dated May 11, 2000 – summarized findings from two previous child fatality reviews. Lack of communication between agencies was the systemic factor. This includes communication between county agencies when families move and responsibility for service provision and safety management shift to a new county, and communication on new rules, policies, and oversight from the Division of Child Welfare to the county agencies.

Workers’ characteristics were also examined for those workers involved in the 13 fatalities, and for the most part found to be in compliance. Most workers, by self-report, had the requisite level of education, background checks, and training as required by Volume VII. At a more general level, however, there is a need for increased funding for both CORE training as well as training for the safety assessment. Despite the training being full for all of 2008, training providers reported a much greater demand. This was mostly related to high turnover requiring new workers to complete the required training. The state does not have a continuing education model designed especially for supervisory needs. Also, while an estimated 839 child welfare professionals have been trained in the new safety model, there are still many professional who have not been trained. The Department is unable to give an exact number because there is not a centralized automated method for tracking.

Child welfare professionals interviewed as part of this process indicated that many of the gaps and issues identified were driven primarily by overburdened staff. However, due to the lack of any rigorous workload studies, it is difficult to determine what the current workload level is on average, and impossible to use it as a method for assigning cases.

A number of Volume VII regulations were found to be incomplete, inconsistent with other policies, or simply lacking definitions of key concepts. In addition, numerous areas of practice were identified where Volume VII regulations were not being applied accurately or consistently across county departments.

5. Recommendations
This report concludes with a list of recommendations intended to address many of the issues identified. Specifically, the list is broken into short-term recommendations to be implemented within 90 days of the publication of this report, and long-term recommendations that will require more time to study and craft solutions and/or implement statutory changes, budget requests, and rules and regulations. ..more..