Risk, Not Substantiation, Should Drive Services to Families

A new report from L.A.’s Office of Child Protection (OCP), as recently reported by Daniel Heimpel in The Chronicle, recommends revising current policy to enable the Department of Child and Family Services (DCFS) to offer services to families of children at high risk, even if they do not have a substantiated allegation of maltreatment. But even more importantly, the report points to new thinking about using risk rather than substantiation as the trigger for child welfare services.

The OCP report was established in response to the horrific death of Yonatan Aguilar after four investigations failed to find abuse or neglect in his home. The County Supervisors wanted to know if the County’s Structured Decision Making (SDM) risk assessment tool was at fault in Yonatan’s death.

But as I wrote back in November 2016 (SDM Not the Root of System Failure in Yonatan’s Death), OCP found that the problem was not a misuse of the SDM risk-assessment tool. That tool correctly identified the child as being at high risk for future maltreatment. The problem was the action, or lack thereof, by DCFS.

Specifically, the problem was DCFS practice when an SDM risk level is high, or very high, but a referral allegation is not substantiated. While OCP found agency policy to be unclear, guidance issued to staff indicates that a case should not be opened when the allegation has not been substantiated. As Heimpel pointed out, this conflicts with the recommendation of California’s SDM Manual, which recommends that a case be opened most all high-risk or very high-risk families, and that an explanation be provided when a case is not opened .

OCP recommended that DCFS’ policy for cases where the SDM risk level is high or very high, but referral allegations are unfounded or inconclusive, should be revised so efforts are made to connect the family with voluntary services and supports. Such a policy is in effect in other jurisdictions, including the District of Columbia.

Such a change would be an improvement but might not save many children. That is because, as Heimpel points out, efforts to involve families with maltreatment allegations in voluntary services through differential response have had mixed results.

Part of the solution may lie in an idea that was buried in the middle of the OCP report. OCP analyzed 1,225 referrals investigated by DCFS between 2012 and 2016 where the child was later seriously injured or killed. They found that as in the case of Yonatan Aguilar, more than half of the fatalities and near-fatalities occurred when the allegation was not substantiated.

These findings are consistent with research studies over the years that have found little or no difference in future reports of maltreatment of children who were the subject of substantiated or unsubstantiated reports. For example, studying all infants born in 2006 who were reported to CPS and remained at home, Emily Putnam-Hornstein and colleagues found that the proportion reported again hardly varied between those who were screened out, unsubstantiated, and substantiated; more than half of babies in all of these groups were reported again within five years.

According to OCP, its data “support the growing conversation in the field that allegation substantiation may be inadequate to identify those children who are most at risk for future safety concerns and negative outcomes.” The authors go on to report that they spoke with “child welfare experts who expressed interest in placing more importance on levels of risk (instead of on allegation dispositions) when making case decisions, and in offering services and supports to families that may help to reduce this risk.”

The OCP closes the report by saying it will “partner with DCFS and other key stakeholders to further explore the larger question of the role that assessing risk should play in child welfare,” as well as “the effectiveness of allegation substantiation.”

Moving away from allegation dispositions (and toward level of risk) as a trigger for case decisions makes a lot of sense. While imminent danger should remain the trigger for removal of a child, high risk of future harm should be a trigger for the agency to open a case for monitoring or services.

But what happens if the parent does not agree to participate? In these cases, the agency could bring the family to court to require participation. This is already allowed in Los Angeles and other jurisdictions, although not used as often as needed. In Los Angeles County, the agency can order a “non-detention petition” when it determines that the safety and protection of a child require judicial intervention but it has not removed the child from the home.

Breaking the link between substantiation and services offers one avenue for preventing more at-risk children from falling through the cracks. But unless the agency is able to require parents to participate in services and accept agency supervision, children like Yonatan Aguilar will continue to die.

This column was published in the Chronicle of Social Change on May 30, 2017.