At any given time, some 19,000 children are in foster care in Ontario, placed there for their protection by one of the province’s 53 Children’s Aid Societies (CAS).
While CAS caseworkers write meticulous reports on the children who pass through the system, the details contained in the thousands of files are not aggregated. As a result, the provincial Ministry of Community and Social Services lacks a comprehensive statistical snapshot on which to base child welfare policies and services.
Shlonsky and his research team are striving to close that gap. Through the creation of the Ontario Child Abuse and Neglect Data System (OCANDS), they will provide the CAS agencies with their first unified database, an advance that could result in improved child welfare policies and services.
While this system will be developed over the next two years, Shlonsky’s database work has already had a significant impact in evaluating the province’s risk-assessment tool. Every family investigated for child abuse, the CAS caseworker has to use a tool designed to identify elements in a family’s story that suggest how likely their child is to return to foster care.
In 2000, Ontario began using a tool called the Ontario Risk Assessment Model (ORAM) for this purpose. Shlonsky and his research team examined the predictive ability of the ORAM and disputed its validity and reliability. (Not coincidentally, the rates of child maltreatment almost tripled between 1998 and 2005.)
In 2007, Community and Social Services heeded the critique and replaced the ORAM with a new risk-assessment tool that is proving to have a much higher probability of forecasting correctly which children will be re-abused.
“For the first time, we have a risk-assessment tool in Ontario that has been validated,” says Shlonsky. But he did not stop there. Since the tool was imported from California, the OCANDS team is adapting it to fit the Ontario context. “We’re combining computer science with social research to optimize the predictive value of the tool.”
There’s more valuable research to come. The University of Toronto team will structure the data emerging from OCANDS to determine which children are coming into the system, how long they stay in foster care, how long they are in a particular foster home and whether they return to their family or become a ward of the province.
“Currently, if I’m a casework and I look at a particular family to see whether it has a history with the CAS, that information will come up on my screen,” says Shlonsky. “But if I’m a manager and wish to know, say, the median length of time for family reunification for all cases in a given year by age, I typically can’t do that.” That’s because the existing system is designed to manage information at the case level. It’s not designed to pool that information into a larger report to indicate trends either for the agency or for the province as a whole.
Once complete, OCANDS will allow each Children’s Aid Society to query its own data, as well as those of other agencies. The idea is to mine the aggregated data for useful insights that can guide child welfare agencies and ministry policy-makers. For example, if a high proportion of the children placed in foster care are babies, then child welfare services can reflect that.
Similarly, the unified database will allow officials to “map” the maltreatment landscape in the province. If certain geographic areas have a relatively high concentration of child-abuse cases, then more social workers and services can be deployed in those areas.
OCANDS won’t solve all the data-mining difficulties. If a family with an abused child moves from one CAS jurisdiction to another, the new system won’t be able to track that movement any more than does the existing system. Such transient families, however, account for only a small percentage of CAS clientele.
The project’s CFI-funded resources are being housed in a newly established OCANDS laboratory in the social work building at University of Toronto. The facilities include offices, a secure data-processing and data-storage area and a large, open laboratory.
The configuration will promote a clustering of resources, the protection of sensitive data and the development of a community of child welfare researchers, including professors, graduate students and agency and government partners. In the end, Shlonsky hopes the new system will lead to more innovative policies and services to aid one of the province’s most vulnerable populations.