Recurrent child protection involvement in Quebec: A multinomial logistic regression analysis

Children who are subject to a substantiated child protection report are likely to experience this multiple times during childhood. While socioeconomic vulnerabilities may contribute to recurrent involvement, they are not predictive.

The purpose of the present study is to analyze reasons for recurrent child protection involvement among children experiencing recurrence.

This study included all children in the Canadian province of Quebec who experienced an initial child protection substantiation and subsequent case opening from 2002 to 2017 (n = 24,809). The recurrence sample constituted about one third of the total clinical population of children with an initial substantiation (n = 76,176).

Using a multinomial logistic regression analysis, the study draws on anonymized administrative child protection data to predict the likelihood of different initial case reasons at the recurrence stage. Seven categorical reasons for child protection concern constituted the multinomial variable at the initial and subsequent stages. A socioeconomic vulnerability index, age, and sex were used as control variables in the analysis.

Over one third of recurrent cases returned to the child protection system for the same reason as the initial substantiation. The odds of recurring for the same reasons were significant for most categories and socioeconomic vulnerability was not significantly associated with higher odds of recurrence for any reason.

Understanding the odds of re-entry into child protection for the same or different reasons can inform improved policy and practice design and more efficient allocation of funding for both prevention and support to families at risk of recurrence.