Differential response and traditional protection investigation: a comparison of two child welfare investigation streams

Abstract

INTRODUCTION

Determining an appropriate threshold for which children and families should receive child welfare services has been widely debated in literature. Workers are faced with multiple child welfare orientations when determining the appropriate level of intrusiveness during child maltreatment investigations [3]. Parental rights, family preservation, and safety of the child are some of the competing tensions in the child welfare model that investigating workers must balance [2,9]. Differential Response (DR) model is a flexible approach to child maltreatment investigations developed to provide a range of customized investigative streams incorporating family-centered, and strengths-based practices [4]. This model is a less intrusive approach to investigations for low-risk cases that do not present imminent risk to the child, compared to Traditional Protection Investigations (INV) [5,7]. There is a high degree of jurisdictional variation in screening, intake and administration of DR models resulting in a disparity of findings throughout child welfare research [4]. In 2002, the Alberta Response Model was created to allow child welfare agencies to provide flexible investigative streams to families based on the type and severity of child maltreatment [1,8]. The Family Enhancement Program works as part of the Alberta Response Model to provide children and families with a DR investigation stream [1]. High-risk cases are streamed to INV while lower risk families who are willing are streamed to the Family Enhancement Program [1]. This article will inform practitioners’ understandings of DR model by providing a provincial profile of DR streamed investigations while comparing child, household, maltreatment characteristics and case factor associated with each child welfare investigative stream based on the Alberta Incidence Study of Reported Child Abuse and Neglect 2008 (AIS-2008) [6].

METHODS

This analysis is based on secondary analysis of the AIS-2008 dataset. The AIS-2008 is the second cycle of a provincial study that examines incidences of reported child abuse and neglect in Alberta for children 17 years and younger. This paper reports on findings based on a total weighted number of 26,957 child investigations that were noted as either DR streamed investigations, or INV streamed investigations. A total of 6,666 child investigations noted use of DR model, while an additional 20,291 child investigations noted use of INV model. Select comparisons of child, household, case and maltreatment factors from two child welfare investigation streams are presented in this analysis. Bivariate analyses and Pearson’s chi-squared test were used to compare each investigative stream and associated factors. For further information on methodology of the AIS-2008 refer to Chapter 2 of the full report [7].

RESULTS

Figure 1 shows primary categories of all reported child maltreatment incidents and risk investigations streamed to DR and INV. Significant differences in child investigations were noted for younger children. Eleven percent of DR streamed investigations involved children who were less than 1 year old, and 8% of INV streamed investigations. Caseworkers were asked to report on nine caregiver risk factors, such as alcohol and/or drug abuse, cognitive impairment, and/or victim of intimate partner violence (IPV) [7]. While differences between the two investigative streams for caregiver risk factors would be expected, there were little differences noted.

A risk investigation refers to situations where a specific incident of maltreatment has not yet occurred, however circumstances indicate that there is a significant future risk of maltreatment [7]. DR child investigations more frequently noted risk investigations than INV child investigations. For DR investigations, neglect was noted most frequently (38%), followed by exposure to intimate partner violence (23%), physical abuse (10%), and emotional maltreatment (9%). Ten percent of DR investigations noted emotional harm requiring treatment, compared to 16% of INV investigations. One percent of DR investigations noted physical harm severe enough to require medical attention, and 2% of INV investigations. The percentage of DR child investigations resulting in a formal child welfare placement was significantly less (5%) than INV child investigations (10%). Twenty-five percent of DR investigations remained open for ongoing services compared to 32% of INV investigations. DR investigations less often resulted an application to child welfare court (4%), than INV streamed investigations (13%).

Figure 1. Primary Categories of All Reported Child Maltreatment Incidents and Risk Investigations in Alberta in 2008 by Child Welfare Investigation Streams. Percentages of type of maltreatment includes total number of child investigations (n=26,957) streamed to differential response (n=6,666), and traditional protection investigation (n=20,291).

DISCUSSION AND CONCLUSIONS

This secondary data analysis examined 26,957 child maltreatment investigations from the AIS-2008 dataset. The increased flexibility of DR model enables workers to balance some of the competing tensions in child welfare. DR model allows child welfare agencies to provide less intrusive approach to investigations for low-risk investigations, compared to INV model which is intended for investigations presenting greater severity of harm. This exploratory analysis examined differences in characteristics between children and families of child maltreatment investigations streamed to DR approach and INV approach. Future research should examine long term outcomes of investigations streamed to DR and INV. Exploring long term outcomes will provide greater insight into the effectiveness of DR model child maltreatment investigations.

LIMITATIONS

While the AIS-2008 dataset provides a unique opportunity to examine the child welfare response to reported maltreatment in Alberta, a number of considerations for this secondary analysis must be made when interpreting these findings. The AIS-2008 dataset; 1) only tracked reports investigated by child intervention services and did not include reports that were screened out, only investigated by police, and never reported; 2) is based on the assessments provided by the investigating child intervention workers and could not be independently verified; 3) is weighted using annual estimates which included counts of children investigated more than once during the year, therefore the unit of analysis for the weighted estimates was a child investigation; 4) as weighted estimates provided some instances where sample sizes were too small to derive publishable estimates [7].