Addressing Decision Errors in Child Maltreatment Investigations: Can Practice Be Improved?

Background and Purpose.  Child protection decision making typically occurs in highly stressful, complex environments.  Mistakes in judgment leading to inaccurate assessments of risk are inevitable and occasionally result in tragic case outcomes. While individual caseworkers may be blamed for poor decision making, mistakes are likely to result as much from problems with organizational processes as individual misjudgments. Research question: How can understanding sources of error provide a foundation upon which to build better organizational processes that will reduce the likelihood of similar mistakes occurring in the future and enhance child safety? Objectives: (1) identify decision errors and organizational weaknesses contributing to serious child injury or death in families known to a state child welfare system; (2) identify and implement strategies to improve child protection investigations; 3) monitor implementation, uncover and begin to address barriers to change.

            Methods. Mixed methods. Ten cases of child maltreatment fatality and serious injury investigations were selected for in-depth analysis to identify common patterns of error in child protection decision making. Root cause analysis was used to identify multi-level individual and organizational factors contributing to the tragic outcomes. A stratified random sample of 300 child maltreatment investigations involving cuts, welts, and bruises was examined to determine the prevalence of errors and inform the development of intervention to correct problematic practice.  Subsequent random samples examined between 6 months and 1 year later provided data on implementation and barriers to change.

            Results. Analyses identified staff level factors leading to decision errors, including failure of child protection investigators to use multiple sources of information, failure to recognize cumulative risks, failure to consider other adults with access to the child, failure to access expertise, and failure to establish or monitor a realistic safety plan. Higher level organizational factors contributing to errors in child maltreatment investigations included staff shortages, supervision problems and informal incentives for practice shortcuts, communication problems, lack of support for good work, and a defensive work environment, among other issues. Based on findings from the analyses, initial steps to address identified problems included revision of the risk assessment and safety planning protocol, development of checklists to prompt better communication between child protection workers and medical/mental health professionals, and training to complement their use for child protection staff and supervisors across the state. These first efforts to reduce errors and improve child protection have met with limited success.

            Conclusions and Implications. This experience represents an innovative use of root cause analysis in human services and an attempt to apply some High Reliability Organization (HRO) principles to reduce the likelihood of tragic case outcomes.  Our findings suggest that child risk/safety assessments could be improved in a number of specific ways across organizational levels in the child welfare system. However, significant organizational barriers persist and prevent anything more than an incremental approach to intervention.  While initial efforts to implement changes have had mixed results, this project influenced recent state legislation establishing a child welfare error reduction team. Next steps for the error reduction initiative are currently being considered.