2018 • Interdisciplinary Topics, Centers and Institutes
Early Prediction of Child Maltreatment Using Electronic Medical Records
Lead Presenter: Samantha Ernst
PI: Erin Dunn
Child maltreatment is common in the US and associated with lifelong health impacts, including increased risk for depression, anxiety, and poor physical health. To date, the main strategies to identify children experiencing maltreatment focus on detecting children in emergency departments or primary care who already have a clinical consequence of maltreatment (e.g., head injury, fracture, behavioral issues), plus assessments by child protection services, clinical interviews, parental screening questionnaires, or retrospective child reports. However, these strategies are limited in ways such as delayed assessment (detection of abuse after the fact), biased reporting (e.g., a parent perpetrator unlikely to report abuse), poor quality of maltreatment screening tools, and lack of a standardized maltreatment detection method. The goal of this study is to describe the limitations of current measures to identify maltreatment and explore the opportunity to utilize electronic medical records to identify children experiencing maltreatment faster and more accurately. We will perform a review of the literature using several databases (PubMed, PsychINFO, and Web of Science) to identify empirical articles on the topics of measurements of maltreatment exposure and the use of electronic medical records in predicting future health outcomes. Results of our literature review will be summarized into three parts: (1) key issues that arise when measuring exposure to child maltreatment, (2) strengths and challenges of using electronic medical records to build practical early prediction alert systems, (3) gaps in the literature to target for future research.