456 - Characteristics of Patients and Families Receiving Aggression Prevention Team Alerts in a Pediatric Hospital

Poster Number: 456
Publication Number: 456.216

Angela Amaniampong, Indiana University School of Medicine, Indianapolis, IN, United States; Kristin Fauntleroy, Riley Hospital for Children at Indiana University Health, Indianapolis, IN, United States; Katherine Krause, Indiana University School of Medicine, Indianapolis, IN, United States; Rachel Peterson, Indiana University School of Medicine, Indianapolis, IN, United States

Background: Aggression Prevention Teams (APTs) were formed to mitigate patient or visitor aggression toward healthcare providers. Little has been understood about the population receiving APT alerts when compared with the general inpatient population in a pediatric hospital setting. 

Objective: Our intent in this study was to determine whether the characteristics of admitted patients receiving APTs at our single institution differed significantly from the inpatient general population, which would signify a disproportionate use of APTs in certain demographics of patients or families. 

Design/Methods: A retrospective chart review of all inpatient APT alerts was conducted between January 1, 2018 and December 31, 2020. Initial data was obtained via records from the hospital-based police department and cross-checked with APT notes in the electronic medical record (EMR). Analyses of hospital admission variables were performed with the primary outcome being whether an APT alert occurred.  

We used mixed-effects logistic regression for all bivariate comparisons of demographics since in some cases there were multiple admissions per patient. We then conducted a mixed-effects logistic regression model with all demographic characteristics included in the model. Similar analyses were conducted for any patient aggression alert and parent aggression alert.


Results: Admissions with Black patients had higher odds (OR=2.5, 95% CI=[1.9, 3.3]) of having an APT alert compared to admissions with White patients. Non-Hispanic patient admissions had higher odds (OR=1.8, 95% CI=[1.1, 3.2]) of an APT alert compared to Hispanic patients. Admissions with managed care insurance had lower odds of any aggression alert than patients with Medicaid or Medicare insurance. Admissions with patients aged 13 to 18 had higher odds of an APT alert than admissions with patients aged 0 to 3. The association of race and insurance status with APT alert remained significant when patient alerts and parent alerts were separated.

Results from the multivariate regression are similar to the bivariate regressions. After adjusting for age, gender, ethnicity and insurance, admissions with Black patients still have a higher odds of APT alerts compared to admissions of white patients. 


Conclusion(s): Disparities exist between the general inpatient population and the patients who receive behavioral emergency alerts regarding (1) race, (2) age, and (3) insurance/payor. Further work is needed to approach how to decrease these disparities and the trauma that may occur for patients and families who are Black and/or on public health insurance.