588 - Tied Up in Bias? Factors Associated with the Use of Restraint during Pediatric Behavioral Health Hospitalizations
Friday, April 28, 2023
5:15 PM – 7:15 PM ET
Poster Number: 588 Publication Number: 588.126
Samantha Crowley, Baylor College of Medicine, Houston, TX, United States; Andrea L. Dean, Baylor College of Medicine/Texas Children’s Hospital, Houston, TX, United States
Assistant Professor Baylor College of Medicine/Texas Children’s Hospital Houston, Texas, United States
Background: Pediatric behavioral health (BH) hospital encounters are surging. Use of behavioral (i.e physical) restraints, used to prevent injury from a patient to self, others, or property, is associated with mental and physical trauma and, rarely, death. In adults, studies show that risk factors for restraint use include Black or African-American race, male sex, non-Hispanic ethnicity, lack of private insurance, and/or homelessness. While the 1998 Hartfort Courant exposé demonstrated the unregulated restraints use and poor outcomes in children, little research has focused on this topic.
Objective: To determine factors associated with use of behavioral restraints in pediatric BH encounters. We hypothesized that non-white race would be associated with higher odds of restraint use.
Design/Methods: We performed a retrospective chart review of BH encounters emergency department (ED) and acute care from 1/1/2019 to 7/21/21 from a single, urban, university-affiliated, freestanding children’s hospital in the Southeast. Encounters eligible for review were identified by the presence of “Sitter to Bedside” order. Sex, language, BMI, race, ethnicity, length of stay, and age were extracted into the data set for analysis. Inclusion criteria included ages 5-18. Exclusion criteria included admission to the intensive care unit or the use of medical restraints. Data was analyzed and compared using odds ratio (OR) within the factors of interest. 95% confidence interval (CI) was used to identify significance.
Results: 4071 BH encounters were identified using the “Sitter to Bedside” order. Restraints were applied in 42 (1%) of these encounters. Factors associated with higher odds ratio were male sex (OR 3.13, CI 5.86-1.67) and non-English language (OR 2.72, CI 5.26-1.41). Ethnicity, age, and race did not show significantly higher OR, both when analyzed by non-white and Black race.
Conclusion(s): It is essential to understand the impact of racial and other internalized biases on restraint use in children and adolescents. This brief study was limited: it was a single institution, the “Sitter to Bedside” order may have over or under included encounters, and the use of odds ratio did not
adjust for length of exposure. However, the results provide an interesting contrast to those of Nash et al, whose 2021study showed higher odds ratio of restraint use in Black, male, and/or publicly-insured children in New England EDs (JAMAPediatr.) Both studies highlight that, as pediatricians tackle the national mental health crisis, research in various settings must be done to mitigate potentially unjust practices in the field.
