388 - Trends in Ketamine Utilization in Hospitalized Infants with Bronchopulmonary Dysplasia

Publication Number: 1372.388

Meaghan A. Ransom, Monroe Carell Jr. Children's Hospital at Vanderbilt, Nashville, TN, United States; Philip T. Levy, Boston Children's Hospital, BROOKLINE, MA, United States; Matthew Huber, Childrens Hospital of Philadelphia, Philadelphia, PA, United States; Jennifer Sucre, Vanderbilt University School of Medicine, Nashville, TN, United States; Jennifer M. Hale, Monroe Carell Jr. Children's Hospital at Vanderbilt, Franklin, TN, United States; Matthew Kielt, Nationwide Children's Hospital, Columbus, OH, United States; Nicolas A. Bamat, Children's Hospital of Philadelphia, Swarthmore, PA, United States

Background: Ketamine is a commonly used sedative in children, particularly as an adjunct when target sedation depth is unmet. Infants with severe bronchopulmonary dysplasia (sBPD; grades 2-3) often have elevated airway resistance and are bronchodilator responsive. Further, they are often prescribed multiple sedatives during invasive mechanical ventilation. Owing to the combined sedative and bronchodilatory effects of ketamine, this may be a preferable sedative in infants with sBPD. However, use of ketamine in this population is not well characterized.

Objective: To characterize recent temporal trends in the use of ketamine infusion in preterm infants with sBPD and identify patient and clinical characteristics associated with use.

Design/Methods: We conducted a 15-year (2010-2024) retrospective cohort study of extremely preterm infants ( < 28 weeks' gestation) with sBPD using data from PHIS. Our primary predictor variable was time period, evaluated as a categorical variable with 3 progressive 5-year groups. Exposure to ketamine during the study period (36 weeks PMA to discharge or 18 months of age) was identified using CTC codes. Our primary outcome was ketamine infusion, inferred from the presence of ketamine in ≥2 consecutive days. We also described any exposure irrespective of duration. We used descriptive statistics to describe ketamine exposures and multivariable logistic regression to determine association between time period and exposure to ketamine infusion, adjusting for patient characteristics.

Results: We identified 8,888 extremely preterm infants with sBPD with a median GA of 25 weeks (IQR 24-26) (Table 1). 112 (1.3%) received a ketamine infusion at a median PMA of 53 weeks (IQR 46-60) over a median duration of 7 days (IQR 3-14). 814 infants (9.2%) were exposed to ketamine at any time. Ketamine infusions increased over time, particularly from 2020-24 (Fig. 1). A significant association between time period and ketamine infusion remained after multivariable adjustment. Characteristics associated with greater odds of ketamine infusions were pulmonary hypertension (PH), necrotizing enterocolitis (NEC), and tracheostomy (Table 2).

Conclusion(s): In this large registry-based study, ketamine infusions among infants with sBPD remain relatively uncommon but increased over time, particularly since 2020. Clinical characteristics reflecting illness severity, such as PH, tracheostomy, and NEC were associated with increased use. Increasing rates of ketamine utilization observed in this study underscore the need for clinical trials to evaluate the safety and efficacy of ketamine in infants with sBPD.

Cohort Characteristics
Abbreviations: IQR, interquartile range; PMA, post-menstrual age; IVH, intraventricular hemorrhage; NEC, necrotizing enterocolitis

Bivariable and Multivariable Associations Between Cohort Characteristics and Ketamine Infusion Exposurea Between 36 Weeks PMA and 18 Months of Age in Infants with Grade 2 and 3 Bronchopulmonary Dysplasia
Abbreviations: OR, odds ratios; CI, confidence intervals; IVH, intraventricular hemorrhage; NEC, necrotizing enterocolitis
a Ketamine infusion exposure defined as 2 or more consecutive calendar days with exposure
b Estimates of bivariable (characteristic vs ketamine infusion exposure) logistic regression models clustering by center
c Estimates of a multivariable logistic regression model including center as a covariate and clustering by center; all characteristics associated at p < 0.10 with ketamine infusion exposure in bivariable analysis included in multivariable model

The Proportion of Infants with Grade 2-3 BPD Exposed to Ketamine Infusion Between 36 Weeks PMA and Discharge Over Time