Withdrawal of Life-Sustaining Treatment in Children Managed in Nine PICUs in the United States: A Secondary Analysis of the Death One Hour After Terminal Extubation (DONATE) Dataset, 2009-2021.

OBJECTIVES

To describe medical management surrounding withdrawal of life-sustaining therapy (WLST) in nine U.S. PICUs.

DESIGN

Retrospective, secondary analysis of the "Death One Hour After Terminal Extubation" (DONATE) cohort (2009-2021) assessing usage patterns of: 1) analgesics and sedatives; 2) vasoactive infusions; 3) neuromuscular blockade; and 4) post-extubation respiratory support.

SETTING

Nine U.S. PICUs.

PATIENTS: Children and adolescents 0-21 years old, who had died after WLST (discontinuation of invasive mechanical ventilation).

INTERVENTIONS: None.

MEASUREMENTS AND MAIN

RESULTS

Of 905 patients, 680 (75.1%) died within 1 hour of WLST. Opioids were administered in 721 of 905 patients (79.7%); across sites the range was 68-89% ( p < 0.001). We did not observe a temporal trend. Benzodiazepines were used in 507 of 905 patients (56.0%; site range, 41-66%; p < 0.001), with lower odds of usage per year (odds ratio [OR], 0.95 per year; 95% CI, 0.90-0.99 per year; p = 0.04). Dexmedetomidine was used in 140 of 905 patients (15.5%; sites range, 4-21%; p = 0.002), with greater odds of usage per year (OR, 1.16 per year; 95% CI, 1.05-1.27 per year; p = 0.004). Vasoactive infusions were discontinued in 458 of 520 patients (88.1%) receiving this medication (site range, 59-100%; p < 0.001), with greater odds of discontinuation per year (OR, 1.15 per year; 95% CI, 1.04-1.26 per year; p = 0.007). Neuromuscular blockade was used in 46 of 905 patients (5.1%; sites range, 0-13%; p < 0.001), with greater odds of usage per year (OR, 1.23 per year; 95% CI, 1.08-1.40 per year; p = 0.002). Use of any post-extubation respiratory support occurred in 50 of 905 patients (5.5%), and we did not identify an association with site or year-on-year trend.

CONCLUSIONS

The 2009-2021 DONATE dataset shows substantial institutional and temporal variability in WLST practices across our nine collaborating PICUs in the United States. Future studies should focus on understanding the drivers of variability to improve the consistency and quality of end-of-life management.