Introduction: Little is known about how critically ill children with chronic respiratory technology dependence (rTD) compare to other children with critical illness. We sought to evaluate the association of pre-existing respiratory technology dependence (rTD) with clinical characteristics and in-ICU outcomes. Methods: Multicenter cross-sectional study of encounters using Virtual Pediatric Systems, a large, contemporary cohort of pediatric intensive care unit (PICU) admissions. We included patients < 21 years discharged from an index admission 4/2017-12/2020. We defined rTD as a Functional Status Scale (FSS) respiratory score of ≥3, indicating tracheostomy and/or use of non-invasive/invasive ventilation. FSS was assessed at ICU admission (reflecting baseline prior to acute illness) and ICU discharge. We used univariate analysis to evaluate the association between rTD and clinical characteristics and outcomes. All p-values were < 0.001 unless noted. Results: Of 33,945 index encounters, 2,016 (6%) children had rTD and 31,929 (94%) did not. Children with rTD were typically admitted for respiratory/ENT reasons (51% v. 30%) and less commonly for neurologic or trauma reasons (9% v. 13% and 3% v. 15%, respectively). They were more likely to be admitted from a procedural area (32% v. 25%). Median severity of illness did not differ (p=0.8). Children with rTD had higher mortality (3% v. 2%), longer median lengths of stay (2.4 days IQR 0.9-6.6 v. 1.2 0.7-2.8), and were more likely to be discharged home instead of the ward (55% v 24%). They were less likely to develop new morbidity during the index admission (3% v. 8%,) and more likely to survive with improvement in functional status (13% v. 0%). Mean change in total FSS was -0.34 (SD 2.01) in those with rTD and 0.56 (1.73) in those without. Resolution of rTD during the admission occurred in 277 patients (14%). Of those without rTD, 748 (2.3%) developed new rTD. Conclusions: In-ICU outcomes of PICU patients with rTD are more likely to show improvement, with 14% of patients having resolution of rTD during the index admission. Improved outcomes may reflect procedural nature of many admissions, which may represent interventions to optimize function or address subacute on chronic issues. Conversely, development of new rTD was seen in a small percentage of patients.
Heneghan et al. (Sun,) studied this question.