Neuromuscular Blocker Use in Critically Ill Children: Assessing Mortality Risk by Propensity Score-Weighted Analysis

Abstract

Objectives: We aim to describe the use of continuous infusion of neuromuscular blocking agents in mechanically ventilated critically ill children and to test its association with in-hospital mortality.

Design: Multicenter, registry-based, observational, two-cohort-comparison retrospective study using prospectively collected data from a web-based national registry.

Setting: Seventeen PICUs in Italy.

Patients: We included children less than 18 years who received mechanical ventilation and a neuromuscular blocking agent infusion from January 2010 to October 2017. A propensity score-weighted Cox regression analysis was used to assess the relationship between the use of neuromuscular blocking agents and in-hospital mortality.

Interventions: None.

Measurements and main results: Of the 23,227 patients admitted to the PICUs during the study period, 3,823 patients were included. Patients who received a continuous infusion of neuromuscular blocking agent were more likely to be younger (p < 0.001), ex-premature (p < 0.001), and presenting with less chronic respiratory insufficiency requiring home mechanical ventilation (p < 0.001). Reasons for mechanical ventilation significantly differed between patients who received a continuous infusion of neuromuscular blocking agent and patients who did not receive a continuous infusion of neuromuscular blocking agent, with a higher frequency of respiratory and cardiac diagnosis among patients who received neuromuscular blocking agents compared with other diagnoses (all p < 0.001). The covariates were well balanced in the propensity-weighted cohort. The mortality rate significantly differed among the two cohorts (patients who received a continuous infusion of neuromuscular blocking agent 21% vs patients who did not receive a continuous infusion of neuromuscular blocking agent 11%; p < 0.001 by weighted logistic regression). Patients who received a continuous infusion of neuromuscular blocking agent experienced longer mechanical ventilation and PICU stay (both p < 0.001 by weighted logistic regression). A weighted Cox regression analysis found the use of neuromuscular blocking agents to be a significant predictor of in-hospital mortality both in the unadjusted analysis (hazard ratio, 1.7; 95% CI, 1.3-2.2) and in the adjusted one (hazard ratio, 1.6; 95% CI, 1.2-2.1).

Conclusions: Thirteen percent of mechanically ventilated children in PICUs received neuromuscular blocking agents. When adjusting for selection bias with a propensity score approach, the use of neuromuscular blocking agent was found to be a significant predictor of in-hospital mortality.