Effect of tracheostomy timing in pediatric patients with traumatic brain injury
Irim Salik,Ankita Das,Alexandria F Naftchi,Sima Vazquez,Eris Spirollari,Jose F Dominguez,Vishad Sukul,Dylan Stewart,Augustine Moscatello
DOI: https://doi.org/10.1016/j.ijporl.2022.111414
Abstract:Background: Traumatic brain injury (TBI) is a prevalent cause of disability and death in the pediatric population, often requiring prolonged mechanical ventilation. Patients with significant TBI or intracranial hemorrhage require advanced airway management to protect against aspiration, hypoxia, and hypercarbia, eventually necessitating tracheostomy. While tracheostomy is much less common in children compared to adults, its prevalence among pediatric populations has been steadily increasing. Although early tracheostomy has demonstrated improved outcomes in adult patients, optimal tracheostomy timing in the pediatric population with TBI remains to be definitively established. Objective: This retrospective cohort analysis aims to evaluate pediatric TBI patients who undergo tracheostomy and to investigate the impact of tracheostomy timing on outcomes. Design/methods: The Healthcare Cost and Utilization Project (HCUP) Kids' Inpatient Database (KID), collected between in 2016 and 2019, was queried using International Classification of Disease 10th edition (ICD10) codes for patients with traumatic brain injury who had received a tracheostomy. Baseline demographics, insurance status, and procedural day data were analyzed with univariate and multivariate regression analyses. Propensity score matching was performed to estimate the incidence of medical complications and mortality related to early versus late tracheostomy timing (as defined by median = 9 days). Results: Of the 68,793 patients (mean age = 14, IQR 4-18) who suffered a TBI, 1,956 (2.8%) received a tracheostomy during their hospital stay. TBI patients who were tracheostomized were older (mean age = 16.5 vs 11.4 years), more likely to have injuries classified as severe TBIs and more likely to have accumulated more than one indicator of parenchymal injury as measured by the Composite Stroke Severity Scale (CSSS >1) than non-tracheostomized TBI patients. TBI patients with a tracheostomy were more likely to encounter serious complications such as sepsis, acute kidney injury (AKI), meningitis, or acute respiratory distress syndrome (ARDS). They were also more likely to necessitate an external ventricular drain (EVD) or decompressive hemicraniectomy (DHC) than TBI patients without a tracheostomy. Tracheostomy was also negatively associated with routine discharge. Procedural timing was assessed in 1,867 patients; older children (age >15 years) were more likely to undergo earlier placements (p < 0.001). Propensity score matching (PSM) comparing early versus late placement was completed by controlling for age, gender, and TBI severity. Those who were subjected to late tracheostomy (>9 days) were more likely to face complications such as AKI or deep vein thrombosis (DVT) as well as a host of respiratory conditions such as pulmonary embolism, aspiration pneumonitis, pneumonia, or ARDS. While the timing did not significantly impact mortality across the PSM cohorts, late tracheostomy was associated with increased length of stay (LOS) and ventilator dependence. Conclusions: Tracheostomy, while necessary for some patients who have sustained a TBI, is itself associated with several risks that should be assessed in context of each individual patient's overall condition. Additionally, the timing of the intervention may significantly impact the trajectory of the patient's recovery. Early intervention may reduce the incidence of serious complications as well as length of stay and dependence on a ventilator and facilitate a timelier recovery.