Impact of real-time ultrasound guidance on complications of percutaneous dilatational tracheostomy: a propensity score analysis

INTRODUCTION Recent studies have demonstrated the feasibility of real-time ultrasound guidance during percutaneous dilatational tracheostomy, including in patients with risk factors such as coagulopathy, cervical spine immobilization and morbid obesity. Use of real-time ultrasound guidance has been shown to improve the technical accuracy of percutaneous dilatational tracheostomy; however, it is unclear if there is an associated reduction in complications. Our objective was to determine whether the peri-procedural use of real-time ultrasound guidance is associated with a reduction in complications of percutaneous dilatational tracheostomy using a propensity score analysis. METHODS This study reviewed all percutaneous dilatational tracheostomies performed in an 8-year period in a neurocritical care unit. Percutaneous dilatational tracheostomies were typically performed by trainees under guidance of the attending intensivist. Bronchoscopic guidance was used for all procedures with addition of real-time ultrasound guidance at the discretion of the attending physician. Real-time ultrasound guidance was used to guide endotracheal tube withdrawal, guide tracheal puncture, identify guidewire entry level and confirm bilateral lung sliding. The primary outcome was a composite of previously defined complications including (among others) bleeding, infection, loss of airway, inability to complete procedure, need for revision, granuloma and early dislodgement. Propensity score analysis was used to ensure that the relationship of not using real-time ultrasound guidance with the probability of an adverse outcome was examined within groups of patients having similar covariate profiles. Covariates included were age, gender, body mass index, diagnosis, Acute Physiology and Chronic Health Evaluation II score, timing of tracheostomy, positive end-expiratory pressure and presence of risk factors including coagulopathy, cervical spine immobilization and prior tracheostomy. RESULTS A total of 200 patients underwent percutaneous dilatational tracheostomy during the specified period, and 107 received real-time ultrasound guidance. Risk factors for percutaneous dilatational tracheostomy were present in 63 (32%). There were nine complications in the group without real-time ultrasound guidance: bleeding (n = 4), need for revision related to inability to ventilate or dislodgement (n = 3) and symptomatic granuloma (n = 2). There was one complication in the real-time ultrasound guidance group (early dislodgement). The odds of having an adverse outcome for patients receiving real-time ultrasound guidance were significantly lower (odds ratio = 0.08; 95% confidence interval, 0.009 to 0.811; P = 0.032) than for those receiving a standard technique while holding the propensity score quartile fixed. CONCLUSIONS The use of real-time ultrasound guidance during percutaneous dilatational tracheostomy was associated with a significant reduction in procedure-related complications.