L. Toelle1, M. Zaza1, S. Leonard1, E. A. Taub1, B. A. Cotton1 1McGovern Medical School at UTHealth,Acute Care Surgery,Houston, TEXAS, USA
Introduction: Early tracheostomy is associated with a reduction in ventilator-associated pneumonia, tracheal stenosis, time spent on ventilator and mortality. Despite these benefits, early tracheostomy is often not performed due to high ventilatory requirements. We hypothesized that patients who undergo tracheostomy under high ventilator settings would have similar complication rates compared to those performed under standard ventilator settings.
Methods: We performed retrospective review of all tracheostomies performed by Acute Care Surgery Faculty between 01/2015 and 12/2017. Patient demographics, ventilator settings, type (open or closed) and location of tracheostomy (operating room or ICU), and complications were recorded. Patients were divided into HIGH-SETTING and LOW-SETTING. HIGH-SETTING tracheostomies were defined as those on FiO2 >50%, PEEP >10, on PRVC mode, on APRV mode, or on nitric oxide. Complications were defined as: loss of airway, hypoxia requiring intervention, intra-operative ACLS, bleeding requiring return to OR, and death related to procedure. Hypoxia requiring intervention was defined as the need for bagging greater than one minute, any need for chest compressions, or need for emergent bronchoscopy post-procedure. Statistical analysis was performed using STATA 12.1. Continuous data are presented as medians (25th-75th percentile interquartile range, IQR) with comparisons performed using Wilcoxon ranksum. Categorical data are reported as proportions and tested for significance using Fisher’s exact test.
Results: 690 tracheostomies were performed during this time frame. 154 were HIGH-SETTING tracheostomies, while 536 were LOW-SETTNG. HIGH-SETTING patients were younger (median 53 vs. 57; p=0.028) and more likely to be male (76 vs. 66%; p=0.012). While there was no difference in intra-operative vasopressors drips (2.6 vs. 2.5%; p=0.992), HIGH-RISK tracheostomies were more likely to have a cervical spine fracture with spine immobilization (27 vs. 17%; p=0.008). HIGH-SETTING tracheostomies were more likely to be done through a percutaneous approach than LOW-SETTING tracheostomies (40 vs. 32%; p=0.082). However, there was no difference in whether the procedure was performed in the ICU setting (23 vs. 20%; 0.362). HIGH-SETTING patients were on higher PEEP, had higher FiO2, were more likely to be on advanced ventilator modes, and to have been on inhaled NO2. While there was a trend towards more ACLS interventions with HIGH-SETTING patients, there was no difference in complications overall or individually (TABLE).
Conclusion: Tracheostomies can be performed safely on high ventilator settings without increased complications. Careful pateint selection and hightened pre-procedure planning, however, is warranted.
