 An RT-driven extubation pathway for safe extubation practices and preparedness for extubation failure in the ICU Houman Khosravani MD PhD FRCPC, John Fuller MD FRCPC Impetus “Patients failing extubation experience longer lengths of stay, experience higher intensive care unit mortality, and incur greater hospital costs” 1. Re-intubation costs an average of ~$34,000 2. Overall extubation failure is a more difficult entity to predict than difficult intubation. With an evidence-driven approach the incidence of extubation failure may be reducible. Hypothesis Implementation of an extubation pathway/checklist can flag patients with potential for extubation failure and reduce morbidity and mortality. Methods We conducted a literature review using PubMed (Medline) and Google Scholar (1995-2015) for manuscripts addressing factors that anticipate extubation failure in critically-ill patients. We did our best to review the literature in a way to separate factors (studies) that refer more to patient readiness or preparedness for extubation rather than the evidence for factors that actually anticipate failure of extubation. We focused on the latter in order to provide a synopsis of the available literature. Findings The need to re-intubate patients in the ICU within 24-72 Hrs of a planned extubation occurs anywhere from 2-25% of the time. It would be reasonable to expect that this number may be slightly higher in unplanned extubations. Several risk factors have been identified, as outlined in the studies below, however factors such as age (in adults > 70 years of age), longer-durations of ventilation (> 3-7 days), and use of continuous sedation are important factors. In addition, conditions such as fluid balance and presence of anemia at the time of extubation may also be important contributors 3. Post extubation stridor can occur in up to 37% of patients, and can result in increased cost, morbidity, and mortality. Patients with a leak of < 10% of the inhaled volume have a risk of laryngeal edema that is significant and may cause respiratory distress 4; this has a specificity of 96%. Patients at risk include: traumatic intubations, females, prolonged > 7 days, oversized ETT, failed cuff-leak test. We elected not to list gender as a risk factor for extubation failure given the paucity of data showing that it has a high positive predictive value. Meta-analysis of cuff-leak test (i.e. positive cuff-leak test = absence of leak) showed a pooled sensitivity of 63%, and specificity of 86%, LR 4.04, OR 10.37 for extubation failure. In other words, a positive cuff-leak study (absence of leak) should be considered a high-risk airway during the peri-extubation period 5. Overall, the literature appears to support the notion that qualitative cuff-leaks are just as good as quantitative measurements and that limiting the cuff-leak test to the quantitative approach likely lowers the sensitivity of this test. Overall, the absence of a cuff-leak is a more specific finding. Peak cough expiratory flow (PCEF) was shown in study of 130 patients, median duration of 8 days on IPPV to be the sole factor associated with extubation failure. The rate of failure in this study was 10%. Higher rates of re-intubation were observed for patients unable to cough. Inability to cough, PCEF < 35 L/min, had sensitivity of 79% and specificity of 71%. Risk of extubation failure was 24% for patients who did not cough or PCEF < 35 L/min 6. Similarly, absence of cough or leak, or both was more frequently observed with PES (post-extubation stridor). The study effectively showed that presence of a leak effectively ruled out PES 7. In a study where patients had passed a spontaneous breathing trial (SBT)*, extubation failure occurred in 13.4% of extubations. Associated variables were rapid-shallow breathing index (RSBI) (OR 1 per unit value of the RSBI), Positive fluid balance (OR 1.70), Pneumonia as reason for initial intubation (OR 1.77) 8. Decreased peak expiratory flow rates using a calibrated flow meter (< 60 L/min), increase in sputum (>2.5 ml/hour in 2-3 hours prior to extubation), had RR of 4.8 and 3.0 respectively (for extubation failure). In addition, inability to complete actions upon comment: open eyes, follow with eyes, grasp hand, stick out tongue had a RR of 4.3. Combining all three had a synergistic effect 9. Lastly, in a sub-population of patients, specifically those with COPD and perhaps those intubated for hypercapnic respiratory failure, there is evidence that non-invasive ventilation (NIV) can reduce the rate of re-intubations. These studies are nicely summarized by Epstein3, see table included below. Overall, two randomized controlled studies in patients with COPD show that NIV can facilitate weaning from mechanical ventilation. Moreover, several uncontrolled studies suggest NIV benefits > 2/3 of patients experiencing extubation failure.  Based on the above, we proposed the following attached pathway for consideration by the Quality Improvement committee. If approved, this pathway would be implemented in the unit as an RT-driven intervention to facilitate a reduction in re-intubation rates. At the very least, this pathway inserts an important cognitive pause prior to extubation in the ICU with evidence-based considerations that may predict extubation failure. Being better prepared for extubation failure may reduce morbidity and mortality for patients that experience sudden and clinically consequential hypoxia post-extubation. In addition, we suggest two important interventions: a) pre-oxygenation prior to extubation, and b) BiPAP for all patients admitted because of COPD, with consideration of BiPAP for those admitted for heart failure, and/or neuromuscular weakness and/or obesity hypo-ventilation syndrome. If you are from outside of Western University and are interested in participating in a research study or want additional information about the protocol, please contact us *. Questions, comments or want to implement this protocol at your local institution? - feel free to contact us. Please also checkout our Intubation checklist References
OR = odds ratio LR = likelihood ratio * - The definition of SBT can potentially be a matter of debate. Likely the best definition of SBT is breathing on a T-piece. However, minimal ventilator settings of pressure support of 5/5 with low FiO2 is currently used at our centre. Model ETT image from Wikipedia. |
