• Sonja Suntrup-Krueger, Sarah Schmidt, Tobias Warnecke, Corinna Steidl, Paul Muhle, Jens B Schroeder, Bendix Labeit, Jens Minnerup, and Rainer Dziewas.
• From the Department of Neurology, University of Muenster, Germany (S.S.-K., S.S., T.W., P.M., J.B.S., B.L., J.M., R.D.).
• Stroke. 2019 Aug 1; 50 (8): 1981-1988.

AbstractBackground and Purpose- Predicting safe extubation represents a clinical challenge in acute stroke patients. Classical respiratory weaning criteria have not proven reliable. Concerning the paramount relevance of postextubation dysphagia in this population, criteria related to airway safety seem to perform better, but diagnostic standards are lacking. We compare clinical and instrumental swallowing examination tools to assess extubation readiness and propose a simple Determine Extubation Failure in Severe Stroke score for decision making. Methods- Data of 133 orally intubated acute stroke patients were prospectively collected in this observational study. Classical extubation criteria, a modified semiquantitative airway score, and an oral motor function score were assessed before extubation. A 3-ounce water swallow test and validated 6-point fiberoptic endoscopic dysphagia severity scoring were performed thereafter. Association of demographic and clinical parameters with extubation failure (EF) was investigated. Independent predictors of EF were translated into a point scoring system. Ideal cutoff values were determined by receiver operator characteristics analyses. Results- Patients with EF (24.1% after 24±43 hours) performed worse in all swallowing assessments (P<0.001). Fiberoptic endoscopic dysphagia severity scoring was the only independent predictor of EF (adjusted odds ratio, 4.2; P<0.007) with optimal cutoff ≥5 (sensitivity 84.6% and specificity 76.5%). Restricting regression analysis to parameters collected before extubation, a 4-item Determine Extubation Failure in Severe Stroke score (duration of ventilation, the examination of oral motor function, infratentorial lesion, and stroke severity) was derived. The score demonstrated excellent discrimination (area under the curve 0.89; 95% CI, 0.83-0.95) and calibration (Nagelkerkes R2=0.54) with an ideal cutoff ≥4 (sensitivity: 81.3% and specificity: 78.2%). Conclusions- Risk of EF is strongly correlated with postextubation dysphagia severity in stroke. Fiberoptic endoscopic examination of swallowing best predicts necessity of reintubation but requires a trial of extubation. The Determine Extubation Failure In Severe Stroke score is based on easy to collect clinical data and may guide extubation decision making in critically ill stroke patients.

16 keywords...

hide…