The SETscore to Predict Tracheostomy Need in Cerebrovascular Neurocritical Care Patients.

BACKGROUND AND PURPOSE: Patients with severe stroke who require mechanical ventilation and neurointensive care unit (NICU) management often require a tracheostomy (TT). The optimal time point for TT remains unclear and a controversy in everyday NICU life. Here, we prospectively evaluated a score for prediction of TT need in NICU patients with cerebrovascular disease.

METHODS: Seventy-five consecutively ventilated stroke patients were prospectively included in the study and assessed by the stroke-related early tracheostomy score (SETscore) within the first 24 h of admission. Endpoints were TT need, NICU-length of stay (NICU-LOS), and ventilation time (VT). We examined the correlation of these variables with the SETscore using regression analysis and determined a cut-off by receiver operating characteristic (ROC) analysis.

RESULTS: Twenty-six patients had to be tracheostomized. The mean VT was 8.7 ±8 days and the mean NICU-LOS was 11.6 ± 8 days. The SETscore predicted NICU-LOS with a positive predictive value of 0.748 (p < 0.001) and VT with a positive predictive value of 0.799 (p < 0.001). The ROC analysis demonstrated a SETscore value of 8 to be the optimal cut-off to predict prolonged NICU-LOS, VT, and TT need with a sensitivity of 64 % and a specificity of 86 %.

CONCLUSIONS: Based on this monocentric study, the SETscore seems to be a valid tool to indicate prolonged NICU-LOS and VT, as well as TT need in cerebrovascular NICU patients. Confirmation of these results in larger cohorts with various settings may help to develop the SETscore as a decisive tool on primary TT early in time to avoid extubation failure.