A new index, Respiratory Insufficiency index and Modified Early Warning Scores predict extubation failure

BACKGROUND: Extubation failure occurs in 5%–20% of patients and is associated with poor clinical outcomes. The primary aim of this project was to determine the predictive ability of the Respiratory Insufficiency (RI) index, Respiratory Oxygenation (ROX) index and Modified Early Warning Score (MEWS) in identifying extubation failure. METHODS: This was a secondary analysis of a prior cross-sectional retrospective study conducted from February 2018 through December 2018 among adult subjects who received mechanical ventilation for more than 24 h. Extubation failure was defined as the need for reintubation or rescue non-invasive ventilation (NIV) within 48 h after planned extubation. Univariate analysis and logistic regression were used to identify the predictors and final model was validated using 10-fold cross validation. Nomogram was constructed based on the final model. RESULTS: Of 216 enrolled subjects, 46 (21.3%) experienced extubation failure. The median RI index 1-h post extubation was 20 [interquartile range [IQR] 16.33–24.24] for success group and 27.02 [IQR 22.42–33.83] for the failure group (P<0.001). The median ROX index 1-h post extubation was 16.66 [IQR 12.57–19.84] for success group and 11.11 [IQR 8.09–14.67] for failure group (P<0.001). The median MEWS 1-h post extubation was 2 [IQR 1–3] for the success group and 4 [IQR 3–5] for the failure group (P<0.001). In multivariable analysis, age >60 years [OR 3.89 (95% CI 1.56–9.73); P=0.004], MEWS >4 [OR 4.01 (95% CI (1.59–10.14); P=0.003] and, RI index >20 [OR 4.50 (95% CI 1.43–14.21); P=0.010] were independently associated with extubation failure. CONCLUSION: In the present study, RI index and MEWS were independently associated with predicting extubation failure within 1 h of extubation. A prospective validation study is warranted to establish the role of these indices in predicting extubation outcome.