Objective: To evaluate the diagnostic performance of the rapid shallow breathing index (RSBI) in predicting extubation failure among adult patients in the intensive care unit and to determine the appropriateness of the classical RSBI cut-off value. Methods: This was a prospective study conducted in the adult intensive care unit of the Botucatu School of Medicine Hospital das Clínicas. The RSBI was evaluated in 73 consecutive patients considered clinically ready for extubation. Results: The classical RSBI cut-off value (105 breaths/min/L) presented a sensitivity of 20% and a specificity of 95% (sum = 115%). Analysis of the receiver operator characteristic (ROC) curve revealed a better cut-off value (76.5 breaths/min/L), which presented a sensitivity of 66% and a specificity of 74% (sum = 140%). The area under the ROC curve for the RSBI was 0.78. Conclusions: The classical RSBI cut-off value proved inappropriate, predicting only 20% of the cases of extubation failure in our sample. The new cut-off value provided substantial improvement in sensitivity, with an acceptable loss of specificity. The area under the ROC curve indicated that the discriminative power of the RSBI is satisfactory, which justifies the validation of this index for use.

Keywords: Adult; Intensive care; ROC curve; Ventilator weaning; Diagnosis; Respiration, artificial.