USING THE RATIO OF OXYGEN SATURATION (ROX) INDEX TO PREDICT TREATMENT OUTCOME FOR HIGH-FLOW NASAL CANNULA AND/OR NONINVASIVE VENTILATION IN PATIENTS WITH COPD EXACERBATIONS

SESSION TITLE: Late Breaking Posters in Critical CareSESSION TYPE: Original Investigation PostersPRESENTED ON: 10/18/2022 01:30 pm - 02:30 pmPURPOSE: The ratio of oxygen saturation—or ROX—index (SpO2/FiO2/RR) has previously been shown to predict outcome of high-flow nasal cannula (HFNC) support in all-comers with acute hypoxemic respiratory failure, most of which had pneumonia. We aimed to assess the performance of the ROX index on a unique pathophysiology leading to respiratory failure: COPD. In patients admitted primarily for acute exacerbation of COPD, we assessed the performance of the validated ROX index to predict treatment outcome (liberation from HFNC or non-invasive ventilation (NIV) treatment, intubation with mechanical ventilation, or death) for patients requiring HFNC or NIV.METHODS: This study is a retrospective analysis of a cohort of 291 hospitalized patients between 2012-2018 with externally validated COPD with acute exacerbation treated with HFNC and/or NIV (BiPAP or CPAP). ROX index scores were collected at treatment initiation and at predetermined time intervals, thereafter, for the duration of HFNC and/or NIV treatment, or until the patient was intubated or died. Based on prior studies (Roca, 2016), a ROX index score of greater than or equal to 4.88 was applied to the cohort to determine if the score would predict a patient's successful treatment with HFNC or NIV. Accuracy of the ROX index was determined by calculating the area under the receiver operator curve (AUROC).RESULTS: A total of 47 patients (16%) required invasive mechanical ventilation or died while on treatment. Most cases of treatment failure occurred in the first 6 hours of HFNC or NIV initiation (39/47, 83%). The ROX index at initiation of HFNC or NIV (time 0), 1 hour, and at 6 hours from initiation demonstrated the best prediction accuracy for success (AUROC 0.753 [CI 0.69-0.82], 0.752 [CI 0.69-0.82], and 0.751 CI [0.61-0.89], respectively). The optimal cutoff value for sensitivity and specificity was a ROX index score >6.88 (Sn 64%, Sp 57%). When using the previously validated Roca threshold of greater than or equal to 4.88, the sensitivity was increased 84% with a decrease in specificity to 6%.CONCLUSIONS: The ROX index applied to patients with acute exacerbation of COPD and use of HFNC and/or NIV performed similarly when compared to patients with hypoxemic respiratory failure and pneumonia for determining patients who are at low risk of treatment failure. External validation in larger cohorts is required.CLINICAL IMPLICATIONS: The ROX index has shown value in predicting outcome in patients with hypoxemic respiratory failure and pneumonia started on HFNC, but the performance of this index in patients with COPD and HFNC and/or NIV use was not known. Our single-center study suggests similar performance of ROX in COPD for prediction of low risk of HFNC and/or NIV treatment failure, but formal derivation/validation may be required to identify the optimal performance.DISCLOSURES:No relevant relationships by Jason AdamsNo relevant relationships by Timothy AlbertsonNo relevant relationships by Hugh BlackNo relevant relationships by Irene Cortes-PuchNo relevant relationships by Sarina FazioNo relevant relationships by Richart HarperNo relevant relationships by Brooks Kuhn, value=HonorariaRemoved 06/21/2022 by Brooks KuhnNo relevant relationships by Brooks Kuhn, value=Consulting feeRemoved 06/21/2022 by Brooks KuhnNo relevant relationships by Anna LiuNo relevant relationships by Brett SchaefferNo relevant relationships by Jacqueline Stocking SESSION TITLE: Late Breaking Posters in Critical Care SESSION TYPE: Original Investigation Posters PRESENTED ON: 10/18/2022 01:30 pm - 02:30 pm PURPOSE: The ratio of oxygen saturation—or ROX—index (SpO2/FiO2/RR) has previously been shown to predict outcome of high-flow nasal cannula (HFNC) support in all-comers with acute hypoxemic respiratory failure, most of which had pneumonia. We aimed to assess the performance of the ROX index on a unique pathophysiology leading to respiratory failure: COPD. In patients admitted primarily for acute exacerbation of COPD, we assessed the performance of the validated ROX index to predict treatment outcome (liberation from HFNC or non-invasive ventilation (NIV) treatment, intubation with mechanical ventilation, or death) for patients requiring HFNC or NIV. METHODS: This study is a retrospective analysis of a cohort of 291 hospitalized patients between 2012-2018 with externally validated COPD with acute exacerbation treated with HFNC and/or NIV (BiPAP or CPAP). ROX index scores were collected at treatment initiation and at predetermined time intervals, thereafter, for the duration of HFNC and/or NIV treatment, or until the patient was intubated or died. Based on prior studies (Roca, 2016), a ROX index score of greater than or equal to 4.88 was applied to the cohort to determine if the score would predict a patient's successful treatment with HFNC or NIV. Accuracy of the ROX index was determined by calculating the area under the receiver operator curve (AUROC). RESULTS: A total of 47 patients (16%) required invasive mechanical ventilation or died while on treatment. Most cases of treatment failure occurred in the first 6 hours of HFNC or NIV initiation (39/47, 83%). The ROX index at initiation of HFNC or NIV (time 0), 1 hour, and at 6 hours from initiation demonstrated the best prediction accuracy for success (AUROC 0.753 [CI 0.69-0.82], 0.752 [CI 0.69-0.82], and 0.751 CI [0.61-0.89], respectively). The optimal cutoff value for sensitivity and specificity was a ROX index score >6.88 (Sn 64%, Sp 57%). When using the previously validated Roca threshold of greater than or equal to 4.88, the sensitivity was increased 84% with a decrease in specificity to 6%. CONCLUSIONS: The ROX index applied to patients with acute exacerbation of COPD and use of HFNC and/or NIV performed similarly when compared to patients with hypoxemic respiratory failure and pneumonia for determining patients who are at low risk of treatment failure. External validation in larger cohorts is required. CLINICAL IMPLICATIONS: The ROX index has shown value in predicting outcome in patients with hypoxemic respiratory failure and pneumonia started on HFNC, but the performance of this index in patients with COPD and HFNC and/or NIV use was not known. Our single-center study suggests similar performance of ROX in COPD for prediction of low risk of HFNC and/or NIV treatment failure, but formal derivation/validation may be required to identify the optimal performance. DISCLOSURES: No relevant relationships by Jason Adams No relevant relationships by Timothy Albertson No relevant relationships by Hugh Black No relevant relationships by Irene Cortes-Puch No relevant relationships by Sarina Fazio No relevant relationships by Richart Harper No relevant relationships by Brooks Kuhn, value=Honoraria Removed 06/21/2022 by Brooks Kuhn No relevant relationships by Brooks Kuhn, value=Consulting fee Removed 06/21/2022 by Brooks Kuhn No relevant relationships by Anna Liu No relevant relationships by Brett Schaeffer No relevant relationships by Jacqueline Stocking