Below what FEV1 should arterial blood be routinely taken to detect chronic respiratory failure in COPD?

INTRODUCTION: To diagnose and assess chronic respiratory failure in stable chronic obstructive pulmonary disease (COPD) the measurement of arterial blood gases (ABG) is required. It has been suggested that ABG be determined for this purpose when FEV1 ranges between 50% and 30% predicted, but these thresholds are not evidence-based. OBJECTIVE: To identify the post-bronchodilator (BD) FEV1 and arterial oxygen saturation (SaO(2)) values that provide the best sensitivity, specificity, and likelihood ratio (LR) for the diagnosis of hypoxaemic and/or hypercapnic chronic respiratory failure in stable COPD. METHODS: A total of 150 patients were included (39 with PaO2 < 60 mmHg [8 kPa], 14 of them with a PaCO2 ≥ 50 mmHg [6.7 kPa]). The best post-BD FEV(1) and SaO(2) cut-off points to predict chronic respiratory failure were selected using the PC and the Receiver Operating Characteristics (ROC) curves. RESULTS: A post-BD FEV(1) equal to 36% and an SaO(2) of 90% were the best predictive values for hypoxaemic respiratory failure and a post-BD FEV(1) equal to 33% for the hypercapnic variant. An FEV(1) ≥ 45% ruled out hypoxaemic respiratory failure. CONCLUSION: A post-BD FEV(1) of 36% is the best cut-off point to adequately predict both hypoxaemic and hypercapnic respiratory failure in the patient with stable COPD. For its part, an SaO(2) of 90% is the best value for isolated hypoxaemic failure. These values could be considered for future clinical recommendations/guidelines for COPD.