Assessing the accuracy of pulse oximetry to detect severe hypoxemia in chronic lung disease patients using long term oxygen therapy

<bold>Background:</bold> Long-term oxygen therapy (LTOT) is recommended to modify survival in patients with severe hypoxemia. However, the accuracy of pulse oximetry (SpO2) in detecting severe hypoxemia in this population remains uncertain. <bold>Objective:</bold> Evaluate SpO2 accuracy in identifying severe hypoxemia in chronic lung disease patients. <bold>Methods:</bold> Retrospective cross-sectional study from 2022-2023. Demographic data and diagnosis were collected from medical records. SpO2 values (measured using Dixtal DX 2022) and arterial blood gas were evaluated after 30 minutes breathing room air. Cutoff points to assess the accuracy for severe hypoxemia were considered as SaO2 ≤ 88% and PaO2 ≤ 55.0 mmHg. <bold>Results:</bold> 383 patients (mean age: 69.6± 12.5y, 65% females) were included. Main diagnosis included COPD (49.1%), circulatory pulmonary diseases (18.3%), post-infectious (16.2%), interstitial lung disease (15.4%), and cancer (9.4%). The settled threshold for SpO2 identified 11 false positives and 10 false negatives, with 91% sensitivity and 93% specificity. Analyzing COPD vs non-COPD groups, lower values of PaO2 (56.8 ±12.1 vs 61.0±14.1, p=0,002) and higher values of PaCO2 (43.3±7.9 vs 39.8±7.0, p=0,001) were observed in the group of COPD. In the COPD group, 7 patients were identified as false negative and 9 as false positive. <bold>Conclusion:</bold> 7,4% of the evaluated patients were misclassified by pulse oximetry with cutoff of SpO2≤88%. Precise diagnostics, especially in COPD, contribute to informed decisions, enhancing healthcare outcomes.