The Ability of Pulmonary Function Tests to Predict Oxygen Needs in Patients with Chronic Lung Diseases

Background: /Objectives: Chronic lung diseases (CLDs) constitute a major burden on healthcare systems, significantly impairing patients' quality of life. In advanced stages, long-term oxygen therapy (LTOT) is often indicated. While pulmonary function tests (PFTs) are routinely used in clinical evaluation, their role in predicting the need for LTOT remains inadequately defined. This study aimed to determine whether specific PFT parameters, particularly diffusing capacity for carbon monoxide (DLCO), can serve as reliable predictors of LTOT requirement in patients with CLD. Methods: A retrospective observational study was conducted, including 302 patients with CLD who underwent PFTs at Rambam Health Care Campus between December 2021 and December 2023. Patients were stratified into LTOT and non-LTOT groups. Demographic, clinical, and functional variables were analyzed using univariate and multivariate logistic regression models. Receiver operating characteristic (ROC) analysis was performed to assess the discriminative capacity of DLCO. Results: Of the 302 patients, 42 (13.9%) were prescribed LTOT. The LTOT group was older, predominantly male, and had a higher prevalence of COPD and interstitial lung disease. DLCO emerged as an independent and statistically significant predictor of LTOT (p = 0.029). A DLCO value below 39% predicted LTOT requirement with 90% probability, whereas a DLCO above 60% indicated a 90% likelihood of not requiring LTOT. The area under the ROC curve for DLCO was 0.855, indicating excellent discriminative performance. Conclusions: DLCO is a robust independent predictor for LTOT eligibility in patients with CLD. Integration of DLCO thresholds into routine assessment may enhance clinical decision-making. Validation through prospective multicenter studies is warranted.