New software for assessment and interpretation of pulmonary function tests in adults: a comparative validation study

Abstract

Despite standardized international guidelines, interpretation of pulmonary function tests (PFTs) remains subject to substantial inter-observer variability. Automated guideline-based decision support systems may improve the standardization and reproducibility of PFT reporting. This agreement study evaluated the concordance between the interpretation generated by a new automated interpretation software (DrPFT © : Doctor Pulmonary Function Test) and an expert-based American Thoracic Society/European Respiratory Society (ATS/ERS) classification of PFTs. This non-interventional and agreement study included a cohort of 1,819 adult Tunisian participants who underwent complete PFT in outpatient and hospital settings in 2025 by a panel of five experienced pulmonologists ( n = 3) and physiologists ( n  = 2) using the standardized ATS/ERS interpretative criteria. A consensus classification was established for ventilatory patterns (normal, obstructive, restrictive, or mixed). The same datasets were subsequently analyzed by the DrPFT © , which implements ATS/ERS interpretative algorithms in an automated manner. Diagnostic concordance between DrPFT © and expert interpretation was assessed using sensitivity, specificity, positive, and negative predictive values, with ATS/ERS-based expert consensus serving as the reference standard. Based on the ATS/ERS criteria, 26.0% of participants were classified as normal, 25.8% restrictive, 26.4% obstructive, 9.4% mixed, and 12.4% nonspecific ventilatory pattern. When compared with the ATS/ERS reference, pulmonologists/physiologists demonstrated high sensitivity across all classical patterns (90.7%–94.2%) and high specificity, particularly for restrictive (99.3%) and mixed (98.0%) patterns. DrPFT © showed similarly robust performance when compared with the ATS/ERS reference, with sensitivities ranging from 88.6% to 93.6% and specificities ≥84.5% for all ventilatory patterns. Direct comparison revealed excellent agreement between DrPFT © and expert interpretation, with sensitivity ≥96.3% and specificity ≥98.4% across all categories. Sensitivity was comparable between DrPFT © and expert interpretation for restrictive, obstructive, and mixed patterns (all p > 0.050), whereas a small but statistically significant difference was observed for the classification of normal pulmonary function pattern participants, which was higher via expert interpretation than via DrPFT © (90.7% vs. 88.6%, respectively; p  = 0.002). DrPFT © provides automated ATS/ERS guideline-based interpretation of ventilatory patterns using spirometry and lung volume measurements, with a level of concordance comparable to experts. Although the software also incorporates interpretation of carbon monoxide diffusing capacity (DLCO), this component was not formally validated in the present study because of insufficient data. By standardizing the application of ATS/ERS interpretative criteria, this decision support system may improve the consistency and efficiency of routine PFT reporting, particularly in settings with limited access to specialized expertise. Pan African Clinical Trials Registry (PACTR) Identifier: PACTR202601512810753.