A prospective observational investigation examining critically ill patients has demonstrated that the Radiographic Assessment of Lung Edema score reliably quantifies the presence and anatomical extent of diffuse airspace disease on post-intubation portable chest radiographs, yet cannot discriminate between underlying clinical syndromes that manifest with similar radiographic patterns. The study evaluated 4,259 frontal portable chest radiographs obtained from 814 adult patients presenting with acute respiratory failure who were enrolled in the Acute Lung Injury Registry and Biospecimen Repository at UPMC Hospitals between 2011 and 2024. Clinicians routinely obtain portable chest radiographs across emergency departments, hospital wards, and intensive care units to verify device placement and qualitatively evaluate parenchymal abnormalities.
To evaluate the standardized measurement tool across diverse clinical conditions, researchers examined index radiographs taken on the calendar day immediately following endotracheal intubation. The cohort comprised 814 patients with a median age of 59 years, an interquartile range of 47 to 68 years, 57 percent male representation, and a median body mass index of 30 with an interquartile range of 26 to 36. An expert consensus committee consisting of attending pulmonary and critical care physicians adjudicated participants into seven mutually exclusive clinical subtypes: acute respiratory distress syndrome per Berlin criteria, at-risk for acute respiratory distress syndrome, congestive heart failure, airway controls intubated for airway protection, acute-on-chronic hypercapnic respiratory failure, acute exacerbation of interstitial lung disease, and an other or multifactorial presentation category. Adjudication relied on clinical, physiological, laboratory, and qualitative radiographic review without knowledge of quantitative RALE measurements.
The multi-rater scoring framework employed a validated protocol where fifteen trained clinicians, blinded to clinical data, adjudicated diagnostic subtypes, and patient outcomes, evaluated all frontal chest radiographs. The Radiographic Assessment of Lung Edema methodology divides each radiograph into four anatomical quadrants, scoring each quadrant for the extent of opacification from 0 to 4 and the density of opacification from 0 to 3. Quadrant scores represent the product of extent and density, which are then summed across all four quadrants to generate a cumulative score ranging from 0 to 48. Independent clinician ratings demonstrated excellent inter-rater reliability, evidenced by an overall intraclass correlation coefficient of 0.92. Consensus measurements for primary analyses were calculated as the mean score among the trained evaluators, establishing an objective baseline against which expert-adjudicated clinical reference categories were compared.
Clinical subtypes were grouped into three distinct categories reflecting expected parenchymal involvement to test the primary research questions. Airway controls and patients with acute-on-chronic hypercapnic respiratory failure constituted the airspace disease absent group, totaling 114 patients whose lungs were definitionally expected to exhibit minimal or no parenchymal opacities. The airspace disease limited group comprised 279 patients categorized as at-risk for acute respiratory distress syndrome who possessed risk factors but lacked diffuse bilateral opacities. The airspace disease diffuse group included 421 patients diagnosed with acute respiratory distress syndrome, congestive heart failure, acute exacerbations of interstitial lung disease, or multifactorial failure, all defined by bilateral extensive opacification. Across these categories, the worst ratio of arterial oxygen partial pressure to fractional inspired oxygen in the initial 24 hours declined progressively from 202 mmHg in absent disease, to 170 mmHg in limited disease, and 127 mmHg in diffuse disease, while 60-day mortality increased from 21 percent to 39 percent.
Quantitative scoring distributions tracked closely with the established severity gradient across the adjudicated categories. Median RALE scores on index radiographs differed significantly across the cohorts, registering lowest among airway controls at 10.0 and acute-on-chronic hypercapnic failure at 14.7. Intermediate values appeared in patients at-risk for acute respiratory distress syndrome with a median of 16.5, whereas higher burdens were observed in multifactorial cases at 21.8, congestive heart failure at 23.3, acute respiratory distress syndrome at 26.7, and acute exacerbation of interstitial lung disease at 32.0. Grouped by cumulative burden, median scores increased stepwise from 10.6 in absent disease, to 16.5 in limited disease, and 26.0 in diffuse disease. When evaluated as a binary outcome, patients with any airspace disease presented a median score of 22.4 compared to 10.6 in patients without parenchymal disease.
Receiver operating characteristic curve analyses revealed that the index RALE score achieved an area under the curve of 0.814, with a 95 percent confidence interval of 0.772 to 0.855, for discriminating between the presence and absence of airspace disease across the 814 patients. A cohort-derived rule-out threshold of RALE greater than or equal to 7 yielded a sensitivity of 0.96 and a specificity of 0.29. A balanced threshold of RALE greater than or equal to 17, selected by maximizing Youden’s index, produced a sensitivity of 0.72 and a specificity of 0.82 with a positive likelihood ratio of 3.89. Internal validation using tenfold stratified cross-validation confirmed robust calibration. Among the 700 patients exhibiting airspace disease, the score discriminated diffuse bilateral involvement from limited disease with an area under the curve of 0.793, where an operational rule-out threshold of 12 achieved 0.95 sensitivity and a balanced threshold of 20 provided 0.78 sensitivity and 0.66 specificity.
Analysis of acute respiratory distress syndrome risk factors further demonstrated how underlying insults shape radiographic density measurements across patients excluding congestive heart failure. Scores differed significantly across categories, with higher measurements observed in patients possessing direct pulmonary insults such as pneumonia, aspiration, or lung contusion, as well as those with mixed pulmonary and extra-pulmonary insults, compared to patients with indirect extra-pulmonary insults like non-pulmonary sepsis, pancreatitis, or severe trauma. Scores in both direct and mixed risk groups were significantly higher than in individuals without identified risk factors, while the direct and mixed cohorts did not differ from one another. A pre-specified sensitivity analysis that excluded 60 diagnostically ambiguous multifactorial patients maintained an area under the curve of 0.809 for distinguishing diffuse from limited disease, verifying that discriminative findings were not an artifact of unclassifiable clinical presentations.
Despite strong stratification of disease extent, RALE score distributions overlapped broadly among acute respiratory distress syndrome, cardiogenic pulmonary edema, and acute exacerbations of interstitial lung disease. The authors observed that this distribution overlap is not an indicator of diagnostic misclassification, but directly reflects the nature of RALE as a metric of physical radiographic density rather than an etiologic classifier.
Technical image acquisition characteristics exerted a substantial influence on scoring accuracy and discriminative power. Radiographic penetration, categorized by the deepest visible spinal landmark, altered score magnitude markedly, showing progressive reductions in measured scores from the thoracic inlet down to the carina and heart before plateauing across diaphragmatic levels. In a multivariable linear mixed-effects model accounting for repeated images within subjects across 4,258 radiographs, suboptimal penetration was independently associated with an increase of 3.29 points in RALE score. In contrast, overall poor image quality and artifact burden showed no independent association with score inflation. Suboptimal penetration degraded the area under the curve for detecting airspace disease from 0.83 under adequate penetration down to 0.59, demonstrating that underpenetrated images create artifactual opacification that impairs clinical discrimination.
Several critical limitations define the interpretation and transportability of the reported findings. The study was conducted at a single academic medical center utilizing a registry intentionally enriched for acute lung injury, yielding an overall airspace disease prevalence of 86 percent that exceeds typical emergency department or general intensive care populations. Because predictive values depend on disease prevalence, the negative predictive value of 0.52 observed for the rule-out threshold of 7 cannot be directly extrapolated to lower-prevalence settings without formal external validation. Furthermore, the clinical reference standards were established by consensus review incorporating qualitative radiographic assessment, meaning RALE was evaluated against expected patterns rather than an independent continuous anatomical gold standard such as quantitative computed tomography. Finally, image quality attributes were scored subjectively by an expert reviewer, and three scoring clinicians participated in adjudication committees, highlighting the need for automated quality standardization and prospective validation across broader healthcare environments.
Subject of Research: Evaluation of the Radiographic Assessment of Lung Edema score across acute respiratory failure etiologies and image quality conditions
Article Title: Quantifying diffuse airspace disease on portable chest radiographs in acute respiratory failure using the RALE score
Article References: Gipson, V., Qurashi, H., Jain, P., Singh, N., Inman, K., Rizzo, A. N., Sinopoli, J., Mirza, T., Ali, R., Naghshtabrizi, N., Bobrysheva, H., Venkatesan, N., Golubykh, K., Prendergast, N., Hendricks, A., Oliveria, J.-P., Galanter, J., Cluceru, J., Negahdar, M., … Kitsios, G. D. (2026). Quantifying diffuse airspace disease on portable chest radiographs in acute respiratory failure using the RALE score. Intensive Care Medicine Experimental, 14(1), Article 102. https://doi.org/10.1186/s40635-026-00953-x
Image Credits: AI Generated
DOI: 10.1186/s40635-026-00953-x
Keywords: acute respiratory failure, RALE score, chest radiographs, acute respiratory distress syndrome, pulmonary edema, airspace disease, intensive care unit, mechanical ventilation, diagnostic imaging, radiographic scoring, critical care medicine, radiographic penetration, receiver operating characteristic, image quality, interstitial lung disease, cardiogenic edema
News Source: Barbara Leach. (November 19, 2026). RALE Score Quantifies Lung Disease Extent but Cannot Distinguish Causes. Scienmag.



