Objective  To establish and validate a recruitment-to-inflation (R/I) ratio-oriented nomogram prediction model for mortality risk and assess its prognostic value in patients with moderate to severe acute respiratory distress syndrome (ARDS). Methods  A prospective observational study was conducted. Patients diagnosed with moderate to severe ARDS receiving mechanical ventilation in the ICU of the First Affiliated Hospital of Jinzhou Medical University between January 2024 and July 2025 were enrolled. They were divided into a training set and a validation set in a 7:3 ratio. Baseline data and respiratory mechanics parameters, including R/I ratio, best positive end-expiratory pressure (Best PEEP), stress index (SI), driving pressure (DP) , tidal volume (VT), respiratory rate (RR), mechanical power (MP), and oxygenation index (P/F ratio) measured 4 hours after mechanical ventilation. Based on 28-day clinical outcomes, the patients were categorized into a survival group and a non-survival group. Least absolute shrinkage and selection operator (Lasso) regression and multivariate logistic regression were used to identify risk factors for 28-day prognosis in ARDS patients, and a nomogram prediction model was constructed. The predictive value of the model for 28-day mortality was evaluated in the validation set using receiver operating characteristic (ROC) curves and calibration curves (discrimination and accuracy) for internal validation. Decision curve analysis (DCA) was applied to assess the clinical utility of the nomogram. Results  A total of 207 patients with moderate to severe ARDS were included, with 145 in the training set and 62 in the validation set. In the training and validation sets, 44 (30.34%) and 21 (32.26%) patients died, respectively. Compared with the survival group, the non-survival group had higher Best PEEP, DP, RR, and MP, and lower R/I ratio, VT, and P/F ratio (all P<0.05). Lasso regression identified 7 potential predictors for 28-day prognosis from 8 clinical variables in the training set: R/I ratio, Best PEEP, DP, VT, RR, MP, and P/F ratio. Logistic regression results showed that R/I ratio (OR=0.014, 95%CI 0.000-0.924, P=0.014), Best PEEP (OR=1.171, 95%CI 1.001-1.370, P=0.048), DP (OR=1.199, 95%CI 1.009-1.426, P=0.039), VT (OR=0.991, 95%CI 0.982-1.000, P=0.046), RR (OR=1.461, 95%CI 1.222-1.745, P=0.009), MP (OR=1.150, 95%CI 1.005-1.315, P=0.042), and P/F ratio (OR=0.968, 95%CI 0.945-0.992, P= 0.012) were all independent risk factors for 28-day prognosis in the ARDS patients. A nomogram model was constructed using these variables. In both the training and validation sets, ROC curves showed that the area under the curve for predicting 28-day mortality was 0.884 (95%CI 0.828-0.939) and 0.814 (95%CI 0.691-0.937), respectively. Calibration curves demonstrated good consistency between predicted and actual mortality, indicating satisfactory accuracy. DCA results indicated that the nomogram prediction model provided greater net benefit than the "treat all" or "treat none" strategies. Conclusion  The R/I ratio-oriented nomogram prediction model for mortality risk in patients with moderate to severe ARDS demonstrates good prognostic value.

Citation： LI Tian, DI Xingwei, LIU Jingyu, LI Xiaodong. Recruitment-inflation ratio-based analysis of prognostic risk factors and nomogram prediction model for acute respiratory distress syndrome. Chinese Journal of Respiratory and Critical Care Medicine, 2026, 25(7): 486-493. doi: 10.7507/1671-6205.202509019 Copy

Copyright ? the editorial department of Chinese Journal of Respiratory and Critical Care Medicine of West China Medical Publisher. All rights reserved