To investigate the influencing factors of endotracheal intubation and mechanical ventilation (ETI-MV) in patients with acute respiratory distress syndrome (ARDS) caused by viral pneumonia, and to provide evidence for individualized use of ETI-MV. Patients with ARDS due to viral pneumonia admitted to the respiratory intensive care unit (RICU) of the First Affiliated Hospital of Xinjiang Medical University were retrospectively analyzed from November 2017 to March 2022. The gender, age, concomitant diseases, clinical symptoms and signs, complications, lab results, ARDS severity, infectious virus type, acute physiology and chronic health evaluation II (APACHE II), respiratory support methods and prognosis-related variables were collected. Univariate analysis was performed on each factor, and the variables with statistical significance in the univariate analysis were subjected multivariate logistic regression analysis. The receiver operating characteristic curve (ROC curve) was drawn to evaluate the predictive value of each index for the implementation of ETI-MV. A total of 117 patients were enrolled in the study, including 61 patients in the ETI-MV group, and 3 patients (4.9%), 39 patients (63.9%) and 19 patients (31.1%) with mild, moderate and severe ARDS, respectively. There were 56 patients in non-ETI-MV group, and the mild, moderate and severe ARDS cases were 16 cases (28.6%), 38 cases (67.8%) and 2 cases (3.6%), respectively. There was significant difference between the two groups (P < 0.05). Univariate analysis showed that during 24 hours admitted to RICU, the levels of interleukin-6 [IL-6 (ng/L): 104.0±90.0 vs. 62.4±76.0], oxygenation index [PaO2/FiO2 (mmHg, 1 mmHg ≈ 0.133 kPa): 123.9±30.9 vs. 173.6±28.5], the proportion of cases with pulmonary infiltrating opacity distribution range ≥ 3/4 lung fields [85.3% (52/61) vs. 21.5% (12/56)], APACHE II score ≥ 16.5 [67.2% (41/61) vs. 42.9% (24/56)], the rate of nosocomial invasive aspergillus infection [14.8% (9/61) vs. 3.6% (2/56)], the percentage of nosocomial bacterial infection [16.4% (10/61) vs. 3.6% (2/56)], and the lowest CD4+ T lymphocyte count in the course of the disease [cells/mm3: 192.2±35.8 vs. 215.0±58.3] had significant differences between ETI-MV and non-ETI-MV group (all P < 0.05). Multivariate Logistic regression analysis showed that during 24 hours admitted to RICU the distribution range of pulmonary infiltrating opacity ≥ 3/4 the lung fields [odds ratio (OR) = 12.527, 95% confidence interval (95%CI) = 3.279-47.859, P < 0.001], APACHE II score ≥ 16.5 (OR = 30.604, 95%CI = 4.318-216.932, P = 0.001), PaO2/FiO2 (OR = 0.948, 95%CI = 0.925-0.972, P < 0.001), CD4+ T lymphocytes cell count (OR = 0.975, 95%CI = 0.955-0.995, P = 0.015), and nosocomial bacterial infection (OR = 38.338, 95%CI = 1.638-897.158, P = 0.023) were independent risk factors for ETI-MV. The area under the ROC curve (AUC) of ROC showed that PaO2/FiO2 had the greatest predictive value for ETI-MV, with AUC of 0.903, sensitivity of 91.1% and specificity of 95.1% in case of cutoff value of 151 mmHg. The AUC of pulmonary infiltrating opacity distribution range was 0.809, the sensitivity of 85.2%, specificity of 78.6% when the cutoff value was ≥ 3/4 lung field. APACHE II scores had the lowest predictive value for selecting ETI-MV, with AUC of 0.704, sensitivity of 83.6% and specificity of 57.1% under the cutoff value was 16.5. For patients with ARDS caused by viral pneumonia, PaO2/FiO2 is still the classic reference for selecting ETI-MV, however, the distribution range of pulmonary infiltrating opacity and the systemic severity of the disease during 24 hours admitted to the RICU may provide supplemental helpful information to determine whether the patients choose ETI-MV, especially for moderate ARDS.
Read full abstract