A combination of the APACHE II score, neutrophil/lymphocyte ratio, and expired tidal volume could predict non-invasive ventilation failure in pneumonia-induced mild to moderate acute respiratory distress syndrome patients.

Background: Noninvasive ventilation (NIV) failure rate is relatively high in patients with acute respiratory distress syndrome (ARDS). Currently the data regarding prediction of NIV failure of pneumonia-induced mild to ARDS patients were scarce. Methods: A total of 364 patients (from January 2016 to December 2020) diagnosed with hypoxemic respiratory failure and managed with NIV were initially included and finally 131 pneumonia-induced mild to moderate ARDS patients were enrolled in this study. Electronic medical records were reviewed to determine whether NIV succeeded or failed for each patient. The relationship between the Acute Physiology And Chronic Health Evaluation II (APACHE II) score , neutrophil/lymphocyte ratio (NLR), expired tidal volume (Vte) and NIV failure were specifically analyzed. Multivariate logistic regression analyses were conducted to identify the independent factors of NIV failure. Receiver-operating characteristic curves were used to assess the efficacy of the variables in predicting NIV failure. Kaplan-Meier curves for 28-day survival were used to compare the mortality rates of different groups. Results: Of the patients, 64 (48.9%) experienced NIV failure, APACHE II score [odds ratio (OR) =0.77; P=0.002], NLR (OR =0.838; P=0.046), and Vte (OR =0.343; P=0.009) were independent factors for predicting NIV failure. A combined value comprising the APACHE II score, NLR, and Vte had better efficacy at predicting NIV failure [area under the curve (AUC) =0.9; 95% confidence interval (CI): 0.845-0.955] than the APACHE II score (AUC =0.818; 95% CI: 0.745-0.891), NLR (AUC =0.839; 95% CI: 0.765-0.913), or Vte (AUC =0.805; 95% CI: 0.729-0.881) alone. The cumulative survival probability within 28 days was lower in patients with a combined value >59.17 (P<0.001 by the log-rank test), an APACHE II score >16.5 (P<0.001 by the log-rank test), an NLR >7.22 (P<0.001 by the log-rank test), and a Vte >8.96 mL/kg (P<0.001 by log-rank test). Conclusions: A combined value comprising an APACHE II score >16.5, a NLR >7.22, and a Vte >8.96 mL/kg may be a useful surrogate for predicting NIV failure among pneumonia-induced ARDS patients, and patients with a combined value >59.17 should be cautiously monitored during NIV. A further study with a larger sample size is warranted.