An S-like curve relationship between systemic inflammation response index (SIRI) and respiratory failure in GBS patients

Background Our study examined the association between the initial systemic inflammation response index (SIRI) and respiratory failure in patients with Guillain-Barré syndrome (GBS). Methods The weighted linear regression model, weighted chi-square test, logistic regression models, smooth curve fittings, and the two-piece linear regression model were utilized for data analysis. Results Among the 443 GBS patients, 75 (6.9%) had experienced respiratory failure. According to logistic regression models, there existed no consistent linear relationship between respiratory failure and SIRI in model 1 (OR = 1.2, p < 0.001), model 2 (OR = 1.2, p < 0.001), and model 3 (OR = 1.3, p = 0.017). However, smooth curve fittings found an S-like curve relationship between SIRI and respiratory failure. Furthermore, when SIRI was < 6.4, there existed a positive correlation between SIRI and respiratory failure in model 1 (OR = 1.5, 95% CI = (1.3, 1.8), p < 0.0001), higher correlation in model 2 (OR = 1.6, 95% CI = (1.3, 1.8), p < 0.0001), and highest correlation in model 3 (OR = 1.6, 95% CI = (1.3, 2.5), p < 0.0001). Conclusions SIRI can be used as a predictor of respiratory failure in GBS, and an S-like relationship exists between SIRI and respiratory failure at an infliction point of 6.4. When the SIRI was less than 6.4 and increased, SIRI was associated with a higher occurrence of respiratory failure. The risk of respiratory failure was no longer increased when the SIRI was over 6.4.