Deciphering the Breathless Future: A Novel Approach to Predicting Respiratory Failure in Children With Guillain-Barré...

The applicability and utility of clinical predictive models for respiratory failure in the pediatric Guillain-Barré syndrome (GBS) and Asian population are significantly constrained. Therefore, we aim to develop and validate a clinical prediction model to predict respiratory failure risk in pediatric GBS patients in China, alongside the economic immunological biomarkers in decision-making, and evaluate the Erasmus GBS Respiratory Insufficiency Score (EGRIS)-Kids score's efficacy.

The retrospective study originated from our pediatric GBS cohort at Children's Hospital of Chongqing Medical University during 2014-2022. We utilized logistic regression to identify predictors and construct a nomogram and web-based dynamic nomogram. The net reclassification index and integrated discriminant improvement index were used to compare models after incorporating new indices, with bootstrapping validation.

Our study included 175 children, among which 23 (13%) patients have developed respiratory insufficiency. Variables included in our score were: age (odds ratio [OR] 1.23), bulbar palsy (OR 23.17), bilateral hip flexion (OR 0.75), and platelet-to-lymphocyte ratio (PLR; OR 2.47). The area under the receiver operating characteristic curve of the nomogram was 0.899 (95% confidence interval 0.839-0.960). The calibration plots showed an adequate consistency between the reported and predicted occurrence. The EGRIS-Kids model yielded an area under the receiver operating characteristic curve of 0.849 (95% confidence interval 0.754-0.944) in our cohort and the incorporation of PLR exhibited an incremental value of 12.51% (P = 0.03).

We performed the first external validation of the EGRIS-Kids score in Chinese children with GBS. Furthermore, we developed a new predictive model incorporating the PLR, which shows promise and additional value but requires further external validation.