# The clinical application of subcutaneous thoracic ratio and capillary leakage index on the occurrence of capillary leak syndrome in neonates with sepsis

## TL;DR

This study shows that two clinical measures, S/T ratio and CLI, can help predict capillary leak syndrome in sepsis-affected newborns, but they don't predict survival outcomes.

## Contribution

The study identifies S/T ratio and CLI as independent predictors of capillary leak syndrome in neonatal sepsis.

## Key findings

- S/T and CLI are independent risk factors for capillary leak syndrome in neonatal sepsis.
- Combining S/T and CLI improves prediction accuracy for capillary leak syndrome.
- Neither S/T nor CLI are associated with survival outcomes in neonatal sepsis.

## Abstract

To evaluate the predictive and prognostic value of the subcutaneous-thoracic ratio (S/T) and capillary leakage index (CLI) for capillary leak syndrome (CLS) in neonatal sepsis.

A cohort of 196 neonates with sepsis, admitted to a tertiary children's hospital in southwestern China between January 2019 and March 2021, was included in the study. The neonates were divided into two groups: the CLS group (n = 55) and the non-CLS group (n = 55). Multivariate logistic regression and receiver operating characteristic (ROC) curve analysis were performed to identify key predictors of CLS.

Both S/T and CLI were found to be independent risk factors for CLS in neonatal sepsis (P < 0.05). The median S/T values for the CLS group and non-CLS group were 9.0% and 7.1%, respectively, while the median CLI values were 8.5 and 3.2. The optimal thresholds for predicting CLS were identified as 8.1% for S/T (sensitivity: 67.3%, specificity: 70.9%) and 3.3 for CLI (sensitivity: 78.2%, specificity: 56.4%). Notably, the combination of S/T and CLI yielded improved predictive performance, with a sensitivity of 81.8% and specificity of 60.0%. However, neither S/T nor CLI were significantly associated with prognosis, as no difference was observed between survivors and non-survivors (P > 0.05).

The combined application of S/T and CLI provides an effective tool for predicting the occurrence of CLS in neonatal sepsis. However, these indicators do not demonstrate prognostic value for survival outcomes.