Research Article: Influencing factors of peripheral blood indicators on 30-day mortality risk in children with hemophagocytic lymphohistiocytosis based on weighted quantile sum regression: a single-center retrospective cohort study
Abstract:
To investigate the joint effect of peripheral blood parameters on the 30-day mortality risk in children with hemophagocytic lymphohistiocytosis (HLH) using weighted quantile sum (WQS) regression, and to elucidate the clinical weights of each feature.
A retrospective cohort design was adopted. A total of 133 pediatric HLH patients diagnosed at the Children's Medical Center of the Affiliated Hospital of Guangdong Medical University between January 1, 2015, and December 30, 2024, were included. The patients were divided into a non-survival group ( n =?29) and a survival group ( n =?104) based on their 30-day survival outcomes following diagnosis. Baseline peripheral blood parameters within 24?h prior to or on the day of diagnosis were collected. After preliminary screening via univariate logistic regression, a WQS regression model was utilized to evaluate the joint effect of peripheral blood parameters and extract empirical weights. In addition, restricted cubic spline (RCS) analysis was employed to explore the dose-response relationship between core parameters and mortality risk. The predictive efficacy and clinical net benefit of the core parameters and their combined model were evaluated using receiver operating characteristic (ROC) curves and decision curve analysis (DCA).
Univariate analysis indicated that red blood cell distribution width (RDW), mean corpuscular hemoglobin (MCH), hematocrit (HCT), and hemoglobin (Hb) were significantly associated with early mortality risk. The negative-direction WQS regression model revealed that the overall improvement of peripheral blood parameters had a significant joint protective effect against 30-day mortality risk ( ? =??0.936, P =?0.012), with MCH (weight 62.6%) and HCT (weight 29.8%) exhibiting the highest relative weights within the evaluated mixture. RCS analysis demonstrated a significant overall association between MCH, HCT, and mortality risk, approximating a linear relationship. The area under the curve (AUC) of the combined predictive model constructed by MCH and HCT improved to 0.728, indicating superior predictive efficacy over single indicators. Moreover, DCA confirmed that the combined predictive model provided a favorable clinical net benefit across a broad range of threshold probabilities compared to single parameters.
The analytical strategy incorporating the WQS regression model effectively mitigates collinearity interference among clinical variables. Abnormalities in peripheral blood parameters, primarily driven by HCT and MCH, serve as a exploratory composite indicator for predicting early fatal outcomes in pediatric HLH, offering a potential early warning signal for critical triage in the early stages of the disease.
Introduction:
Hemophagocytic lymphohistiocytosis (HLH) in children is an immune dysregulation syndrome triggered by predisposing factors such as genetic defects, infections, malignancies, or autoimmune diseases. Its core pathological mechanism lies in the functional defects of cytotoxic T lymphocytes and natural killer cells, leading to the overactivation of macrophages and the massive secretion of pro-inflammatory cytokines, which subsequently trigger a cytokine storm and multiple organ dysfunction syndrome ( 1 , 2 ). With the…
Read more