Prognostic factors and outcomes in pediatric posterior reversible encephalopathy syndrome: a 6-year single-center PICU study.

UNLABELLED: This study aims to evaluate the demographic, clinical, laboratory, and radiological characteristics of pediatric patients with posterior reversible encephalopathy syndrome (PRES) admitted to the pediatric intensive care unit (PICU) and to identify prognostic factors associated with mortality. This retrospective cohort study included 52 MRI-confirmed pediatric PRES cases managed in a tertiary-level PICU between January 2020 and January 2026. Demographic features, etiologies, clinical findings, severity scores (PRISM/PELOD), laboratory parameters, treatment modalities, MRI patterns, and outcomes were compared between survivors and non-survivors. Univariable logistic regression and receiver operating characteristic (ROC) curve analyses were performed to identify unadjusted clinical correlates of mortality; given the small number of deaths (n = 8), a multivariable model was not fitted, and Firth's penalized logistic regression was used as a sensitivity analysis for the main associations. The study adhered to STROBE guidelines. The mean age was 120.0 ± 50.8 months; 65.4% had hematologic-oncologic diagnoses. Overall mortality was 15.4% (8/52). Seizures occurred in 90.4% and hypertension in 88.5% of patients. MRI most frequently demonstrated occipital and parietal involvement (90.4% each). Non-survivors had significantly lower admission GCS scores (median 10.0 [IQR 8.0-12.5] vs 13.0 [12.0-15.0], p = 0.012), and longer PICU length of stay (median 14.5 days [IQR 9.2-26.5] vs 5.0 days [3.0-12.2], p = 0.016). GCS demonstrated an area under the curve (AUC) of 0.777 (95% CI 0.619-0.918) with optimal cutoff ≤ 10 (sensitivity 62%, specificity 84%). In univariate logistic regression with Firth penalized correction, vasopressor requirement was strongly associated with mortality (unadjusted penalized OR = 29.75, 95% CI 2.800-315.530, p = 0.005); the wide confidence interval reflects near-complete separation (7/8 non-survivors vs 6/44 survivors required vasopressors). The Glasgow Coma Scale (GCS) score at admission was associated with survival; each one-unit increase in the GCS score was associated with a 27% lower unadjusted odds of death (Firth-penalized estimate OR 0.73, 95% CI 0.550-0.940, p = 0.016).

CONCLUSION

In this large single-center pediatric series, lower GCS score, vasopressor requirement, and prolonged PICU stay were associated with mortality in univariate analysis. Vasopressor requirement may reflect the degree of hemodynamic compromise and the severity of illness rather than serve as an early clinical warning. These exploratory findings require validation in larger multicenter cohorts before clinical application.

WHAT IS KNOWN: • PRES typically presents with acute hypertension, seizures, headaches, visual disturbances, and altered mental status. Radiologically, brain imaging reveals cerebral edema, which is more pronounced in the parietal and occipital regions.

WHAT IS NEW: • In univariate logistic regression with Firth penalized correction, vasopressor requirement was associated with mortality (OR = 29.75, 95% CI 2.800-315.530, p = 0.005). • Non-survivors had lower admission GCS, and GCS ≤ 10 predicted death (AUC 0.777, 95% CI 0.619-0.918; 84% specificity).