Eosinophil-To-Lymphocyte Ratio and Neutrophil-To-Lymphocyte Ratio in Erythroderma: An Exploratory Subtype-Stratified...

Erythroderma is most often secondary to psoriasis, atopic dermatitis (AD), or drug eruption. These three subtypes have different immune mechanisms, but most biomarker studies treat erythroderma as one condition. We examined whether routine complete blood count (CBC) indices behave differently across the three subtypes. This was a single-center retrospective study of inpatients admitted between April 2015 and March 2026. We compared eosinophil-to-lymphocyte ratio (ELR), absolute eosinophil count (AEC), and neutrophil-to-lymphocyte ratio (NLR) between erythroderma cases and non-erythrodermic controls within each subtype, with Benjamini-Hochberg correction. We assessed ROC discrimination using a 5 000-iteration bootstrap and Harrell's optimism correction. Of 375 inpatients (18, 19, and 19 erythroderma cases in psoriasis-type, AD-type, and drug eruption-type), ELR was higher in cases than controls in all three subtypes (all q ≤ 0.007), with corrected AUCs of 0.84, 0.69, and 0.80. NLR was higher in cases for psoriasis-type (AUC 0.71) and AD-type (AUC 0.75) but not for drug eruption-type (AUC 0.56). In AD-type, a combined model with ELR and NLR gave a corrected AUC of 0.76; no gain was seen in the other two subtypes. In this small exploratory cohort, ELR appears higher in erythroderma across the three subtypes, while NLR carries information only in psoriasis-type and AD-type. The pattern is consistent with eosinophil-compartment activity being a shared feature and neutrophil-compartment activity being subtype-restricted. The findings are hypothesis-generating and need confirmation in prospective multicenter cohorts that can also control for pre-admission medication.