Evidence Hub

Lactate dehydrogenase-to-albumin ratio predicts 30-day and 90-day mortality in glucocorticoid-treated ICU patients with pneumonia: a secondary analysis of a multicenter cohort

Dandan Han, Shengfang Sun, Ai-xin Liu, Cui-Ping Hao, Jian-Ping Sun

Frontiers in Medicine, 2026

499 participants · 1 citation

Abstract

Background Glucocorticoid therapy in pneumonia patients confounds conventional inflammatory biomarkers, hindering accurate risk stratification. The lactate dehydrogenase-to-albumin ratio (LAR), a proposed steroid-resilient index reflecting cellular injury and nutritional reserve, may offer prognostic utility in this population. Methods This secondary analysis of a multicenter cohort included 499 ICU patients with pneumonia receiving glucocorticoids. The optimal LAR cutoff for 30-day mortality was determined via the Youden index. Associations with 30- and 90-day mortality were assessed using multivariable Cox regression and propensity score matching (PSM), with doubly robust estimation serving as the primary analysis in the matched cohort. Incremental predictive value was evaluated using the net reclassification improvement (NRI) and integrated discrimination improvement (IDI). Results The optimal LAR cutoff was 13.39. High LAR (≥13.39) was associated with significantly higher 30-day (45.8% vs. 12.1%) and 90-day (51.4% vs. 14.3%) mortality (both p < 0.001). After full multivariable adjustment, high LAR remained an independent predictor for 30-day (HR 2.07, 95% CI 1.34–3.20) and 90-day (HR 1.93, 95% CI 1.29–2.89) mortality. In the PSM cohort, the doubly robust estimate yielded consistent findings (30-day HR 2.07, 95% CI 1.34–3.21; 90-day HR 1.93, 95% CI 1.29–2.90). Sensitivity analyses treating LAR as a continuous variable and as tertiles confirmed a dose–response relationship…

How strong is the evidence?

Evidence Hub reads this as a secondary / pooled analysis, done in people. Its evidence level is 3 of 5 (Moderate · trial / cohort / secondary analysis). Trials without a large randomized design, cohorts and secondary analyses give moderate evidence. Reported size: 499 participants or records.

How the five levels of evidence work. The level is assigned automatically from the abstract and can be wrong, so check the paper.

Read the paper

Related studies

Go deeper in Evidence Hub

Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.

Open this study Sign up free