Key result
Adding NLR and LAR to the GWTG-HF score improves in-hospital mortality risk discrimination.
Why the study?
Severe heart failure carries high in-hospital mortality, yet the GWTG-HF score omits prognostically linked markers of inflammation, hypoxic injury, and nutritional status. It was unknown whether the neutrophil-to-lymphocyte ratio (NLR) and lactate dehydrogenase-to-albumin ratio (LAR) add incremental prognostic value beyond the GWTG-HF score and NT-proBNP.
Does adding the neutrophil-to-lymphocyte ratio (NLR) and lactate dehydrogenase-to-albumin ratio (LAR) to the GWTG-HF score improve the prediction of all-cause in-hospital mortality compared with the GWTG-HF score alone or with NT-proBNP in hospitalized adults with severe heart failure?
Cohort (n=1,974)
Yes
Does adding the neutrophil-to-lymphocyte ratio (NLR) and lactate dehydrogenase-to-albumin ratio (LAR) to the GWTG-HF score improve the prediction of all-cause in-hospital mortality compared with the GWTG-HF score alone or with NT-proBNP in hospitalized adults with severe heart failure?
Combining routine admission biomarkers NLR and LAR with the GWTG-HF score provides consistent, low-cost incremental value for predicting in-hospital mortality in severe heart failure patients across external cohorts, outperforming the addition of NT-proBNP in external validation.
Answers are grounded in this paper and cite their sources.
May enhance in-hospital mortality prediction in severe HF when added to GWTG-HF; hypothesis-generating and requires prospective validation before clinical adoption.
Background Severe heart failure (HF; Killip or NYHA class III–IV) carries high in-hospital mortality, but the Get With The Guidelines–Heart Failure (GWTG-HF) score omits inflammation, hypoxic injury, and nutritional status, factors linked to prognosis. We tested whether the neutrophil-to-lymphocyte ratio (NLR) and lactate dehydrogenase-to-albumin ratio (LAR) add incremental value beyond NT-proBNP and this score. Methods In a dual-center retrospective cohort (derivation, n = 1596, 223 deaths; external validation, n = 378, 48 deaths), four nested logistic models, GWTG-HF alone or with NLR, LAR, or both, were assessed in derivation for discrimination, reclassification, and calibration; the combined model was bootstrap-validated for optimism and compared with GWTG-HF + NT-proBNP in both cohorts. Independent associations of NLR and LAR were assessed by multivariable logistic regression adjusting for GWTG-HF score, age, sex, log-transformed NT-proBNP, myocardial infarction history, diabetes, and moderate-to-severe chronic kidney disease. Without refitting (intercept recalibrated), discrimination, calibration, and decision-curve analysis, but not reclassification, were re-evaluated externally. Results LAR and NLR showed higher univariable discrimination than NT-proBNP in this severe-HF cohort (area under the curve [AUC] 0.732 and 0.704 vs. 0.673). Adding both raised derivation AUC from 0.750 to 0.777 (optimism-corrected 0.776; integrated discrimination improvement 0.032, p < 0.001; net reclassification improvement 0.059, p = 0.013), persisting externally (0.670 to 0.710, p = 0.039). The NT-proBNP model matched the gain (AUC 0.765) but lost significance externally (p = 0.914). LAR remained externally significant (p = 0.039); NLR did not (p = 0.085); only the combined model was significant on every metric assessed in both cohorts. NLR was independently associated with mortality (adjusted odds ratio 1.032, p = 0.001); LAR did not retain independent significance (p = 0.066); variance inflation factors were low (1.13–1.16), indicating partial informational overlap rather than collinearity. Conclusions Adding NLR and LAR to GWTG-HF improved mortality discrimination and, unlike NT-proBNP, this improvement persisted externally, albeit with uncertainty from case-mix differences; the two augmented models were not directly compared with each other.This suggests that markers related to inflammation, tissue injury, and nutritional status may carry prognostic information beyond current tools. Because each marker dominated in a different cohort, combining both, not either alone, provided the most consistent incremental value, supporting their use as a low-cost candidate for risk-stratification.
No takes yet. Share an insight, caveat, or question.
Xu et al. (2026) conducted a cohort in Severe heart failure (Killip class III–IV or NYHA functional class III–IV) (n=1,974). Combined neutrophil-to-lymphocyte ratio (NLR) and lactate dehydrogenase-to-albumin ratio (LAR) added to the GWTG-HF score vs. GWTG-HF score alone (and GWTG-HF score + NT-proBNP) was evaluated on All-cause in-hospital mortality (OR 1.032 for NLR; AUC 0.777 vs 0.750 (derivation) and 0.710 vs 0.670 (external validation), 95% CI 1.014-1.051, p=0.001). Adding combined NLR and LAR to the GWTG-HF score improved in-hospital mortality discrimination in derivation and external validation.
Synapse has enriched 10 related papers. Consider them for comparative context: