Abstract
BACKGROUND: Sepsis with high incidence and mortality, is a major public health problem. Anemia is one of the common complications of sepsis; a single-point value of hemoglobin cannot provide enough information. This study aims to identify hemoglobin trajectories in sepsis patients, assess their relationship with clinical outcomes, and develop a nomogram to predict high-risk trajectories, providing a basis for disease monitoring and blood management. METHOD: Sepsis patients in the ICU for over 3 days with continuous hemoglobin data were included. excluding those with conditions affecting hemoglobin. The latent growth mixture model was used to identify the latent trajectories of hemoglobin, and Cox regression and survival analysis were performed to evaluate the association between hemoglobin trajectories and clinical outcomes. And a nomogram was constructed to predict the risk of rapid hemoglobin decline, and its performance was thoroughly evaluated. RESULTS: Of 811 patients ultimately met the inclusion criteria, and after latent growth mixed model analysis, optimal 3 potential hemoglobin trajectories were detected: slow decline (54.4%), persistently low (34.7%), and rapid decline (10.9%). Persistently low hemoglobin correlated with higher 28-day hospital mortality (HR 2.39, 95% CI 2.11–2.64) and organ dysfunction (HR 3.45, 95% CI 2.48–4.80). Rapidly decreasing hemoglobin significantly increased 28-day hospital mortality (HR 2.89, 95% CI 2.47–3.37 and organ failure risks (HR 5.44, 95% CI 4.09–7.25). Survival analysis confirmed poorer outcomes in the rapid decline group. The nomogram for predicting rapid decline of hemoglobin demonstrated good discrimination and calibration. CONCLUSION: Hemoglobin trajectory patterns are important prognostic indicators in patients with sepsis. A rapid decline in hemoglobin is associated with substantially increased risks of 28-day hospital mortality, organ dysfunction, and poor survival outcomes. The developed nomogram may serve as a useful clinical tool to identify high-risk patients. These findings highlight the importance of dynamic hemoglobin monitoring in sepsis and suggest that risk stratification and targeted clinical interventions based on Hb trajectory may improve patient outcomes. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12879-025-12300-8.