Prognostic performance of an 8-zone lung ultrasound B-line score combined with 24-hour lactate clearance for 28-day all-cause mortality in adults with sepsis: a single-center retrospective cohort study

Background/objectives Sepsis is frequently accompanied by pulmonary interstitial edema and impaired tissue perfusion, which contribute to poor outcomes. Although transpulmonary thermodilution provides a reference assessment of extravascular lung water, its invasive nature limits routine use. This study aimed to evaluate whether a combination of lung ultrasound B-line score and 24-hour lactate clearance could improve prediction of 28-day mortality in adults with sepsis.Notably, due to the requirement of 24-hour lactate measurement for model construction, this prognostic model is only applicable to sepsis patients who survive until the 24-hour assessment, and cannot be directly generalized to all sepsis patients admitted to the ICU. Methods This retrospective single-center cohort study enrolled adult patients with sepsis admitted to the intensive care unit (ICU) between January 2023 and December 2025. Archived 8‑zone lung ultrasound cine‑loop images were independently re‑assessed by two blinded, specially‑trained ICU nurses. Both nurses completed the Advanced Training Program in Critical Care Ultrasonography, obtained corresponding professional training certificates, and have accumulated more than 10 years of clinical experience in ICU nursing. Patients were classified into low-, intermediate-, and high-risk groups according to predefined cutoffs for B-line score and 24-hour lactate clearance. The prognostic performance of individual and combined markers was assessed using receiver operating characteristic curves and multivariable Cox regression analysis. The conditional target population of this study is defined as: adult sepsis patients admitted to the ICU who have been hospitalized for at least 24 hours and survive until the completion of 24-hour lactate assessment. Patients who die or are discharged within 24 hours after ICU admission are excluded, as they cannot provide complete 24-hour lactate clearance data; thus, the conclusions of this study do not apply to this subgroup. Results A total of 560 patients were included, and 153 (27.3%) died within 28 days. The inter-rater reliability of the B-line score was excellent (ICC, 0.912; 95% CI, 0.895–0.927). Compared with survivors, nonsurvivors had higher B-line scores and lower 24-hour lactate clearance (both P < 0.001). The combined model showed better discrimination for 28-day mortality than either marker alone, with an area under the curve of 0.889 (95% CI, 0.853–0.925). Compared with the low-risk group, the intermediate- and high-risk groups showed significantly higher mortality risk (adjusted HR, 2.164 and 4.327, respectively). Conclusions Among adult sepsis ICU patients who survive until the 24‑hour assessment, the combination of an 8‑zone lung ultrasound B‑line score and 24‑hour lactate clearance showed within‑cohort mortality risk‑stratification ability. Its incremental value over conventional severity scores remains unproven. Prospective multicenter validation is required before any clinical application.