Four distinct dynamic D-dimer trajectories and their prognostic implications in acute pulmonary embolism: A...

Acute pulmonary embolism remains a major cause of cardiovascular mortality, and static D-dimer measurements may not fully capture short-term mortality risk.

We conducted a retrospective cohort study of adult intensive care unit patients with acute pulmonary embolism. Latent class growth modeling (LCGM) identified D-dimer trajectories during the first 10 days following acute pulmonary embolism diagnosis. Associations with 30-day all-cause mortality were assessed using Cox proportional hazards models and Kaplan-Meier analysis. Incremental prognostic value beyond the baseline prediction model was evaluated using area under the curve (AUC), integrated discrimination improvement (IDI), net reclassification improvement (NRI), and the median improvement in risk score. Sensitivity analyses included landmark analyses and an alternative five-class Trajectory Model.

Among 303 patients, four D-dimer trajectories were identified: low and stable, mildly elevated and stable, highly elevated and decreasing, and persistently elevated. Compared with the "low and stable" trajectory, the "persistently elevated" trajectory was associated with the highest 30-day all-cause mortality in both unadjusted (HR, 15.06; 95% CI: 7.61-29.80; P < 0.001) and the fully adjusted model (HR, 5.08; 95% CI: 2.32-11.09; P < 0.001). Survival curves differed significantly across trajectory classes. Adding D-dimer trajectory membership to the Baseline Model improved discrimination (increased the AUC from 0.90 to 0.93; P = 0.04) and reclassification (IDI: 0.07; NRI: 0.28). The findings remained robust in subgroup analyses, landmark analyses, and the alternative five-class model.

Distinct D-dimer trajectories were associated with 30-day all-cause mortality. The persistently elevated trajectory identified a high-risk subgroup and improved mortality prediction.