Sulodexide versus low-molecular-weight heparin in critically ill neonates: a propensity score–matched retrospective cohort study

Background Critically ill neonates are at high risk of thrombosis, inflammation, and antimicrobial exposure in the neonatal intensive care unit, but evidence guiding adjunctive anticoagulant selection remains limited. We compared the clinical, laboratory, and safety outcomes associated with sulodexide (SDX) and low-molecular-weight heparin (LMWH). Methods This single-center retrospective cohort study included critically ill neonates treated with sulodexide or low-molecular-weight heparin. Propensity score matching (PSM) yielded 85 matched pairs. The primary outcome was relative percentage change in D-dimer from baseline to treatment end. Secondary outcomes included other coagulation, hematological, blood-gas, and respiratory-support measures. Inflammatory status and antimicrobial defined daily doses (DDDs) were exploratory outcomes. Bleeding events first documented after treatment initiation were assessed as safety outcomes. Results Among 234 eligible neonates, 85 matched pairs were analyzed. D-dimer change did not differ significantly between groups [−70.2% (−82.0, −50.5) vs. −72.0% (−87.1, −42.2); between-group effect, 1.4 percentage points; 95% CI, −8.3 to 9.7; P = 0.765], and no other coagulation, hematological, or blood-gas outcomes differed significantly. Among neonates with abnormal baseline inflammatory status, CRP/hs-CRP improvement occurred in 12/12 (100.0%) and 7/11 (63.6%) in the SDX and LMWH groups, respectively (P = 0.037), suggesting a potential difference in inflammatory response. SDX remained associated with lower observed DDDs after multivariable adjustment, although this association was not robust in calendar-period sensitivity analyses. Bleeding was first documented after treatment initiation in 24/85 (28.2%) SDX-treated and 18/85 (21.2%) LMWH-treated neonates, respectively. Among neonates with bleeding, first documentation occurred during active treatment in 9/24 (37.5%) and 14/18 (77.8%), respectively. Conclusion No significant between-group differences were observed in the primary D-dimer or other coagulation outcomes. A higher proportion of CRP/hs-CRP improvement was observed in the SDX group, although this finding was exploratory and based on a small subgroup. The lower observed DDDs were not robust to calendar-period analyses and should not be attributed to a drug-specific effect. Bleeding was numerically more frequent in the SDX group overall, whereas a greater proportion of bleeding cases in the LMWH group were first documented during active treatment. The observed timing pattern does not establish a causal relationship with treatment. These retrospective findings require cautious interpretation and prospective confirmation.