Epidemiology, diagnosis and management of neonatal thrombosis: a single-center cohort study.

The incidence of neonatal venous and arterial thrombosis ranges from 6.9 to 15/1000 neonatal ICU (NICU) admissions, and is likely an underestimate based on population demographics, frequency of surveillance and vascular catheterization. This retrospective study involving 234 infants reviewed the epidemiology, diagnosis, and management of neonatal thrombosis in a single, tertiary care institution over more than 10 years. The incidence of thrombosis was 25/1000 NICU admissions, with a preterm to term infant ratio of 1.5 : 1 and a slightly higher proportion of male sex (55.1%). The mean (range) gestational age and birth weight was 33.8 weeks (23-41.6) and 2360 g (512-5890). The median age (IQR) of thrombus diagnosis was 7 (3-17) days. Portal vein thrombosis was most prevalent (59.4%) compared with other sites of thrombosis. Almost three-quarter (171/234; 73.1%) of the thrombotic episodes were line-related, while infection and surgery were associated with 19.7% (46/234) and 10.7% (25/234), respectively. Twenty patients (8.3%) were screened for thrombophilia and 3 were positive; 2 for antithrombin deficiency, 1 for factor V Leiden gene mutation. Subjects were followed with imaging for 3 months with a treatment duration, mean (IQR) of 33.5 (10.8-42.5) days. Complete clot resolution was significantly higher in the anticoagulation group (48%; 17%; P = 0.03) compared with untreated patients. No group difference was noted for partial thrombus resolution (33.3%; 12.4%; P = 0.313). Anticoagulation halted thrombus progression (2.6 versus 12.4%; P = 0.025) and fewer treated patients failed to attend follow-up visits (6.5 versus 18.6%; P = 0.022). Well designed, multicenter prospective studies with larger sample sizes are required to confirm these findings.