Facebook Twitter YouTube

Hyperglycemia Increases the Risk of Deep Venous Thrombosis In Non-Diabetic Critically Ill Children

Sunday, October 21, 2012: 8:30 AM
Room 210 (Morial Convention Center)
Joana Anjeh Tala, MD, Pediatric Intensive Care Unit, Yale New Haven Hospital, New Haven, CT, Sree Pemira, MD, Pediatrics, Yale University School of Medicine, New Haven, CT, Cicero T. Silva, MD, Diagnostic Radiology, Yale School of Medicine, New Haven, CT and E. Vincent S. Faustino, MD, Pediatrics, Yale School of Medicine, New Haven, CT

Purpose /Background 

Deep venous thrombosis (DVT) is common in diabetic children admitted to the intensive care unit (ICU) with ketoacidosis. It is unclear whether hyperglycemia, in the absence of diabetes mellitus, increases the risk of DVT in critically ill children.

Objective: to determine the incidence, association and acute outcomes of DVT in non-diabetic critically ill children with hyperglycemia


We performed a retrospective cohort study using data from critically ill children admitted to a tertiary ICU from 1/1/2007 to 12/31/2010. We included non-diabetic children ≤18 years on invasive mechanical ventilation (MV) or vasopressor support with blood glucose ≥150mg/dl. From each subject’s medical records, we collected data on the presence of radiologically confirmed DVT, maximum blood glucose prior to the diagnosis of DVT and patient characteristics. We calculated ventilator (VFD), ICU (ICUFD) and hospital free days (HFD) (defined as the number of days within 28 days after study inclusion that the subject is alive and off MV, ICU and hospital, respectively). We presented the incidence of DVT as proportion of subjects with DVT. Logistic regression was used to determine the association between DVT and hyperglycemia. We compared the VFD, ICUFD and HFD between subjects with and without DVT using t-test and linear regression


Of the 475 subjects included in the study, 27 developed DVT for an incidence of 5.7 per 100 subjects. This was significantly higher than the DVT rate for all ICU admissions during the study period (0.9 per 100 subjects, P<.001). Maximum blood glucose was significantly higher in those with DVT (309±170 mg/dl vs. 232±89 mg/dl, P=.03). After adjusting for age, diagnosis, presence of central venous catheter and vasopressor support, maximum blood glucose was independently associated with DVT with an adjusted odds ratio of 1.04 (95% confidence interval: 1.01-1.06, P=.01) for every 10 mg/dl increase in maximum blood glucose. Subjects with DVT had worse acute outcomes with less VFD, ICUFD and HFD. After adjusting for age, severity of illness score and maximum blood glucose, subjects with DVT had 2.9 less VFD (P=.11), 4.8 less ICUFD (P=.007) and 3.7 less HFD (P=.06) compared to those with no DVT.


In this cohort of critically ill non-diabetic children, hyperglycemia increases the risk of DVT. The incidence of DVT is higher than the general ICU population and children with DVT have worse acute outcomes.  The increased risk of DVT in this cohort of critically ill children suggests that in the appropriate setting, clinicians should have a low index of suspicion for working up these children for DVT. The data also suggest that these children may be targeted for future studies on thromboprophylaxis.