Incidence and risk factors associated with venous thrombotic events in pediatric intensive care unit patients* Renee A. Higgerson, MD; Karla A. Lawson, PhD; LeeAnn M. Christie, MSN; Ann-Marie Brown, PNP; Jennifer A. McArthur, DO; Balagangadhar R. Totapally, MD; Sheila J. Hanson, MD; for the National Association of Children’s Hospitals and Related Institution’s Pediatric Intensive Care Unit FOCUS group S troke, renal, or hepatic vein thrombosis, pulmonary embo- lism, and deep vein thrombosis comprise a group of thromboem- bolic events, primarily venous thromboem- bolic events (VTEs), that cause significant morbidity and mortality in adults older than 40 yrs. Until recently, thromboembolic dis- ease has been considered a rare entity in pe- diatric patients (1–3). Hospitalized children do not routinely receive prophylaxis against VTEs, as is the standard of care for many hospitalized adult patients. The advent of computerized databases has allowed the col- lection of data over long periods of time and from multiple centers simultaneously. Anal- yses of these large databases demonstrate that the incidence of VTEs in pediatric patients is significantly higher than previously sus- pected, the incidence is increasing, and it still may be under-reported (1–5). More recently, prospective studies also have shown higher rates of VTEs than pre- viously suspected, especially those related to central venous catheters (CVC) (6, 7). However, these data primarily represented asymptomatic VTEs. We prospectively stud- ied the incidence of clinically detected VTEs in a diverse group of patients admitted to 11 North American pediatric intensive care units (PICUs) over a rolling 6-month interval. EXPERIMENTAL DESIGN AND METHODS Cohort and Case Ascertainment After obtaining local Institutional Review Board approval or exemption, institutions were provided with training to collect data through the National Association of Children’s Hospitals and Related Institution’s PICU FOCUS Group. Data for the entire cohort were collected through the VPS LLC, a clinical da- tabase that connects a global network of chil- dren’s hospitals and standardizes data sharing and benchmarking between PICUs. These data included patient demographics (age, race, gender), visit-specific data (reason for admis- sion to the PICU, date of admission), and clin- ical data (presence and dates of CVC and me- chanical ventilation). Additional data for cases were collected separately by clinicians, includ- ing patient symptoms leading to suspicion of VTEs, location of VTEs, CVC characteristics (if applicable), and confirmatory studies. Patient history of diagnosed prothrombotic disorder and therapeutic interventions were collected. All data underwent a blinding process before export to National Association of Children’s Hospitals and Related Institution in accor- dance with Health Insurance Portability and Accountability Act regulations. The cohort contained all patients admitted to the participating PICUs over a rolling 6-month period between April 2006 and De- cember 2007. Each unit collected data on se- *See also p. 678. From the Pediatric Intensive Care Unit (RAH, LMC) and Trauma Services (KAL), Dell Children’s Medical Center of Central Texas, Austin, TX; Pediatric Intensive Care Unit (A-MB), Akron Children’s Hospital, Akron, OH; Critical Care Section (JAM, SJH), Children’s Hospital of Wisconsin and Medical College of Wisconsin, Milwau- kee, WI; and Pediatric Intensive Care Unit (BRT), Miami Children’s Hospital, Miami, FL. The authors have not disclosed any potential con- flicts of interest. For information regarding this article, E-mail: kalawson@seton.org Copyright © 2011 by the Society of Critical Care Medicine and the World Federation of Pediatric Inten- sive and Critical Care Societies DOI: 10.1097/PCC.0b013e318207124a Objective: To evaluate the incidence and risk factors associ- ated with venous thromboembolism (VTE) in children admitted to pediatric intensive care units (PICUs). Design: Prospective observational study. Setting: Eleven tertiary care PICUs in the United States. Patients: Children who were admitted to PICUs and had radio- graphically confirmed VTE over a rolling 6-month period were enrolled in the study. Demographic, patient-related, and out- comes data were collected and compared with all children ad- mitted during the same period. Interventions: None. Results: Sixty-six symptomatic VTE were documented in sixty- two patients among 6653 patients admitted to 11 PICUs. Thirteen (19.7%) of the thrombi were present on admission. The incidence rate was 0.74% (range, 0 –2.7% per PICU) with a point prevalence of 0.93%. Doppler ultrasound was most frequently used to diagnose or confirm a suspected VTE. Variables associated with unadjusted risk for VTE include: younger age (3.8 months for patients with VTE vs. 51 months for non-VTE patients, p < .001), cardiac diagnosis (41% in VTE cases vs. 15% in non-VTE, p < .001), pre-/post-operative status (63% in VTE cases vs. 40% in non-VTE, p  .001), presence of central venous catheter (88% in VTE case vs. 17% in non-VTE, p < .001), or mechanical ventilation (85% in VTE cases vs. 30% non-VTE, p < .001). Multivariate analysis showed increased risk of VTE with CVC (odds ratio 6.9; confidence interval 2.7–17.5) and mechanical ventilation (odds ratio 2.8; confidence interval 0.98 –7.93). Children with VTE were sicker (Pe- diatric Index of Mortality 2 score risk of mortality of 3.0% vs. 0.9%; p<0.0001), stayed longer in the ICU (21.2 days vs. 1.6 days; p < .0001) and had increased mortality (10.2% vs. 2.6; p < .0001). Conclusions: Children admitted to the PICU have an increased risk of VTE. The presence of a CVC is the strongest risk factor for VTE in this PICU population. Children with VTE were younger, sicker, stayed longer in PICU, and had a higher mortality rate. (Pediatr Crit Care Med 2011; 12:628 – 634) KEY WORDS: anticoagulation; central venous catheter; intensive care unit; pediatric thrombus; venous thromboembolism 628 Pediatr Crit Care Med 2011 Vol. 12, No. 6