Central Venous Catheters, a Major Culprit for Pediatric Venous Thrombosis: A Report from the Children's Hospital Acquired Thrombosis (CHAT) Project

Introduction: Incidence of venous thromboembolism (VTE), a potentially life-threatening condition, has increased in hospitalized pediatric patients by over 70% in the last decade leading to increased morbidity, mortality, and longer hospitalizations. In contrast to standardized thromboprophylaxis interventions for hospitalized adults, routine implementation of VTE prevention strategies have not been established in children. To develop a risk-prediction model for pediatric hospital-acquired VTE (HA-VTE) and identify a high-risk group in need of thromboprophylaxis, an observational case-cohort study was created, using the multi-institutional Children's Hospital Acquired Thrombosis (CHAT) Registry. Preliminary CHAT data confirm prior reports that central venous catheters (CVCs) are the greatest risk factor for pediatric HA-VTE. We describe clinical and CVC characteristics for Registry subjects with a CVC-related HA-VTE.

Methods: An institutional review board-approved, multi-center CHAT Registry was created for this project to collect retrospective clinical and laboratory data on hospitalized subjects age 0-21 years diagnosed with imaging-confirmed HA-VTE, along with admission year matched controls. Seven hundred and fifty HA-VTE subjects hospitalized since January 1, 2012 are included in the Registry. Data include demographics (age and past medical history); hospital course (duration, infection, mobility, surgeries, trauma, etc.); VTE specifics (location, symptoms, treatment, prophylaxis); CVC details (type, brand, lumen number, duration, infection history); and pertinent laboratory data (thrombophilia and coagulation testing). Pertinent CVC related data were selected for analysis.

Results: To date, 658 subjects with HA-VTE have been entered into the CHAT Registry from 01/01/12 to 07/01/2017 from five pediatric tertiary care centers. CVC-VTE was identified in 486 (74%) subjects. Most subjects with a CVC-VTE were male (58%) and the median age at diagnosis was 2 years (IQR 0.3-12.7). Infants <1 year of age accounted for 214 (44%) of CVC-VTE cases (Figure 1). In contrast, subjects diagnosed with a non-CVC-related VTE had a median age of 5 years (IQR 0.3-13.3) at diagnosis, with only 45 (29%) being <1 year. The average time from CVC placement to VTE diagnosis was 23 days. Time from femoral line placement to VTE was 7 days, peripherally inserted central catheters (PICC) placement to VTE was 16 days and tunneled line to VTE was 114 days. Mean hospital length of stay was 48 days for those diagnosed with a CVC-VTE.

Subjects with PICCs made up 62% (310) of the CVC-VTE cases and 19% (95) were in subjects with femoral catheters (Figure 2). The majority of CVC-VTEs, 336 (67%), were with multi-lumen CVCs.Upper extremity CVCs were placed in 271 (54%) cases and 203 (41%) were in the lower extremity. The remaining 26 (5%) had CVCs in the umbilical vein or were unknown. CVC's were found to have a malfunction (damage to the line, fibrin sheath, medication precipitation or other blockage) prior to VTE diagnosis in 122 cases (24%), and tissue plasminogen activator (tPA) was used in 91 (75%) cases of malfunction.

Conclusions: CVCs are a prominent risk factor for pediatric HA-VTE, yet a paucity of evidence exists to guide clinicians toward strategies to decrease pediatric CVC-VTE. Our preliminary analysis has shown the majority of subjects with a CVC-related HA-VTE were male, had a PICC, and almost half were less than 1 year of age. Subjects also tended to have multi-lumen CVCs placed in the upper extremity. This novel analysis of specific CVC characteristics in CVC-related HA-VTE among hospitalized patients from the CHAT Registry represents a first step in determining the true risk factors for CVC-related VTE in children.

Once all control subjects are entered a more robust analysis of risk factors will be performed and a VTE risk prediction model will be created and prospectively validated for HA-VTE, including a specific CVC-related VTE model. This information will be essential for future risk-prediction models of pediatric HA-VTE that may guide randomized control trials to evaluate safety and efficacy of various thromboprophylaxis strategies in this population.

View largeDownload slide

View largeDownload slide

Close modal