Risk Factors for Venous Thromboembolism In Hospitalized Cancer Patients with Central Catheters

Abstract

AbstractAbstract 810 Background:The incidence of venous thromboembolism (VTE) and central venous catheter (CVC) related thrombosis is increased in cancer patients, in particular those with advanced hematologic malignancies. Bloodstream infection (BSI) is thought to increase CVC associated thrombosis in several small studies. In July 2007, systematic changes including intensive education, process changes and use of the Biopatch.. (Ethicon 360) were effective in decreasing the incidence of BSI (from 3% to 0.7%) at the Hospital of the University of Pennsylvania (HUP).Our primary objective was to identify risk factors for VTE in hospitalized cancer patients, particularly hematologic malignancy and stem cell transplant (SCT) patients. Secondary objectives were to evaluate the effect of VTE prophylaxis and infection control measures in preventing VTE and CVC thromboses and describe treatment strategies in this population. Methods:After IRB approval was obtained, patients admitted to oncology floors were identified using an inpatient database. This was cross-referenced with an Interventional Radiology Hickman catheter placement database. Demographic and clinical data were retrieved retrospectively from the electronic medical record (EMR). Clinical events including objectively documented VTE were confirmed by review of the EMR by investigators. A case-control study was performed. Odds ratios were calculated for risk factors associated with thrombosis and univariate and multivariate regression analysis was performed. Results:Between July 1, 2005 and June 30, 2009, 770 oncology inpatients had a tunneled catheter placed. 50 patients with a VTE after the CVC was placed were identified as cases. 182 controls without VTE were identified randomly. Mean patient age with and without VTE was 52 and 53 years respectively. Women constituted 42% and 41% of patients with and without VTE. Diagnoses in cases include leukemia in 33 (66%), lymphoma in 7 (14%), myeloma in 4 (8%), MDS and aplastic anemia in 1 each (4%) and solid tumor in 4 (8%). Types of SCT in cases were allogeneic in 16 (62%), non-myeloablative in 4 (15%) and autologous in 6 (23%).Of the 50 cases, 15 (31%) had multiple VTEs, 26 (52%) were CVC associated, 17 (34%) were DVT and 4 (8%) were PE. No VTE prophylaxis was used in 22 (47%) of cases. Median platelet count at time of thrombosis was 86. Anticoagulant therapy was used in 33 patients (67%), including UFH in 28 (85%) and LMWH in 5 (15%). Bleeding during anticoagulation occurred in 4 cases (12%). With regard specifically to CVC thrombosis, 13 (50%) cases presented before infection control measures were put in place, and 13 (50%) after, vs. controls (38% before and 62% after) (p= 0.23). BSI was present in 16 (62%) cases and 51 (28%) of controls (p = 0.27). In a SCT subset analysis, VTE occurred in 26 (31%) of 83 SCT patients vs. 25 (17%) of 150 nonSCT patients (p=0.01); CVC thrombosis occurred in 14 (20%) SCT patients vs. 12 (9%) nonSCT patients (p=0.02).Risk factors for venous thrombosis by univariate analysis include:VariableOR95% CIp-valueRelapsed/refractory Heme Malignancy2.151.11–4.230.01History of VTE3.011.00–8.990.02History of Obesity6.251.84–24.070.003Post SCT10.203.52–29.55<0.001VTE prophylaxis0.530.28–0.990.03Left sided CVC2.331.03–5.140.02BSI1.550.82–2.910.14Biopatch ..1.540.83–2.780.14Active chemo0.420.20–0.880.01In multivariate regression analysis, the following risk factors remained statistically significant: status post SCT (p = 0.001); obesity (p = 0.02); line type- PICC (p = 0.02); CVC not in the internal jugular (IJ) vein (p=0.03), and CVC tip not in the superior vena cava (SVC) (p=0.003). Conclusions:Subjects with a history of VTE, obesity, advanced hematologic malignancies and SCT recipients were more likely to develop a VTE. Active chemotherapy did not increase the risk of VTE. CVC related risk factors include left sided placement, non- IJ vessel, and tip not in the SVC. While the presence of BSI was minimally associated with CVC- thrombosis, use of infection control measures did not significantly reduce VTE risk. VTE prophylaxis appears to be effective in preventing VTE; however 47% of cases were not on any, suggesting that at least mechanical measures may be warranted in high risk patients. Thrombocytopenia is not protective against VTE and should not preclude prophylaxis in these patients. Additional studies are warranted to further define risk factors for VTE in this population. Disclosures:Trerotola:Bard: Consultancy; MedComp: Consultancy; Cook: Consultancy; Teleflex: Consultancy; Grant Adler: Consultancy.