Prevalence and Predictors of DVT in Medical ICU Patients Undergoing Diagnostic Duplex Ultrasonography

Abstract

SESSION TITLE: Hematology/Oncology in the ICU SESSION TYPE: Original Investigation Slide PRESENTED ON: Monday, October 30, 2017 at 04:30 PM - 05:30 PM PURPOSE: Deep vein thrombosis (DVT) is a recognized but preventable cause of morbidity and mortality in the ICU. There is limited data on the risk of DVT or predisposing factors among non-surgical ICU patients with multiple medical comorbidities. Our aim is to retrospectively examine the prevalence and risk factors for proximal DVT and potential effect on clinical outcomes in critically ill medical patients in which diagnostic duplex ultrasonography (DUS) was performed. METHODS: Retrospective case-control study examining the prevalence of DVT in 678 patients admitted to two tertiary care level academic Medical ICUs (MICU) from July 2014 to July 2015 in the Bronx, NY. Persons with a diagnosis of proximal DVT on DUS were compared to a cohort of patients without DVT. Sequential Organ Failure Assessment (SOFA) scores were calculated for all patients to assess disease severity and comparability. Primary outcomes were ICU length of stay and mortality. Secondary outcomes were need for invasive mechanical ventilation (IMV), hemodialysis, sepsis, SOFA scores, vasopressor/inotrope use, central venous catheter (CVC) placement, history of DVT and malignancy. Stepwise logistic regression analysis was utilized to determine predictors of DVT occurrence. RESULTS: Of the 678 patients admitted to the MICU, 245 (36%) patients underwent duplex ultrasonography (DUS) to evaluate for DVT. 141 (57%) were male, mean age was 64.4±16.7 years, and BMI 29.8±11.5 kg/m2. Dyspnea (31%) and immobility (23%) were the most common indications for DUS testing. The prevalence of DVT was 19% (47) amongst tested patients. Of the patients with DVT, we noted increased rates of CVC placement (25% vs.15% p=0.05), and vasopressor/inotrope use (26% vs.16% p=0.05) compared to persons without DVT on DUS. Those on routine DVT thromboprophylaxis, chemical or mechanical, had fewer DVTs (15% vs 29% p=0.01). Between cases and controls, there were no significant differences in the need for IMV, hemodialysis, sepsis, SOFA scores, and history of malignancy. There were also no significant differences in ICU length of stay (DVT:4.4±6 days vs no DVT:4.8±59 days p=0.35) or mortality (DVT:10.6% vs no DVT:11.8% p=0.34). Stepwise regression modeling demonstrated a prior history of DVT was a significant predictor for development of DVT (OR 5.68, 95% CI [1.81, 17.83] p<0.01). CONCLUSIONS: A significant percentage of patients tested had proximal DVT even with appropriate use of DVT thromboprophylaxis. A history of DVT in a patient was the sole predictor for development of DVT on DUS testing in our MICU population. Our study demonstrated vasopressor/inotrope use and the central venous catheters used for medication delivery were associated with presence of DVT. Sepsis was not an independent risk factor for DVT. CLINICAL IMPLICATIONS: Despite the use of universal, guideline-recommended thromboprophylaxis, critically ill patients still demonstrate high rates of DVT. In the presence of known risk factors elicited from our cohort - history of DVT, CVC placement, and use of vasopressors/inotropes - our study highlights the need for targeted screening via DUS in the medical ICU. Further studies with larger patient populations are needed to fully elucidate the impact of acute DVT on morbidity and mortality in the medical ICU. DISCLOSURE: The following authors have nothing to disclose: Charlisa Gibson, Mai Colvin, Michael Park, Qingying Lai, Gavin Harris, Chirag Shah, Matthew Langston No Product/Research Disclosure Information