Critically ill patients have multiple risk factors for deep venous thrombosis (DVT), a leading preventable cause of death among hospitalized patients. Upper extremity (UE) DVT occurs more commonly than previously thought, and some UEDVT may result in pulmonary embolism (PE). We examined risk factors and outcomes among ICU patients with UEDVT.
We performed a retrospective case-control study of 147 critically ill patients from July 2000 through June 2003 in a 64 bed multidisciplinary ICU of a large university hospital to evaluate risk factors for UEDVT, including previous DVT, UE central venous catheterization (CVC), obesity (BMI > 30), malignancy, mechanical ventilation, DVT prophylaxis, surgery within 30 days, and vasopressor use. Cases were all patients with ultrasound evidence of UEDVT (n= 45). One control group included all patients with negative UE ultrasounds (n= 57), and another control group included ICU patients matched by age, gender, and primary diagnosis (n= 45). We excluded patients with known DVT or PE, or anticoagulant use prior to admission. Outcomes included documented PE, duration of mechanical ventilation, mortality, hospital length of stay (LOS), and ICU LOS.
Significant risk factors for UEDVT included history of DVT, UE CVC, requirement for vasopressors, and mechanical ventilation . 6 patients had documented PE, all of whom had UEDVT. There were no differences in mortality or ICU LOS ( Tables 1Table 1.
Case vs. control.Odds Ratio (95% CI)PRisk factorHistory of DVT6.8 (1.4-33.7)<.01Upper extremity CVCNC1.07OutcomesPulmonary embolism2.8 (0.6-11.8).181
Not calculated. 45 out of 97 (46%) patients with upper extremity CVC had UEDVT, 0 out of 5 (0%) patients without upper extremity CVC had UEDVT. Odds ratio could not be calculated. P value trended towards significance by Fisher’s Exact Testand 2Table 2.
Case vs. matched controlOdds Ratio (95%CI)PRisk factorHistory of DVT11.0 (1.3-90.5).02Upper extremity CVCNC1<.01Vasopressor requirement2.8 (1.2-6.6).02Mechanical ventilation6.3 (2.5-6.0)<.01OutcomesPulmonary embolismNC2.031
Not calculated. 45 out of 79 (57%) patients with upper extremity CVC had UEDVT, 0 out of 11 (0%) patients without upper extremity CVC had UEDVT. Odds ratio could not be calculated. P value obtained significance by Fisher’s Exact Test.2
Not calculated. 6 out of 45 (13.3%) patients with UEDVT had PE, 0 out of 45 (0%) patients without UEDVT had PE. Odds ratio could no be calculated. P value obtained significance by Fisher’s Exact Test.).
Critically ill patients with history of DVT, UE CVC, need for vasopressors, or mechanical ventilation are at greater risk for development of UEDVT. UEDVT appears to be a risk factor for PE in critically ill patients.
Critically ill patients with prior DVT, UE CVCs, or prolonged mechanical ventilation should be closely monitored for the development of UEDVT and pulmonary embolism.
P. Lee, None.