Characteristics of Upper Extremity Deep Vein Thrombosis in Hospitalized Patients. a Review of 8350 Admissions from the 2012 National Inpatient Sample

Background: The incidence of upper extremity deep vein thrombosis (UEDVT) is increasing due to the increased use of central venous catheters. In contrast to lower extremity DVT (LEDVT), there is limited data to guide management, with many current management recommendations based on extrapolation from LEDVT studies. The aim of this study is to evaluate the characteristics and mortality associated with hospitalized patients with UEDVT from a large national database.

Methods: Using the 2012 National Inpatient Sample (NIS), admissions with the International Classification of Diseases, 9^th Revision, Clinical Modification (ICD-9-CM) code for UEDVT (453.82) were extracted, and data collected on age, sex, length of stay, mortality, and associated diagnoses and procedures. LEDVT (ICD-9-CM 453.40) data were extracted for comparison. Collected data were used to calculate odds ratios with corresponding 95% confidence intervals and p-values to show associations where applicable.

Results: Of the 7,296,968 unweighted admissions in the 2012 NIS, 8,350 were associated with UEDVT, and 18,194 were associated with LEDVT (prevalence of 0.11% vs 0.25%, respectively). UEDVT and LEDVT rates were similar in men and women. However, compared to LEDVT, patients with UEDVT were younger (61yr vs. 67 yr; p<0.0001), had longer mean LOS (14 days (range 0-309) vs. 5 days (range 0-345); p<0.0001), and higher inpatient mortality (7.75% vs 5.34%; OR 1.45, 95% CI 1.31, 1.61; p<0.0001).  Pulmonary embolism was less common in UEDVT than in LEDVT (6.2% vs. 25%; OR 0.24, 95% CI 0.22, 0.27; p<0.0001), but significantly more common than the general hospitalized population (6.2% vs. 0.08%; OR 7.7, 95% CI 7.04, 8.42; p<0.0001). Malignancy was slightly more common in patients with UEDVT compared to those with LEDVT (30.1% vs 27.1%; OR 1.11, 95% CI 1.05, 1.17; p=0.0002). UEDVT was associated with central venous catheter placement (OR 10.1, 95% CI 9.56, 10.68; p<0.0001), other venous catheter placement (OR 21.9, 95% CI 20.9, 22.9; p<0.0001), and red blood cell transfusion (OR 3.6, 95% CI 3.47, 3.83; p<0.0001) in the hospitalized population. Procedures most commonly associated with LEDVT were interruption of the vena cava, red blood cell transfusion and other venous catheter placement.

Conclusions: UEDVT was less common than LEDVT, and associated with central venous catheter placement and blood transfusions in hospitalized patients. UEDVT tended to occur in younger patients with longer LOS, and was associated with a higher mortality and a slightly higher prevalence of malignancies than LEDVT. PE was less common during admissions with UEDVT than LEDVT, however UEDVT admissions were associated with an increased incidence of PE compared to the general inpatient population. The results of this large epidemiologic study of UEDVT in hospitalized patients contribute to our understanding of this increasingly common disease and will help define strategies for prophylaxis and treatment.