Central venous catheters and upper extremity deep vein thrombosis in medical inpatients: the Medical Inpatients and Thrombosis (MITH) Study

Background: Upper extremity deep vein thrombosis (UEDVT) is an increasingly recognized complication in medical inpatients, with few data available regarding the incidence, risk factors and association with central venous catheter (CVC) use. Methods: Between 2002 and 2009 all cases of hospital-acquired venous thromboembolism (VTE) at a university hospital were frequency matched 1 : 2 to non-cases without VTE by admission year and medical service. Records were abstracted to identify, characterize and assess risk factors for UEDVT. Weighted logistic regression was used to calculate odds ratios (ORs) for UEDVT associated with use of a CVC, adjusting for known VTE risk factors. Results: Two hundred and ninety-nine cases of VTE complicated 64 034 admissions to medical services (4.6 per 1000 admissions). UEDVT constituted 51% (91/180) of all deep vein thrombosis (DVT), for an incidence of 1.4 per 1000 admissions (95% confidence interval [CI], 0.8-1.7). There were 247 CVCs placed per 1000 admissions (95% CI, 203-292). The use of a CVC was associated with a 14.0-fold increased risk of UEDVT (95% CI, 5.9-33.2), but was not associated with a significantly increased risk of PE (OR, 1.3; 95% CI, 0.8-2.1). Peripherally inserted central catheters had a higher OR for UEDVT (OR, 13.0; 95% CI, 6.1-27.6) than centrally inserted central venous catheters (CICC) (OR, 3.4; 95% CI, 1.7-6.8). Conclusion: UEDVT is a relevant complication affecting medical inpatients, accounting for half of hospital-acquired DVTs. Use of CVCs was strongly associated with risk of UEDVT.