{Online Passive-Aggressive Algorithms for Non-Negative Matrix Factorization and Completion}

Abstract

It has been suggested that upper extremity deep venous thrombosis (UEDVT) is as common and dangerous as lower extremity deep venous thrombosis. Pulmonary embolism (PE) is often found with no evidence of associated lower extremity deep venous thrombosis and could have originated from UEDVT. Routine screening is well accepted for lower extremity deep venous thrombosis but not for UEDVT. We hypothesized that UEDVT in trauma is frequent but undetected; therefore, routine screening of trauma patients at risk will increase the UEDVT rate and decrease the PE rate due to early diagnosis and treatment. We evaluated the incidence of UEDVT and PE over 6 months before (Group BEFORE) and 6 months after (Group AFTER) implementing a policy of screening patients at high risk for deep venous thrombosis with Duplex ultrasonography. Group BEFORE was evaluated retrospectively and group AFTER prospectively. There were 1110 BEFORE and 911 AFTER patients. The two groups were similar. Of the AFTER patients, 86 met predetermined screening criteria and were evaluated routinely by a total of 130 Duplex exams. One patient in each group developed UEDVT (0.09% vs. 0.11%, P = 1.00). The brachial vein was involved in both patients. Six BEFORE (0.54%) and 1 AFTER (0.11%) patients developed PE (P = 0.137). The single AFTER patient with PE was not screened for UEDVT because he had no high-risk criteria. UEDVT is an uncommon event with unclear significance in trauma. Aggressive screening did not result in a higher rate of UEDVT diagnosis, nor an opportunity to prevent PE.