Lisette van Dam

General discussion and summary 13 193 incidence of recurrent VTE was low in patients with MRDTI negative for DVT (1.1%; 95% confidence interval (95%CI) 0.13-3.8) and in patients with MRDTI negative for DVT and thrombophlebitis, who were not treated with any anticoagulant during follow-up (1.7%; 95%CI 0.20-5.9). Moreover, with an excellent interobserver agreement and MRDTI not being available in only 3.6% of patients, this technique was proven to be a feasible and reproducible diagnostic test. Additionally, the use of MRDTI possibly resulted in 19% fewer false positive diagnoses. However, as a MRI scan is more expensive than a CUS examination, the cost-effectiveness of the addition of MRDTI to the diagnostic management of suspected recurrent DVT needed to be proven before this technique can be widely implemented. In Chapter 3 , the one-year healthcare costs of 10 diagnostic strategies including the Wells score in combination with a D-dimer test, CUS and/or MRDTI in suspected recurrent DVT was assessed and compared. We showed that the healthcare costs of strategies with MRDTI were generally lower than of strategies without MRDTI, due to superior specificity resulting in less false-positive diagnosis and overtreatment. Therefore, it was concluded that application of MRDTI to the diagnostic management of suspected recurrent ipsilateral DVT will not increase healthcare costs. Healthcare costs can also be reduced with the use of diagnostic algorithms in which VTE can be ruled out without additional imaging tests, i.e. in case of a low clinical probability in combination with a negative D-dimer test. As discussed, the safety of these algorithms has not yet been proven in large prospective studies exclusively in patients with suspected recurrent DVT. In Chapter 4 , we therefore evaluated the diagnostic accuracy of the combination of the (modified) Wells rule for DVT and D-dimer for suspected recurrent ipsilateral DVT in the prospective Theia study. We showed that excluding recurrent DVT based on a low clinical probability according the (modified) Wells rule in combination with a negative D-dimer test would have resulted in an unacceptable high failure-rate (6.1-11%). Our data therefore suggest not to routinely apply assessment of a clinical decision rule and D-dimer in the diagnostic workup of suspected recurrent DVT, but to directly perform a CUS to exclude or diagnose recurrent DVT. In patients with a suspected recurrent ipsilateral DVT and inconclusive CUS, MRDTI should be performed to provide a definitive diagnosis. The diagnostic management of upper extremity deep vein thrombosis (UEDVT) is also very challenging due to the anatomic location of the deep veins. Adequate visualization is difficult as the upper extremity deep veins lies partly within the