In Reply We thank Di Nisio and colleagues for their insights on our prospective management study1 for suspected upper-extremity deep vein thrombosis (DVT) in which anticoagulation was withheld only on the basis of initial negative ultrasound test results. Di Nisio and colleagues correctly point out a limitation of our study (ie, study participants from a single tertiary care center with 5 of 337 [1.5%] with a negative ultrasonographic reading lost to follow-up) in comparison to their ARMOUR diagnostic management study.2 In the latter study, an algorithm including a clinical prediction rule (ie, the Constans score3) and D-dimer was employed to limit the number of ultrasonographic screenings performed resulting in ultrasonography being withheld in 21% of participants.2 However, the sample of patients with an unlikely score and normal D-dimer level was limited and is reflected by the confidence interval of the failure rate of 0 of 87 (95% CI, 0.0%-4.2%) in this subgroup of subjects.