Below-knee deep vein thrombosis (DVT): Diagnostic and treatment patterns

Background: Guidelines are lacking for isolated below-knee deep vein thrombosis (BKDVT). The American College of Chest Physicians (ACCP) suggests anticoagulation only if symptoms are severe, otherwise surveillance with follow-up ultrasound in 2 weeks is recommended. Yet clinical practice remains highly variable. Little is known about the natural history of BKDVT and how to best manage these patients. Methods: A retrospective analysis was conducted of medical records from 2014-2016. Risk factors were assessed such as age, gender, malignancy, recent surgery, and history of DVT. Treatment decisions and outcomes were identified. Radiology reports were graded for clarity. Chi-square and logistic regression were used to correlate risk factors with outcomes and determine odds of treatment. Results: New isolated BKDVT was identified in 102 patients. Patients were symptomatic or had previous pulmonary embolism (PE) or DVT in all cases. Eighteen were positive for PE at diagnosis (17.6% of all patients or 62.1% with chest CT). Malignancy was independently associated with PE (P=0.015); no other risk factors were significant. Treatment was not associated with clinical risk factors. The language used for radiology reports was highly variable and was associated with the decision to treat with anticoagulation. One hundred percent were treated (n=24) when the report stated "positive DVT" and 89.2% (n=58) were treated after an objective description without the word "DVT" (P=0.01). Treatment was much less likely if the report described BKDVT anatomically but was said to be "negative for DVT" (P<0.001). A total of 86.3% (n=88) of all patients were treated, compared to 46.2% (n=6) of patients in this group (n=13). IVC filters were placed in 3 patients. Of the 14 untreated patients, 5 received surveillance, 3 developed new proximal DVT, and none developed PE. Conclusions: About 90% of patients diagnosed with BKDVT received anticoagulation after the initial diagnosis. Surveillance was not commonly recommended and is likely underutilized. Radiology reporting was highly variable and correlated with clinical treatment decisions, whereas other clinical risk factors did not. Describing BKDVT findings only in terms of being "positive" or "negative" for DVT may be inadequate.