Radiology: when to test
Symptoms of VTE are nonspecific, and only about 20% of patients suspected of having VTE ultimately have the diagnosis confirmed.2 Thus, objective diagnosis with radiological investigation is always required. Unfortunately, in the primary care setting immediate radiological investigation is not always feasible. In this setting, brief courses (less than 24 hours) of presumptive anticoagulation are appropriate until radiological investigation can be arranged.
Radiological studies are associated with significant cost and inconvenience to patients, especially those living in rural areas (Box). Imaging can be avoided in patients with a negative d-dimer test result and low risk scores as calculated by validated clinical prediction tools (Table 1). The tools with the most evidence to support their use are the Geneva and Wells scores.3 The pulmonary embolism rule out criteria (PERC) are useful in the exclusion of PE; however, they should only be used in young patients (less than 50 years of age), in whom the likely background rate of VTE is about 15% or less.4,5
Radiology: choice of test
The investigation of choice for suspected DVT is venous duplex ultrasound. In Australia, to investigate lower-limb DVT, the entire deep venous system of the leg from groin to ankle is imaged (Figure). A negative study result effectively excludes the diagnosis of lower-limb DVT.6 A two-point compression ultrasound may be helpful to exclude a proximal thrombosis but does not detect distal thromboses, and should always be followed up with serial ultrasound. A second ultrasound within one week from the original imaging is sufficient.
In most patients, CT pulmonary angiography (CTPA) is recommended as the investigation of choice for suspected PE.7 Its worth should be weighed against the significant radiation and nephrotoxic dye exposure. Ventilation-perfusion (VQ) scanning is preferred in pregnant patients and those with renal failure; however, this test is associated with a significant number of indeterminate results. In the setting of an indeterminate VQ finding, the next investigation is determined by the risk-benefit ratio of CTPA. If the risks of CTPA are too high, serial duplex ultrasound of the lower limbs is an effective alternative. The absence of DVT makes embolic events unlikely, and therefore anticoagulation can usually be safely withheld.7
When to treat
Most patients with VTE require anticoagulation therapy. Apart from the rare patient with a contraindication due to bleeding, all proximal DVTs and PEs should be treated. Less clear is the management of isolated distal DVT – thromboses confined to the vessels below the popliteal vein, without associated PE. Symptomatic distal DVT should receive anticoagulation, but the duration may be limited to six to 12 weeks. Those with increased risk for bleeding may have anticoagulation deferred but should be monitored with two duplex ultrasounds over two weeks.8,9 Any thrombosis that progresses during this time should be treated.8,10 Subsegmental PE may be managed with a similar observational protocol in low-risk asymptomatic patients; the THANZ recommend serial bilateral ultrasound examination of the legs at diagnosis and after one week.11
At the other end of the scale, haemodynamic instability resulting from massive PE is associated with high rates of mortality. These patients should be urgently transferred via ambulance to a centre with thrombolysis capabilities.12,13
The benefit of treatment with anticoagulation must be weighed against the risk of major bleeding. The most significant risk for bleeding in patients on anticoagulation therapy is a previous history of bleeding. Anticoagulation should be avoided in most patients with active bleeding or unremedied blood loss over the preceding 30 days.14 Other predictors of bleeding include potential bleeding lesions, recent surgery (within 14 days), severe kidney disease and active cancer. It should be noted that the latter three criteria listed are also risk factors for thrombosis, making the decision to anticoagulate challenging, and often subjective, in these patients. Nevertheless, despite the high level of concern among patients and clinicians, rates of bleeding in patients without risk factors are reassuringly low, in the order of 0.8 to 1.6% per year. Bleeding rates in patients receiving low-dose NOACs were similar to age-matched controls.8,15