We know that the d-dimer can be a helpful test for patients who have a low pre-test probability of pulmonary embolism. But can the test be pushed into use for higher risk patients? Will it still have useful negative predictive value or will we risk missing too many PEs?