A 31-year-old woman comes to the emergency department due to shortness of breath and anxiety. She has no associated chest pain or palpitations. She recently returned from a trip to Thailand. The patient does not smoke and drinks alcohol occasionally. She is married and uses barrier contraception. The patient has a history of anxiety and sleep problems, and her family history is negative for heart attack or blood clots but positive for colon cancer (her uncle at age 58). She is afebrile. Blood pressure is 110/70 mm Hg, pulse is 96/min, and respirations are 24/min. Estimated jugular venous pressure is normal. The chest is clear to auscultation. There is no extremity edema and no skin rashes are present. Arterial blood gas analysis shows:
| pH | 7.43 |
| PaCO2 | 35 mm Hg |
| PaO2 | 90 mm Hg |
Chest x-ray and D-dimer are normal. Her symptoms resolve over the next hour. Which of the following is the most appropriate next step in management of this patient?
Show Explanatory Sources
This patient has self-resolving dyspnea with no chest pain or fever and unremarkable laboratory and radiologic studies, possibly due to underlying anxiety. Her normal chest x-ray helps exclude pulmonary edema and pneumonia. Pulmonary embolism (PE) is on the differential. Due to the variability in presentation of PE, the modified Wells criteria can help assess its pretest possibility. Long-distance flights, although a risk factor for venous thromboembolism (VTE), do not qualify as immobilization per Wells criteria, and therefore this patient's modified Wells score is 0. This result gives her an unlikely pretest probability for PE, and her normal D-dimer (fibrin degradation product) result effectively rules out PE.
D-dimer has 95%-97% sensitivity and 45% specificity for VTE. The positive predictive value (PPV) and negative predictive value (NPV) vary depending on pretest probability (eg, disease prevalence, individual pretest probability). In patients with unlikely probability of VTE, D-dimer has a NPV >95%, so a normal result (<500 ng/mL) reliably excludes VTE. In patients with likely probability of VTE, the prevalence of VTE is higher, lowering the NPV and rendering a normal D-dimer result unreliable in excluding VTE (lower NPV = more false negatives). In such patients, a CT angiogram or a ventilation/perfusion scan (rather than D-dimer) should be used to reliably exclude PE.
Regardless of pretest probability, an elevated D-dimer has poor PPV and is not diagnostic of VTE.
(Choices A, B, C, and D) A normal D-dimer result reliably rules out VTE in this patient with unlikely pretest probability. Additional testing with CT angiogram, compression ultrasound of the legs, or echocardiography (which might show suggestive right ventricular changes in the presence of large PE) is not necessary, would increase cost, and, in the case of CT scan, would provide unnecessary radiation exposure. Initiation of anticoagulation is rarely appropriate in clinically stable patients with unlikely pretest probability of VTE until the diagnosis has been confirmed.
Educational objective:
A normal D-dimer result is useful in excluding pulmonary venous thromboembolism (VTE) in patients with unlikely pretest probability. However, a normal D-dimer result is not useful in patients with likely pretest probability. An elevated D-dimer result in any patient is not diagnostic of VTE and must be followed by more specific studies.