Ms. Sarah Bellum (if you're not smiling, try saying the name out loud or the joke is just lame …) is a 32-year-old Caucasian female who presents to your ED with shortness of breath. She just returned home after the International Conference on Coordination in London. Immediately after walking through her front door, she became acutely short of breath (not attributable to the Justin Bieber poster in her living room). This is associated with some moderately sharp chest pain located along the left side of her chest. The pain seems worse when she attempts to breathe deeply.
Question 3.5.1 Which important question(s) do you next ask Ms. Bellum?
A) Do you smoke cigarettes?
B) When was your last menstrual period?
C) Have you had surgery recently?
D) Do you have a history of kidney disease?
Answer 3.5.1 The correct answer is "E." Each of these questions addresses risk factors associated with pulmonary embolism (PE) and/or deep vein thrombosis (DVT). Smoking cigarettes and recent surgery are strong risk factors, as is an active pregnancy. Addressing a patient's menstrual cycle serves as a natural segue to a discussion about the use of oral contraceptives, which, too, is a prominent risk factor. As for renal disease, nephrotic syndrome has been associated with an increased risk of PE.
After further verbal probing, you discover that Ms. Bellum recently completed her menstrual cycle and the other presented questions turned up no risk factors. However, your smooth segue did reveal that she takes low-dose estrogen for birth control. During this discourse, you also learn that her aunt had a blood clot in her leg once. She has no further details but does not think that her aunt had any further complications from this condition. Regardless, PE just took a violent leap to the top of your differential. You glance at her vitals (temperature 37.1°C, heart rate 92 bpm, blood pressure 129/68 mm Hg, respiratory rate 21, SpO2 95% on room air) and notice that she appears mildly uncomfortable but not in any acute distress. Her physical examination is entirely unremarkable. You order an ECG on the patient to evaluate for potential cardiac etiologies for her symptoms.
Question 3.5.2 Assuming Ms. Bellum does have a PE, what is her ECG most likely to show?
B) Nonspecific ST-T wave changes.
E) Multifocal atrial tachycardia.
Answer 3.5.2 The correct answer is "D." The most common ECG finding associated with the diagnosis of PE remains normal sinus rhythm. With that said, the most common arrhythmia found in patients with a PE is sinus tachycardia. But alas, Sarah had a normal heart rate. The other choices can certainly be found with this condition but are far less frequent. Of note, the "textbook" S1Q3T3 ECG rarely occurs and historically traces back to a handful of patients in the 1930s that had massive pulmonary emboli. Even if you do spot this pattern on an ECG, it is not specific enough to confirm the diagnosis. In the end, the clinical signs attributed to pulmonary emboli (such as the shortness of breath and chest pain) are frequently more valuable than any abnormal ECG finding.
As you attempt to rule out other potential etiologies for the patient's symptoms (e.g., pneumonia, atelectasis, and pneumothorax), you order a trusty chest radiograph.
Question 3.5.3 What is the most common radiographic finding in a patient with a PE?
B) No acute cardiopulmonary processes.
D) Hilar/mediastinal enlargement.
Answer 3.5.3 The correct answer is "B." Admittedly, this one is a bit tricky. Approximately 75% of the chest radiographs in the setting of PE are abnormal. However, there are numerous causes for these abnormalities and none of them individually surpass the frequency of the normal chest radiographs. Specifically, the "textbook" findings of Westermark sign (loss of peripheral vascular markings) and Hampton hump (a wedge-shaped opacity due to pulmonary infarction) are infrequent, and both have a low sensitivity and low specificity. In short, all the other options can be seen as the result of a PE but none are more frequent than a normal chest radiograph.
ECG and chest radiograph in hand, you turn your attention toward ordering the appropriate laboratory tests to solidify your presumptive diagnosis. You are working with a medical student who suggests a number of lab tests. You agree with most of them, but shoot him down on one.
Question 3.5.4 Which test should you AVOID ordering?
D) Basic metabolic panel (Na+, K+, Cl–, CO2–, BUN, Cr–, and glucose).
Answer 3.5.4 The correct answer is "B." The D-dimer can be a blessing for some but is the bane of existence for others. In this patient, a D-dimer is not useful. This test has great sensitivity but poor specificity. It is positive in far more conditions than PE. Used as a "rule-out" test for PE, it only applies in low-risk patients. Ms. Bellum is not a low-risk patient as suggested by the pulmonary embolism rule-out criteria (PERC) rules (see Helpful Tip) due to her use of exogenous estrogen. The Wells criteria for PE place her in the moderate-risk group (16.2% risk of PE). As such, even a negative D-dimer is insufficient for ruling out the diagnosis. As for the other tests, they all serve a valuable role in her evaluation. For instance, the CBC could provide evidence of anemia, while the PT/PTT may reveal a coagulopathy. Assessing her renal function may be needed for her evaluation moving forward, and the same can be said for verifying her nongestational status. Plus, a urine pregnancy test is performed on almost every woman in an ED. It might as well be part of the triage process.
That pesky med student seemed to know a lot about the PERC rules and Wells criteria. But when he listed the Wells criteria, he got one wrong.
Question 3.5.5 The Wells criteria for PE include all of the following EXCEPT:
C) Previous history of venous thromboembolism.
D) Clinical symptoms and signs consistent with PE.
Answer 3.5.5 The correct answer is "A." While an important risk factor for PE, estrogen use is not included in the Wells criteria. All the others count in the Wells criteria (Table 3-3). Using a medical calculator website, such as www.medcalc.com, is most helpful.
TABLE 3-3WELL'S SCORE FOR PEa ||Download (.pdf) TABLE 3-3 WELL'S SCORE FOR PEa
|Clinically suspected DVT ||3 points |
|No alternative diagnosis more likely than PE ||3 points |
|Tachycardia >100 ||1.5 points |
|Immobilization for ≥3 days or surgery in the previous 4 weeks ||1.5 points |
|History of DVT or PE ||1.5 points |
|Malignancy within 6 months ||1 point |
|Presence of hemoptysis ||1 point |
The PERC rules are a validated set of rules that allow categorization of a patient into a low-risk group to rule out PE clinically. If the patient meets all of the following, PE is ruled out (assuming you believe the patient is low risk):
Heart rate <100.
No unilateral leg swelling.
No recent history of trauma or surgery.
No prior DVT or PE.
No hormone use.
Remember not to get a D-dimer on no-risk patients. This simply increases the CT rate and exposure to unnecessary radiation. The fatal cancer rate in a 20-year-old female undergoing a 64-slice chest CT is 1:142 (really) (JAMA. 2007;298:317–323)! See also Lancet. 2009;374:1160 and J Pediatr. 2009;154:912 among others. Plus, 25% of "positive" CTs in low-risk patients are false positive, damning them to the hell of anticoagulation (Am J Roentgenol. 2015;205:271).
The CBC, coagulation studies, and basic metabolic panel all return within normal limits. In addition, Ms. Bellum is not pregnant. Thus, you wish to (finally) solidify that diagnosis you have suspected for quite some time.
Question 3.5.6 Since you do not put her in the low-risk category by your clinical judgment, what diagnostic study should you order?
A) VQ (ventilation–perfusion) scan.
B) CT scan of the chest without contrast.
C) CT scan of the chest with contrast.
E) Compressive Dopplers of the lower extremities.
Answer 3.5.6 The correct answer is "C." The American College of Radiology (ACR) lists the CT scan of the chest with contrast (i.e., CT angiography or CTA) as the modality of choice in stable patients with a suspected PE. The CTA is considered to be the standard of care. Its benefits include the fact that it is noninvasive, cheaper than pulmonary angiography, and far more available than VQ scans. It should be noted that pulmonary angiography still remains the "gold standard" for diagnosing pulmonary emboli, but that is more of an academic point. As for VQ scans, they are not available in many locales and often return nondiagnostic. However, they can be used in a patient with a normal chest x-ray. A chest CT without contrast ("B") will not enhance the pulmonary arteries, making the diagnosis of a PE far more difficult, if not impossible. Since most pulmonary emboli are believed to arise from the lower extremity venous system, "E," Dopplers of the legs, could be considered if the patient was a poor candidate for both CTA and VQ scan (e.g., COPD and Stage 4 chronic kidney disease). But this approach is obviously not diagnostic of PE; it would just help you determine if the patient has an active thrombosis, and you would manage the same whether she has a PE or DVT or both.
Some practitioners will diagnose a PE based on CTA then obtain Dopplers of the legs. Why do both? If you find a DVT and a PE, you would treat it the same as if you only found one or the other. After you have diagnosed a PE, you can be done with imaging. Same with DVT.
In pregnancy, both chest CT and V/Q scanning are acceptable imaging options and involve acceptable radiation levels (Obstet Gynecol. 2011;118:718).
As keenly suspected, Ms. Bellum's CTA of the chest reveals a moderate-sized pulmonary embolus in the left pulmonary artery. Her vital signs are still stable and her pain is well-controlled with intravenous morphine. She is surprised by the diagnosis you give her but appears to be taking it in stride.
Question 3.5.7 What is the optimal management plan for the patient moving forward?
A) Bolus her with unfractionated heparin (UFH), start her on oral warfarin, and discharge her to home.
B) Administer a dose of enoxaparin in the ED, provide education, and discharge the patient to home with primary care follow-up in the next 2 to 3 days.
C) Start the patient on low–molecular-weight heparin (LMWH), initiate oral warfarin therapy, and admit the patient to the family medicine service.
D) Bolus her with UFH, initiate a UFH drip, and admit the patient to the family medicine service.
E) Start her on oral warfarin and discharge her to home with primary care follow-up in the next 2 to 3 days.
Answer 3.5.7 The correct answer is "C." Full anticoagulation is considered mandatory for all patients with a PE. While discharging a patient with only a DVT is considered standard of care, this is not (yet) true for PE. Thus, any plan that centers on a discharge to home is incorrect (a 2014 Cochrane Database review concluded that the available evidence is insufficient to recommend outpatient therapy and that more studies are needed). With regard to selecting an anticoagulant, current evidence does not support the use of one agent over another; UFH, LMWH, and fondaparinux are all appropriate. Regardless of the selected anticoagulant, the current recommendations stipulate the initiation of warfarin at the time of diagnosis as well. The UFH, LMWH, or fondaparinux should be continued until the patient's INR has been therapeutic (2–3) for at least 24 hours and at least 5 days. Note that apixaban, dabigatran, and rivaroxaban can all be used to treat and prevent PE. Like warfarin, dabigatran requires 5 days of overlap with LMWH when used as treatment. Avoid edoxaban (Savaysa) if possible. It can only be used in those with a CrCl of less than 95 mL/min. And, it requires dose adjustment for those with a CrCl of between 50 and 15 mL/min (do not use under CrCl 15 mL/min). Note that the most recent available guideline from the American College of Chest Physicians (ACCP), published online in January 2016, recommends novel oral anticoagulants (e.g., dabigatran, apixaban, and others) over warfarin for treatment of PE in patients without cancer; however, the ACCP grades this recommendation as "weak with moderate quality evidence." Continue to look at the whole patient when making anticoagulation decisions.
Question 3.5.8 How long are you going to maintain this patient on anticoagulation?
Answer 3.5.8 The correct answer is "A." For a PE that has a reversible cause (oral contraceptive pills in this patient with a long airplane trip), 3 months of anticoagulation is adequate. For those with a second PE, lifetime anticoagulation is warranted. For those with a cryptogenic PE or PE from an acquired or inherited thrombophilia (e.g., Factor V Leiden), recommendations are all over the place from 3 months to life. Nine months is likely adequate for a patient with a PE from an irreversible cause, although patients go back to their pre-treatment risk as soon as you stop anticoagulation. And, a first PE trumps everything else in terms for risk factors for a second PE—including all those fancy thrombophilia tests! So, finding that thrombophilia does not necessarily help your decision-making process. Note that some guidelines suggest anticoagulation for life after a first cryptogenic/unprovoked PE if benefit seems to outweigh risks. This should be decided after 3 months of anticoagulation. (Chest. 2012;141(2)(Suppl):7S–47S.)
An elevated A–a gradient suggests a ventilation/perfusion mismatch and occurs in a number of conditions, including atelectasis, right-to-left shunt, acute respiratory distress syndrome (ARDS), air embolism, resolving severe asthma, COPD with oxygen treatment, and bronchiectasis with impaired gas exchange. Thus, an elevated A–a gradient is not specific for PE. Likewise, a normal A–a gradient and normal oxygen saturation do not rule out PE! In fact, in patients without underlying lung disease, the PIOPED study found no difference in the oxygen saturation and A–a gradient among patients with and without PE.
Only 88% of patients with a PE are hypoxic, 70% have dyspnea or tachypnea, 65% have pleuritic pain, and as few as 30% are tachycardic. The point here is to have a high clinical suspicion in the right situation despite the lack of the classic triad. To make things worse, the troponin and BNP can be elevated in patients with a PE.
Compression stockings may prevent postthrombotic syndrome, the recurrent swelling and edema often found after a DVT—or it may not! Evidence is conflicting. Consider prescribing compression stockings routinely in these patients. Thrombolysis of proximal DVTs will also prevent postthrombotic syndrome with a number needed to treat (NNT) of 7 for proximal DVTs. This comes at the number needed to harm (NNH) of 1 in 22 patients requiring a blood transfusion, NNH of 1 in 11 developing a PE, and NNH of 1 in 5 requiring a vena cava filter. (JAMA Intern Med. 2014;174(9):1494–1501.)
Your patient does well, completes her course of warfarin, and has no further episodes over the next 2 years. She develops gallstones and plans to have an elective laparoscopic cholecystectomy. A surgeon colleague sends her back to see you for a preoperative evaluation. You find no evidence of cardiac, pulmonary, or hematologic disease. She is no longer on warfarin and is doing well.
Question 3.5.9 Which of the following postoperative management strategies do you recommend?
A) Aspirin 81 mg PO daily.
B) Warfarin 5 mg PO daily.
C) UFH 5,000 units subcutaneously daily.
D) Enoxaparin 40 mg subcutaneously daily.
E) No antiplatelet or anticoagulant drugs.
Answer 3.5.9 The correct answer is "D." Even for a relatively minor surgical procedure where anesthesia is used for 30 minutes or less and the postoperative recovery is usually quick, your patient is at moderate risk for venous thromboembolism. Her history of PE puts her in a higher risk category, and she requires prophylaxis. Of the choices available, enoxaparin would be the most appropriate. LMWH and UFH are both acceptable for prevention of DVT/PE in the postoperative period, but "C" is wrong because UFH must be dosed every 8 to 12 hours rather than daily. "A" is incorrect. Aspirin is sometimes used postoperatively, but the dose should be 160 mg/day or greater. Also, compared with heparin and its derivatives, aspirin is less efficacious in the prevention of thrombus. "B" is incorrect. Warfarin alone is not appropriate in this setting due to its slow onset of action.
The optimal length of time that patients require prophylaxis for venous thromboembolism after surgery is unknown. Arguments can be made for prophylaxis until the patient is ambulating several hundred feet per day.
Objectives: Did you learn to…
Recognize risk factors for a PE?
Understand the variability of symptoms and signs in PE?
Appreciate the PERC rules and Wells criteria and how they can be used to rule out a PE?
Implement treatment and prevention for PE?
Understand the A–a gradient?