Approach to a Patient with Suspected Pulmonary Embolism


Approach to a Patient with Suspected Pulmonary Embolism

Patients presenting with signs and symptoms of an acute PE require urgent evaluation including a thorough history and physical examination that focuses on assessment of pretest probability scores, as well as laboratory studies and advanced imaging.

In many situations, first-line assessment of pretest probability of PE prior to further testing often reduces unnecessary laboratory and imaging studies.  There are several pretest probability clinical scoring systems for PE such as the Wells clinical prediction rule (Table 1) and the revised Geneva score (Table 2) that can be used to assess patients’ pretest probability for PE.  More recently the pulmonary embolism rule-out criteria (PERC) (details below) have been utilized to help avoid additional testing for the patient deemed at low risk. In addition to pretest probability scores, laboratory testing utilizing the D-dimer can help rule out PE and avoid further imaging such as computed tomography pulmonary angiogram (CTPA). We recommend using a hybrid approach that incorporates pretest probability scores (Wells or Revised Geneva), PERC criteria (Table 3) as well as D-dimer as tools for the approach to suspected PE.  The flow chart outlined below follows this approach.  It is important to recognize that clinical judgment in the hands of an experienced clinician expert on PE trumps any form of pretest probability scoring.

Narration of the flowchart:

When assessing a patient with suspected PE, it is first vital to identify signs and symptoms which require immediate attention or hospitalization such as severe chest pain or shortness of breath, hemodynamic instability, hemoptysis, hypotension, tachycardia and syncope or near syncope.  If such alarming features are not present then we recommend proceeding with Wells or revised Geneva pretest probability scoring.

Based on the number of the points further assessment is provided. Wells scores (Table 1) of 0-2 are considered low PE probability (<3.6% risk of PE). Scores of 3-6 points are considered moderate PE probability (<20.5% risk of PE) and scores of 6 points or greater indicate a high probability for PE (up to 66.7% risk of PE).

The Geneva pretest probability model (Table 2) may be preferred by physicians who would like to avoid the subjective assessments present in the Wells criteria which is that PE is more likely than an alternative diagnosis.

It is recommended that patients with a high pretest probability for PE directly undergo pulmonary imaging, primarily computed tomographic pulmonary arteriogram (CTPA). It is recommended that patients with a moderate pretest probability proceed with D-dimer testing.  If the D-dimer is normal then no further testing (unless clinical probability is high) is necessary and PE is ruled out. If the D-dimer is elevated then further testing (CTPA) is necessary.  The D-dimer is considered abnormal in patients < 50 years of age if it is greater than 500 ng/mL by an enzyme-linked immunosorbent assay [ELISA]. In patients > 50 years of age it is considered abnormal if the D-dimer levels are above the patient’s age in years multiplied by 10.  For example the D-dimer would be consider abnormal in a 65 year old if it is > than 650 ng/mL by ELISA.

Lastly, in our approach, patients with a low pretest probability for PE, further evaluation with the PERC rule (Table 3) should be undertaken and if all eight criteria are absent, no further testing is needed.  However, if any of the clinical characteristics listed in the PERC criteria are present then further testing using the D-dimer is recommended similar to the approach used for patients with a moderate pretest probability using the Wells or Geneva scores. (Figure 1)

During the evaluation process it is ideal to check baseline laboratory studies, including renal function, a complete blood count, liver function, and baseline prothrombin/INR and partial thromboplastin times.

Table 1: The Well’s Clinical Prediction Rule includes the following:

Clinical CharacteristicsPoints
Previous DVT or PE1.5
Heart rate >100 minute1.5
Surgery or immobilization  < 4 weeks1.5
Active malignancy1
Clinical signs of DVT3
Alternative diagnosis less likely than PE3

The Wells score obtained relates to the probability of PE:

0-2 points indicates low probability (<3.6%)
3-6 points indicates moderate probability (<20.5%)
6 points or more indicates high probability (66.6%)

Table 2: The Geneva pretest probability model includes the following:

Age 65 years or over1
Previous DVT or PE3
Surgery or fracture within one month2
Active malignant condition2
Unilateral limb pain3
Heart rate 75 to 94 beats per minute3
Heart rate 95 or more beats per minute5
Pain on deep palpation4

The Geneva score obtained relates to the probability of PE:

  • 0 – 3 points indicates low probability (8%)
  • 4 – 10 points indicates intermediate probability (28%)
  • 11 points or more indicates high probability (74%)

Table 3: The PERC criteria are as follows:

Is the patient older than 49 years of age?0+1
Is the pulse oximetry <95% (while on room air)0+1
Is the heart rate above 99 beats per minute?0+1
Does the patient have a prior history of venous thromboembolism?0+1
Has the patient had recent surgery or trauma or hospitalization in the last 4 weeks?0+1
Is there a history of hemoptysis?0+1
Is the patient taking estrogen?0+1
Does the patient have unilateral leg swelling0+1