Pulmonary embolism: patients with a normal d-dimer and considered at low risk for a PE using a clinical diagnosis rule were unlikely to have one

Clinical bottom line (level 1a)

  1. A tenth of patients attending an emergency department with a suspected PE has one.
  2. A clinical diagnosis rule based on history and examination can rank patients at low, medium or high risk for a pulmonary embolism.
  3. Patients ranked at low-risk and with a normal d-dimer are unlikely to have a PE (LR - 0.023) .
  4. A diagnostic protocol for PE missed 0.6% of PEs.
Wells et al: Ann Intern Med 2001; 135 : 98-107
Expires May 2004

The study

Setting: emergency departments, 4 acute hospitals, Canada

930 patients (aged aged 16 to 93; mean 51, 63% male) with suspected pulmonary embolism with new or worsening dyspnoea or chest pain

Excluded if
  • aged < 18
  • pregnancy
  • geographical inaccessibility preventing follow-up
  • suspected upper extremity DVT as source of PE
  • no symptoms of PE within 3 days of presentation
  • anticoagulation for > 24 hours
  • expected survival < 3 months
  • contraindication to contrast media
  • symptoms for 30 days or more

Patients with a negative d-dimer consider at low risk for a PE on the clinical prediction guide were followed for 3 months and received no diagnostic imaging. All other patients underwent ventilation-perfusion scanning. Patients with non-diagnostic scans then had bilateral leg-vein ultrasonography. Patients with no evidence of DVT at low risk for a PE, or moderate risk with a normal d-dimer were diagnosed with no PE. Patients with no evidence of DVT at moderate risk for a PE with an abnormal d-dimer or patients at high-risk with a normal d-dimer had follow-up ultrasonography in a week. If this was positive they were treated for a PE. Patients at high risk for a PE with a positive d-dimer underwent pulmonary angiography. Patients considered to be negative for PE had anticoagulation withheld.
Independent blinded reference standard, applied in all patients from a consecutive appropriate spectrum.
Reference standard:
  • PE diagnosed if high probability on ventilation-perfusion scanning. If nondiagnostic, abnormal lower extremity ultrasonography abnormal angiography or venous thromboembolic event during 3 months of follow-up.
Diagnostic test: Validation of a clinical diagnosis rule and d-dimer
  • Clinical prediction guide. Sum the following components if present
    • clinical signs and symptoms of deep vein thrombosis (objectively measured leg swelling and pain with palpation in the deep-vein region): 3.0
    • heart rate > 100 beats/min: 1.5
    • immobilisation (bedrest, except access to bathroom, for 3 or more days; or surgery in previous 4 weeks: 1.5
    • haemoptysis: 1.0 points
    • previously objectively diagnosed deep vein thrombosis or pulmonary embolism: 1.0
    • malignancy (patients with cancer receiving treatment, or treatment stopped within previous 6 months, or receiving palliative care): 1.0
    • pulmonary embolism as likely or more likely than an alternative diagnosis (based on clinical information, chest X-ray, ECG and any blood tests required to diagnose PE): 3.0

The evidence

pre-test probability of pulmonary embolism: 9.5%, (95% CI: 7.5% to 11.3%)

diagnostic test pulmonary embolism no PE LR
(95% CI)		 post-test probability
high risk: score > 6.0 24 40 5.9
(3.7 to 9.3) 		38%
moderate risk: score 2.0 to 6.0 55 284 1.9
(1.6 to 2.3) 		16%
low risk: score < 2.0 7 520 0.13
(0.065 to 0.27) 		1%
total 86 844


diagnostic test pulmonary embolism no PE LR+
(95% CI)		 post-test probability LR-
(95% CI)		 post-test probability
diagnosis other than low risk and d-dimer negative 85 408 2.0
(1.9 to 2.2) 		17% 0.023
(0.0032 to 0.16) 		0.2%
total 86 844

Comments

  1. 5 patients developed a PE or DVT who were diagnosed as negative for PE using the above protocol (0.6%: 95% CI: 0.2% to 1.4%). However 4 did not undergo the full diagnostic work-up.
  2. 47% of patients received no imaging, 7% underwent serial ultrasonography and 1.1% had angiography using the diagnostic protocol.

Citation

  1. Wells PS, Anderson DR, Rogers M, et al: excluding pulmonary embolism at the bedside without diagnostic imaging: management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and d-dimer. Ann Intern Med 2001; 135 : 98-107
Search Terms: from ACP Journal Club
Contributor: Chris Ball, May 2002
Reviewer:

Clinical Question.
Patient suspected pulmonary embolism
Intervention or Exposure clinical prediction rule and d-dimer
Outcome pulmonary embolism