Pulmonary embolism: arterial blood gases were unhelpful.

Clinical bottom line (level 1b)

  1. Arterial blood gases were unhelpful in diagnosing or excluding pulmonary embolism.
Stein et al: Chest 1996; 109 (1): 78-81
Expires September 2003

The study

Setting: tertiary medical centre, USA

768 patients (aged ?, ?% male) suspected pulmonary embolism

Excluded if
  • aged <19 years old
  • no information available on prior cardiopulmonary disease, defined as: history or evidence of valvular heart disease coronary artery disease, myocardial infarction, "other heart disease", prior left or right heart failure; asthma, COPD, interstitial lung disease, "other lung disease", concurrent pneumonia, ARDS, prior PE
Independent blinded reference standard, applied in all patients from a consecutive appropriate spectrum.
Reference standard:
  • pulmonary angiogram
Diagnostic test: arterial blood gases done on room air

The evidence

pre-test probability of pulmonary embolism: 36%, (95% CI: 33% to 40%)
pre-test probability of PE in patients with no prior cardiorespiratory disease: 39%, (95% CI: 34% to 45%)
pre-test probability of PE in patients with prior cardiorespiratory disease: 34%, (95% CI: 29% to 38%)

diagnostic test pulmonary embolism (overall) no pulmonary embolism LR+
(95% CI)		 post-test probability LR-
(95% CI)		 post-test probability
pO2 80 mmHg or more (10.2 kPa) 223 374 1.1
(0.98 to 1.1) 		37% 0.82
(0.61 to 1.1) 		32%
pCO2 35 mmHg or more (4.5 kPa) 142 244 1.0
(0.89 to 1.2) 		37% 0.97
(0.83 to 1.1) 		35%
p (A-a) O2 20 mmHg or less (2.6 kPa) 248 421 1.0
(0.99 to 1.1) 		37% 0.73
(0.49 to 1.1) 		29%
total 277 491


diagnostic test pulmonary embolism (pts without prior disease) no pulmonary embolism LR+
(95% CI)		 post-test probability LR-
(95% CI)		 post-test probability
pO2 80 mmHg or more 98 145 1.0
(0.91 to 1.2) 		40% 0.90
(0.61 to 1.3) 		37%
pCO2 35 mmHg or more 64 95 1.0
(0.83 to 1.3) 		40% 0.97
(0.78 to 1.2) 		39%
p(A-a) O2 20 mmHg or less 112 160 1.1
(0.98 to 1.2) 		41% 0.69
(0.42 to 1.2) 		31%
total 130 200


diagnostic test pulmonary embolism (pts with prior disease) no pulmonary embolism LR+
(95% CI)		 post-test probability LR-
(95% CI)		 post-test probability
pO2 80 mmHg or more 125 229 1.1
(0.99 to 1.2) 		35% 0.70
(0.45 to 1.1) 		26%
pCO2 35 mmHg or more 78 149 1.0
(0.86 to 1.3) 		34% 0.96
(0.78 to 1.2) 		33%
p(A-a) O2 20 mmHg or less 135 253 1.1
(0.99 to 1.1) 		35% 0.63
(0.34 to 1.2) 		24%
total 147 291

  • p(A-a) O2= 150-1.25 pCO2-pO2

Comments

  1. ABG may be helpful in the overall assessment and management decisions of a dyspneic patient, but will not help rule in or out a PE.

Citation

  1. Stein PD, Goldhaber SZ, Henry JW, et al: Arterial blood gas analysis in the assessment of suspected acute pulmonary embolism. Chest 1996; 109 (1): 78-81
Contributor: Chris Ball and Clare Wotton, September 2000
Reviewer: Michael Christian

Clinical Question.
Patient suspected pulmonary embolism
Intervention or Exposure arterial blood gases
Outcome diagnosis