Optimization of the diagnostic management of clinically suspected pulmonary embolism in hospitalized patients

Identical diagnostic algorithms for suspected pulmonary embolism (PE) are used for hospitalized patients and outpatients, while D-dimer levels, risk factors and pre-test probability for PE differ, and the percentage of patients managed without computerized tomography pulmonary angiography (CTPA) is lower in hospitalized patients. We aimed to improve the efficiency of the diagnostic algorithm by increasing the threshold of the D-dimer, the threshold of the Wells rule and by adjustments of the Wells rule. Six-hundred and twenty-four hospitalized patients from two previously performed management studies with a PE prevalence of 26% were studied. Adjustments were considered to be safe when the failure rate remained <2%. By applying standard management, 8% (49/624) were managed without CTPA with a failure rate of 0·0% (0/49; 95% confidence interval [CI] 0·0-7·3), and it was 1·7% (8/465; 95%CI 0·8-3·4) for all patients in whom PE was excluded at baseline. All evaluated adjustments resulted in an increase of the failure rate with very small improvements of the efficiency. Given these potentially small improvements and the increasing complexity of clinical practice if adjusted diagnostic algorithms for specific patient categories were introduced, we do not recommend further evaluation of any of the adjustments; we recommend that the standard diagnostic algorithm should continue to be applied.