Clinical Decision Rules in the Diagnostic Work-Up of Patients with Clinically Suspected Pulmonary Embolism

Abstract

Pulmonary embolism (PE) is challenging to diagnose because of the low sensitivity and specificity of individual clinical signs and symptoms. Increasing emphasis is put on excluding the diagnosis in a safe, efficient and non-invasive manner by determining the pre-test probability of PE. Clinical variables of patients with suspected PE can be combined, either by implicit judgement or by explicit, more standardized clinical decision rules. An ‘unlikely’ probability according to the Wells rule or a ‘low’ clinical probability according to the Geneva rule, in combination with a normal d-dimer test, results in a substantial reduction of the necessity for further diagnostic testing in patients with suspected PE. Interobserver variability is similar for the different clinical decision rules, but is higher for implicit judgement. Comparative studies of the different methods show similar accuracy in the determination of pre-test probability. Using a standardized clinical decision score has advantages over implicit judgement, as even less experienced physicians can adequately assess clinical probability. Modified scores have recently been introduced with clear advantages over the more customary score, but these need further validation.