RESUMO
BACKGROUND: Pulmonary embolism (PE) is associated with significant morbidity and mortality. PE is a heterogeneous entity that causes a wide variety of clinical presentations, making it imperative to establish which clinical symptoms, signs and biomarkers can influence the pretest probability of PE to aid clinicians and reduce over testing. AIM: To analyse the clinical parameters used by clinicians to order a computed tomography pulmonary angiogram (CTPA) and establish which were associated with the presence of PE. METHODS: Medical records of patients who underwent CTPA from December 2015 to March 2016 were extracted. Patient demographics, clinical symptoms, diagnostic and radiological results were analysed. RESULTS: The study included 150 CTPA studies. Of the studies, 25 were positive for PE and 125 were negative. There was no significant relationship between the presence or character of chest pain and a positive CTPA result (P = 0.216). Previous history of venous thromboembolism (VTE) (P < 0.0001), one or more risk factors for VTE and positive troponin (P < 0.002) were all predictive of PE. None of the patients with a negative D-dimer had a positive CTPA. CONCLUSION: This study supports the negative predictive value of the D-dimer for excluding PE and demonstrates that the strongest pretest predictors of PE in our population are a prior history of VTE, risk factors for VTE and elevated troponin. None of the parameters that often generate requests for CTPA, including vital signs or the presence of chest pain, was associated with the presence of PE in our study population.