Objectives To compare the diagnostic accuracy of three pre-probability scores in prediction of pulmonary embolism (PE) admitted to the department of cardiology.

Methods One hundred and seventy five consecutive patients with suspected PE underwent prospective CT pulmonary angiography (CTPA) at the time of initial diagnosis. Three clinical predication scoring systems (Wells’, Geneva’ and revised Geneva’) were used to evaluate the probability of PE in these patients. The predictive accuracy of three scores was compared by area under the curve (AUC) of receiver operating characteristic (ROC) curves.

Results The overall prevalence of PE was 18.9% for the patients with suspected PE in the department of cardiology. The Wells score and Geneva score of the patients with PE was higher than that of the patients without PE (P < 0.05). The revised Geneva score of the patients with PE was the same as that of the patients without PE (P < 0.05). Prevalence of PE in the low, moderate and high pretest probability groups assessed by the Wells score, Geneva score and revised Geneva score was respectively 9.3% 15.8% and 14.8%, 32.3% 23.1% and 24.2%, 100% 100% and 100%. The Wells score performed better than the both Geneva score and revised Geneva score in patients with a low or moderate suspicion of PE (P < 0.05). The AUC for the Wells score, Geneva score and the simplified revised Geneva score was 0.77 ± 0.06, 0.63 ± 0.06 and 0.61 ± 0.05 respectively. The AUC for the Wells score was larger than the other two pre-probability scores (P < 0.05).

Conclusions In department of cardiology, the Wells score appeared to be more accurate than the Geneva score and revised Geneva score and could be used as basic method in screening PE.