Acute pulmonary embolism (APE) is an important clinical problem in patients after major surgery and often remains a difficult diagnosis because of unspecific clinical symptoms. Therefore, we investigated the role of N-terminal prohormone of brain natriuretic peptide (NT-proBNP) for the detection of APE.

METHODS: In 44 patients with suspected APE referred to the intensive care unit after major surgery, serum NT-proBNP, troponin-I, and D-dimers were measured according to the standard hospital protocol. To definitively confirm or exclude APE, all patients underwent an angiographic CT scan of the thorax.

RESULTS: APE was confirmed in 28 and excluded in 16 patients by CT scan. NT-proBNP was significantly (P<0.01) higher in patients with APE [4425 (sd 8826; range 63-35 000) pg ml(-1)] compared with those without [283 (sd 327; range 13-1133) pg ml(-1)]. The sensitivity of the NT-proBNP screening was 93%, specificity 63%, positive predictive value 81%, and negative predictive value 83%. There were no significant (P = 0.96) differences in D-dimers between subjects with and without APE [confirmed APE: 511 (sd 207; range 83-750) μg litre(-1); excluded APE: 509 (sd 170; range 230-750) μg litre(-1)]. Troponin-I levels were not elevated in 32% of the patients with APE.

CONCLUSIONS: D-dimer levels are frequently elevated in post-surgical patients and not applicable for confirmation or exclusion of APE. In contrast, NT-proBNP appears to be a useful biomarker for APE diagnosis in the postoperative setting. In the case of NT-proBNP levels below the upper reference limit, haemodynamically relevant APE is unlikely. Troponin-I in contrast is not considered to be helpful.