BACKGROUND: Haemoptysis is a common symptom and can be an early sign of lung cancer. Careful investigation of patients with haemoptysis may lead to early diagnosis. The strategy for investigation of these patients, however, is still being debated.

OBJECTIVES: We studied whether the combination of computed tomography (CT) and bronchoscopy had a higher sensitivity for malignant and non-malignant causes of haemoptysis than CT alone.

METHODS: The study was a retrospective, non-randomised, two-centre study and included patients who were referred from primary care for the investigation of haemoptysis.

RESULTS: A total of 326 patients were included in the study (mean age 60.5 [SD 15.3] years, 63.3% male). The most common aetiologies of haemoptysis were cryptogenic (52.5%), pneumonia (16.3%), emphysema (8.0%), bronchiectasis (5.8%) and lung cancer (4.0%). In patients diagnosed with lung cancer, bronchoscopy, CT and the combination of bronchoscopy and CT had a sensitivity of 0.61, 0.92 (p<0.05) and 0.97 (p=0.58), respectively. In patients with non-malignant causes of haemoptysis, most aetiologies were diagnosed by CT and comprised mainly pneumonia, emphysema and bronchiectasis. Bronchoscopy did not reveal these conditions and the sensitivity to these conditions was not increased by combining CT and bronchoscopy.

CONCLUSIONS: CT can stand alone as a diagnostic workup for patients with haemoptysis referred to an outpatient clinic. Bronchoscopy does not identify any malignant aetiologies not already diagnosed by CT. Combining the two test modalities does not result in a significant increase in sensitivity for malignant or non-malignant causes of haemoptysis.