Background BAE for massive haemoptysis is potentially life-saving with low short-to-medium term failure rates in previous studies. We aimed to characterise patients referred for BAE, to examine long-term treatment success and identify risk factors for requiring repeat BAE.

Methods We retrospectively identified all patients undergoing BAE from 1994 to 2007. We collated data from hospital databases and primary care on demographics, respiratory diagnoses and procedure with follow-up of up to 16 years. Outcomes were all-cause mortality and recurrence of haemoptysis requiring repeat BAE.

Results 158 patients were embolised on 208 occasions. 85 (54%) patients were male and median age was 54 (IQR: 41–67) y. The most common underlying diagnoses were aspergilloma (n=38; 24% of patients), bronchiectasis (n=24; 15%), unidentified cause (n=17; 11%) chronic tuberculosis (n=14; 9%), active tuberculosis (n=12; 8%) and cystic fibrosis (n=11; 7%). All-cause mortality at 1 month and 3 years was 5.3% and 29.7%, and need for repeat BAE was 4.7% and 30.7% respectively. Repeat BAE at 3 years was most common with aspergilloma (50%) and least common with active TB (0%). 3-year mortality was highest in cystic fibrosis (40%) and least with unknown cause (7.7%) Neither number nor location of vessels embolised predicted mortality or need for repeat BAE. No major procedural complications were noted.

Conclusions BAE by experienced operators is a safe, minimally invasive procedure for massive haemoptysis with excellent short-term success. It does not prevent late recurrence of haemoptysis nor obviate the need for repeat BAE, the risk of which is related to the underlying disease rather than to technical aspects of the procedure.