• Stefano Aliberti, Alberto Pesci, Sara Lonni, Lucy Poppelwell, Anthony De Soyza, Melissa McDonnell, James Chalmers, Pieter Goemmine, John Davison, Lieven Dupont, and Thomas Fardon.
• Chest. 2014 Mar 1;145(3 Suppl):428A.

Session TitleBronchiectasisSESSION TYPE: Slide PresentationsPRESENTED ON: Saturday, March 22, 2014 at 09:00 AM - 10:00 AMPURPOSE: Non-cystic fibrosis bronchiectasis is a multicomponent disease associated with increased morbidity and mortality. Using data from 608 patients attending a specialist bronchiectasis clinic in Edinburgh, UK, the Bronchiectasis Severity Index (BSI) has recently been developed as a composite clinical prediction tool in bronchiectasis. The score consists of 8 commonly measured clinical parameters reflecting age, BMI, MRC dyspnoea score, FEV1%, bacterial colonisation, radiological extent, exacerbation frequency and prior hospitalization with subsequent classification into low, intermediate and high risk groups.MethodsProspective observational studies of patients attending each of the four centres was performed. The utility of the BSI to predict mortality, hospital admission rates and exacerbation frequency across the three risk groups were determined in each cohort.ResultsThe validation cohorts totalled 702 patients; median age 65 years (interquartile range 56-74); 46% male. All-cause mortality was 2.3%, 13.5%, 16.6% and 1.9% in Scotland (n=251), England (n=126), Belgium (n=253) and Italy (n=105), respectively. The mortality and hospitalization rates increased with increasing BSI scores across all cohorts. The area under the receiver operator characteristic curve (AUC) for BSI-predicted mortality was 0.84 (Scotland), 0.82 (England) and 0.81 (Belgium). No AUC was calculated for Italy due to their minimal death rate. The AUC for BSI-predicted hospitalization rates was 0.82 (Scotland), 0.80 (England) and 0.88 (Italy) with no data for the Belgian cohort. The BSI predicted exacerbation frequency during follow-up in all cohorts where data was available (Scotland and England, p<0.001, Italy p=0.03). Meta-analysis data comparing results of derivation and validation cohorts (total 1315 patients) demonstrated that the score works consistently with low heterogeneity and high discriminatory ability.ConclusionsThe BSI is a reliable, valid, reproducible clinical prediction tool for calculating the probability of mortality, hospital admissions and exacerbation frequency in bronchiectasis.Clinical ImplicationsThe BSI score provides a much needed clinically useful outcome measure in bronchiectasis that may help risk-stratify patients and assist personalised medicines approaches.DisclosureThe following authors have nothing to disclose: Melissa McDonnell, James Chalmers, Pieter Goemmine, Stefano Aliberti, Sara Lonni, John Davison, Lucy Poppelwell, Walid Salih, Alberto Pesci, Lieven Dupont, Thomas Fardon, Adam Hill, Anthony De SoyzaNo Product/Research Disclosure Information.

19 keywords...

hide…