Introduction and objectives NICE Guidance encourages further research on the efficacy of bronchial thermoplasty (BT). This study uses data from the British Thoracic Society (BTS) UK Difficult Asthma Registry (DAR) and the Hospital Episodes Statistics (HES) database to assess aspects of efficacy and compares these with previous trials.
Methods Lung function (FEV1), quality of life (AQLQ), rescue steroid use, healthcare visits and days lost from work/school were compared at BT baseline and 12 month follow-up in patients for whom DAR data were available. In calculating annualised figures, baseline data were assumed to represent 12 months pre-BT, and 12 month follow-up data were scaled according to the time period that the follow-up represented.
Significance testing for differences in FEV1 and AQLQ used a paired t-test. Differences in event counts were tested using non-parametric bootstrap hypothesis tests.
HES was searched for BT episodes from 1st April 2011 to 31st January 2015. An anonymised matching technique was used to link patients in HES and DAR, and for those whom sufficient time had elapsed since BT, HES A&E attendances were compared in the 12 months pre-BT and the 12 months starting from 30 days post-BT (to exclude any transient increases).
Results 31 patients had 12 month follow-ups in DAR, enabling comparison with BT baseline where data were available. All outcomes from DAR showed improvement at 12 month follow-up compared to BT baseline (Table 1). The mean improvement in AQLQ score (0.92) was smaller than that reported in AIR2 (1.35; n = 190), AIR (1.3; n = 52) and RISA (1.53; n = 15) trials.
From HES, there were 24 A&E attendances (in 5/12 patients) in the 12 months pre-treatment and 15 A&E attendances (in 6/12 patients) in the 12 months post-treatment.
Conclusion To date, efficacy outcomes appear consistent with those observed in previous clinical trials, with a smaller, but statistically significant, improvement in AQLQ score. The reduction in unscheduled healthcare visits and days lost from work/school also reached statistical significance. Although the median number of A&E attendances increased in the 12 patients studied, the annual rate of A&E attendances per patient reduced from 2 to 1.25.