Differences in patient improvements, stratified by hospital site: a) Asthma Control Questionnaire-6 (ACQ-6), b) exacerbation reduction, c) unscheduled care (emergency department attendances and hospitalisation events) reduction, d) forced expiratory volume in 1 s (FEV1) and e) oral corticosteroid (OCS) discontinuation. All models are adjusted for sex, age category, ethnicity, biologic therapy, year of baseline assessment, hospital site and base level of the relevant outcome of interest. Analysis by site was restricted to hospitals with ≥30 patients per reported outcome. The terms “best” and “worst” refer to the magnitude of change for each outcome measured and are not consistent across panels.