Each patient with asthma experiences the disease differently, which makes it hard to put it into categories. This is important because different types of asthma require different treatment strategies. Some people with asthma have whatÂ’s known as difficult-to-treat asthma. Their symptoms do not go away, even with high doses of asthma inhaler medicines. They experience several asthma attacks per year, requiring treatment with steroid tablets. Some of these patients, in particular those who have a specific type of inflamed lining of the lung with a particular white cell called an eosinophil may benefit from medicines that specifically target these cells. However, it is unknown how clinicians could best distinguish these patients from patients with other types of inflammation underlying their asthma.
This study aims to group patients with similar characteristics of asthma together in order to distinguish different types, including the group that may benefit from new medicines, analysing a large anonymous database of patients with asthma that are registered at general practices.
Aim: This study aims to characterise the group of patients with an eosinophilic phenotype of asthma who do not achieve disease control under current guideline-based therapy.
Design: The study will include a broad population of patients, aged ?13 years, who had active asthma at last data extraction from general practice and ?1 blood eosinophil count recorded after first asthma diagnosis.
Active asthma will be defined as ?1 prescription for a controller or reliever inhaler in the last 12 months.
Patients will be classified into four different grades of likelihood of eosinophilic asthma, based on the highest blood eosinophil count recorded combined with information on clinical features previously reported to be associated with T2-high inflammation, such as age of onset and presence of nasal polyps.
In addition, unsupervised k-means clustering analyses will be performed to identify groups of patients with a similar asthma phenotype. Agreement between both classification methods will be assessed.
Demographics, diagnosed comorbidities, clinical characteristics, such as asthma severity and control and health care resource utilisation (HCRU) will be described for different phenotypes (both predefined and clusters) and for the total population over the last 12 months. Analyses will be repeated among patients who underwent a recent annual review within the quality and outcomes framework using additional information on asthma symptoms and in the subgroup of these patients with severe asthma.
Among patients who had ?3 blood eosinophil counts available, characteristics of those with persistently high blood eosinophil counts will be compared with those of patients with intermittently or never high counts using a separate design. These groups will be characterised and compared in a baseline period and clinical outcomes & HCRU will be compared in a follow-up year, using the most recent eosinophil count as the index date.
The following asthma outcomes will be studied over one full year of data: numbers of asthma exacerbations; acute respiratory events, presence/absence of asthma control and all-cause and asthma-specific events of: hospitalisations; admissions to Accident & Emergency; outpatient visits; physician office visits in primary care
David Price - Chief Investigator - OPRI - Observational and Pragmatic Research Institute Pte Ltd
Marjan Kerkhof - Corresponding Applicant - OPRI - Observational and Pragmatic Research Institute Pte Ltd
Derek Skinner - Collaborator - OPRI - Observational and Pragmatic Research Institute Pte Ltd
Trung Tran - Collaborator - Astra Zeneca Inc - USA
HES Accident and Emergency;HES Admitted Patient Care;HES Outpatient