Around 2% of adults with asthma develop pneumonia each year. There is substantial evidence for an increased risk of developing pneumonia in people with the lung disease, chronic obstructive pulmonary disease (COPD). As inhaled steroids are the mainstay of treatment for asthma this could contribute towards their risk of pneumonia. Four studies have previously investigated the risk of pneumonia associated with inhaled steroid use in people with asthma; these studies found inconsistent findings. This is likely in part because the studies had biases. Two of these studies only included a narrow age group of patients, one study did not include pneumonia diagnosed in hospital and all the studies did not adequately account for oral steroid use. Another potential risk factor is the use of a type of drug (proton pump inhibitor, PPI) that reduces stomach acid that causes acid reflux. This is common in asthma and larger proportion of patients use both inhaled steroids and PPIs. There is some suggestion that PPIs cause pneumonia, especially in older people with long-term use, but no study has specifically addressed this in people with asthma.
Population of people with asthma will be identified in the Gold and AURUM database. Patients with COPD, pregnancy, dementia or immunosuppression or previously used inhaled, nasal or oral corticosteroids will be excluded. The primary outcome with be pneumonia, identified either in CPRD, Hospital Episodes Statistics or on death certificate (Office of National Statistics). Secondary outcomes will be lower respiratory tract infection and hospitalised pneumonia only. Primary exposures will be inhaled corticosteroids and proton pump inhibitors. To describe the association between exposures and outcomes, different study designs will be used, survival model using Poisson regression with multivariable adjustment, assessment of drug-drug interaction and on-treatment analysis, case-control with different exposure times measured and self-case controlled series stratified by drug use, drug dosage and types of drugs. Covariates considered in the models will include age, gender, socioeconomic status (IMD), asthma severity (using asthma medication and exacerbations in year prior to study entry), oral corticosteroid use in past three years, BMI, vaccination (influenza and pneumococcal), bronchiectasis, obstructive sleep apnoea, atopy, cardiovascular disease, cerebrovascular disease, renal failure, rheumatic diseases, lung fibrosis, reflux, previous pneumonia, smoking, alcohol, and depression. Restricted cubic splines will be used to assess if there is a non-linear relationship between drug dose and outcomes.
Pneumonia and lower respiratory tract infections.
Chloe Bloom - Chief Investigator - Imperial College London
Chloe Bloom - Corresponding Applicant - Imperial College London
Sriprasad Vikram - Collaborator - Imperial College London
HES Accident and Emergency;HES Admitted Patient Care;ONS Death Registration Data;Patient Level Index of Multiple Deprivation;Practice Level Index of Multiple Deprivation