It is well recognised that people are more likely to have heart attacks and strokes after suffering with pneumonia. We also know that aspirin use can reduce the risk of heart attacks and strokes in certain settings, but not others.
Emerging hospital data suggests that aspirin may have a protective effect, but this has not been tested in a general practice setting.
This study will take a large group of people who develop pneumonia and will see if they are less likely to get heart attacks or strokes if they are on aspirin, and will aim to measure this benefit. Alongside this, we will test a novel statistical approach which tries to improve how accurately we can estimate the benefits of a particular medicine (in this case aspirin).
The study will involve a large cohort of older patients (>50 years old) with pneumonia, identified using the CPRD.
It is well established that pneumonia increases the risk of myocardial infarction (MI) and stroke, both in primary and secondary care datasets. Aspirin use has been shown to be protective in some clinical settings (such as secondary prevention of MI), but not beneficial in others.
We will identify whether or not there is a protective effect of aspirin against MI and stroke in pneumonia using both traditional and novel statistical techniques: multiple variable logistic regression (our primary analysis, and our pre-specified data analysis) and a prior events rate ratio (PERR) analysis.
Our primary analysis will be a multivariable logistic regression, with aspirin use as a covariable. Other relevant confounders available in the CPRD will be included in this model. Alongside this, given the significant risks of confounding, we will perform a PERR analysis, a novel methodology involving self-control of cases within a dataset.
A PERR analysis adjusts for confounders by calculating an event rate prior to the time of diagnosis, and after the time of diagnosis, and comparing the two. In this analysis, we will identify all myocardial infarctions and strokes in the 6 months prior, and 6 months after the diagnosis of pneumonia. We will then compare the ratio of the prior events to the after events in the patients who are taking aspirin as compared to patients not on aspirin.
Primary: A composite of
1. Myocardial Infarction or
2. Stroke
Secondary:
1. Mortality (all cause)
Fergus Hamilton - Chief Investigator - University of Bristol
Fergus Hamilton - Corresponding Applicant - University of Bristol
David Arnold - Collaborator - University of Bristol
Rupert Payne - Collaborator - University of Bristol
William Henley - Collaborator - University of Exeter
HES Admitted Patient Care;ONS Death Registration Data;Patient Level Index of Multiple Deprivation