Patients with conditions of the heart or blood vessels or those with diabetes with complications are at greater risk of cardiovascular related death, heart attack (myocardial infarction), stroke and hospitalisations for heart failure.
Medications to reduce the risk of such cardiovascular events are prescribed based upon evidence from clinical trials which explore treatment effects in a sample of the general population. However, this sample is often not representative of the entire population. Therefore, the effectiveness of medications on groups underrepresented in clinical trials such as those over 75 years or from ethnic minority backgrounds is often unknown.
One of the most commonly prescribed types of medication given to patients at high risk of cardiovascular events are angiotensin-converting-enzyme inhibitors (ACEi), which reduce blood pressure but can cause serious side effects. Based upon results including those from a landmark study (the ONTARGET/TRANSCEND trial) angiotensin II receptor blockers (ARBs) are often now prescribed instead of ACE inhibitors, as in the trials these were shown to perform as well as ACE inhibitors and cause fewer side effects.
Anonymised information routinely collected during general practice (GP) and hospital appointments will be used in this study to assess the effects of ARBs compared to ACE inhibitors, and a combination of the two. First, we will see how well each treatment works when prescribed to people that are similar to those included in the ONTARGET/TRANSCEND trials. Then we will see how well the treatments work when prescribed to patient groups excluded or underrepresented in these trials.
We will use HES linked CPRD data to select patients prescribed ACE inhibitors, ARBs or both and from this select those that are thought to be at high risk of cardiovascular events, i.e. those with coronary, peripheral or cerebrovascular disease or diabetes with end-organ damage. We will then create a population in CPRD with similar characteristics to the ONTARGET/TRANSCEND trial population by applying their inclusion/exclusion trial criteria. We will use multivariable Cox regression to calculate the absolute and relative rates/risks (whichever is more appropriate) for the primary composite outcome of cardiovascular death, non-fatal MI, non-fatal stroke or hospitalisation for congestive heart failure and also the secondary and safety outcomes.
The analysis will be repeated after having applied a matching technique, such as propensity score matching, to ensure that patients in the two comparison groups from both the CPRD cohort and the randomised trials are similar, which we have approved access to anonymised individual participant data. If the absolute and relative rates/risks from analysis of the matched cohort are similar to those from applying the inclusion/exclusion criteria this will suggest that matching has no additional benefit in replicating a clinical trial in this specific therapeutic area.
Either using the CPRD trial-analogous cohort, or a matched cohort (if shown to be important), we will extend the analysis to subgroups that were underrepresented in the trials (adjusting the inclusion/exclusion criteria as necessary) such as those aged >75 years, from ethnic minority groups or with poor renal function. We will use a variety of modelling approaches to determine how to expand the population of trial-underrepresented subgroups within the CPRD population. However, if matching does have an effect in replicating the trial it will also be used here. Finally, we will observe the outcomes over a longer time period than studied in the clinical trial.
The outcomes are listed below, based on the outcomes from the ONTARGET/TRANSCEND trials. How each component of the outcome will be identified in the CPRD cohort is shown in Table 1. in Appendix 1.
Primary outcome: Cardiovascular death, non-fatal MI, non-fatal stroke or hospitalisation for congestive heart failure
Each component of the primary outcome will also be studied as a secondary outcome
Secondary outcomes: Newly diagnosed congestive heart failure, revascularization procedures, and nephropathy
Other outcomes: All-cause mortality and microvascular complications of diabetes mellitus
Safety outcomes: Cough recorded both as a Read Code and adverse event: cough, angioedema, hyperkalaemia or renal dysfunction
Laurie Tomlinson - Chief Investigator - London School of Hygiene & Tropical Medicine ( LSHTM )
Paris Baptiste - Corresponding Applicant - London School of Hygiene & Tropical Medicine ( LSHTM )
Elizabeth Williamson - Collaborator - London School of Hygiene & Tropical Medicine ( LSHTM )
Heide Stirnadel-Farrant - Collaborator - AstraZeneca Ltd - UK Headquarters
Ian Douglas - Collaborator - London School of Hygiene & Tropical Medicine ( LSHTM )
Kevin Wing - Collaborator - London School of Hygiene & Tropical Medicine ( LSHTM )
Liam Smeeth - Collaborator - London School of Hygiene & Tropical Medicine ( LSHTM )
HES Admitted Patient Care;ONS Death Registration Data;Patient Level Index of Multiple Deprivation;Practice Level Index of Multiple Deprivation