Proton pump inhibitor use and the risk of myocardial infarction: a nested case-control study in the UK Clinical Practice Research Datalink.

Date of ISAC Approval: 
13/11/2018
Lay Summary: 
Proton pump inhibitors (PPIs), drugs to treat heart burn, were introduced to the market in 1989. Since their introduction, multiple studies have shown that taking PPIs may increase a person's chances of having cardiovascular events such as a heart attack or stroke if they already have serious cardiovascular disease. However, there have been few published studies that evaluated whether relatively healthy people taking PPIs have a higher chance of heart attacks. In this study, we will explore whether PPI use increases the chances of having a heart attack in a healthy population.
Technical Summary: 
Proton pump inhibitors (PPIs) were introduced to the market in 1989 primarily to treat gastroesophageal reflux disease (GERD) and peptic ulcer disease. Since their introduction, multiple studies have shown that exposure to PPIs among high-risk cardiovascular patients with concomitant use of antiplatelet medication (e.g. clopidogrel) is associated with an increased risk of cardiovascular events. However, little is known about whether PPIs increase the risk of myocardial infarction (MI) among generally healthy populations. Our aim is to evaluate the association between PPI use and risk of developing MI among patients in the CPRD. We plan to conduct a nested case-control analysis in a population of PPI users. First, we will identify all users of PPIs within the CPRD. From this population, we will identify all cases who had a first MI and match them to up to 4 controls based on age (+/- 2 years), calendar year (+/- 2 years), sex, general practice, and index date (same date as their matched case). We will use conditional logistic regression to estimate crude and adjusted odds ratios and 95% CIs of MI among PPI users, compared with unexposed patients (patients who had their first PPI exposure after the index date).
Health Outcomes to be Measured: 
The outcome of interest is a diagnosis of MI in the patient clinical or referral record. MI will be identified using Read codes. We will exclude patients whose medical records started less than 180 days prior to the index date, and patients who were younger than 25 and older than 65 years old at the index date, as well as patients with history of old MI, stroke, type 1 diabetes, autoimmune disorders, malignancy, venous thrombus embolism (VTE) before the index date. See appendix for a list of codes. We will assess the validity of the MI codes by identifying supporting evidence of MI in each patient's record, such as codes for hyperlipidemia, atherosclerosis as well as the use of antiplatelet and antihypertensive drugs. We will conduct a sensitivity analysis to evaluate potential misclassification of the MI diagnosis by conducting an analysis restricted to MI cases with supporting treatments or other codes.
Collaborators: 

Dr Susan S Jick - Chief Investigator - BCDSP - Boston Collaborative Drug Surveillance Program
Dr Ellen Qian - Collaborator - BCDSP - Boston Collaborative Drug Surveillance Program
Dr Susan S Jick - Corresponding Applicant - BCDSP - Boston Collaborative Drug Surveillance Program