select search filters
roundups & rapid reactions
factsheets & briefing notes
before the headlines
Fiona fox's blog

Proton Pump Inhibitors (PPIs) and risk of death

A new study published in BMJ Open reports that there is an increased risk of death among proton pump inhibitor users.


Title, Date of Publication & Journal

Risk of death among users of Proton Pump Inhibitors: a longitudinal observational cohort study of United States veterans

Published: July 3rd 2017

BMJ Open


Study’s main claims – and are they supported by the data

The researchers have accessed records on 3.5 million people who are covered by the US veterans system to look at the risks associated with proton pump inhibitors (PPIs, like omeprazole) and H2 blockers (like ranitidine). They report an increased risk of death for people on PPIs than H2s.

This type of study cannot prove that proton pump inhibitors (PPIs) increase the risk of death.  But it does raise good questions for future research to answer.

It’s important not to over-interpret these results.  Patients benefitting from PPIs could do themselves harm by stopping taking the drugs.




Very large dataset…but see below.

Prospectively collected data using hospital systems – i.e. the researchers knew what the hospital knew.


A dataset this huge means that even chance findings can appear very certain – we must be cautious interpreting these results.

Figure 1 suggests that 5-10% of people on PPIs died within 1 year. That suggests the deaths may not be linked to the treatment but to the underlying condition.

Cause and effect is unclear. We don’t know whether the diseases being treated are different for each drug. So again, we can’t be sure that it is the different type of diseases that’s the problem, and not the treatment.

Deaths (and risks) are reported in relative terms – we might be talking about small absolute numbers here.

There were lots of different analyses but none was specified as a main analysis – this raises the risk of finding false positives.

It’s not clear how relevant this US Veterans hospital data would be to NHS patients across the board.


Any specific expertise relevant to studied paper (beyond statistical)?



Before The Headlines is a service provided to the SMC by volunteer statisticians: members of the Royal Statistical Society (RSS), Statisticians in the Pharmaceutical Industry (PSI) and experienced statisticians in academia and research. A list of contributors, including affiliations, is available here.

in this section

filter Headlines by year

search by tag